Enovari LinkedIn Marketing Strategy
Enovari Marketing Campaign
Total Contacts
0
Completed
0
In Progress
0
Completion Rate
0%
▼
Table of Contents
0 items
▼
1. Company Page Setup & Optimization
7 itemsCreating the Company Page
Medium
Additional Info
1. Go to linkedin.com/company/setup/new 2. Select "Company" (not Showcase Page) 3. Enter: Enovari as the company name 4. Use the LinkedIn URL: linkedin.com/company/enovari 5. Industry: Technology, Information and Internet (primary) or Software Development 6. Company size: 1-10 employees (solo founder counts) 7. Company type: Privately Held
1. Go to linkedin.com/company/setup/new 2. Select "Company" (not Showcase Page) 3. Enter: Enovari as the company name 4. Use the LinkedIn URL: linkedin.com/company/enovari 5. Industry: Technology, Information and Internet (primary) or Software Development 6. Company size: 1-10 employees (solo founder counts) 7. Company type: Privately Held
Visual Assets Needed
Medium
Additional Info
> Fact-check note (April 2026): The company page banner is 1128x191 px. This is correct. However, be aware that LinkedIn occasionally adjusts rendering -- the safe zone for text and key visuals is roughly the center 1000x150 px area. Mobile crops both sides, so keep critical elements centered or left-aligned. The logo is displayed at 300x300 px on the page but should be uploaded at that size or higher for crispness. Left-aligned text (mobile crops the right side) Include the URL: enovari.ai Use a dark, tech-forward color scheme Tagline options for the banner: "AI That Remembers. Across Every Platform." "Persistent, Portable, Structured Memory for AI" "Your AI Finally Remembers Who You Are"
> Fact-check note (April 2026): The company page banner is 1128x191 px. This is correct. However, be aware that LinkedIn occasionally adjusts rendering -- the safe zone for text and key visuals is roughly the center 1000x150 px area. Mobile crops both sides, so keep critical elements centered or left-aligned. The logo is displayed at 300x300 px on the page but should be uploaded at that size or higher for crispness. Left-aligned text (mobile crops the right side) Include the URL: enovari.ai Use a dark, tech-forward color scheme Tagline options for the banner: "AI That Remembers. Across Every Platform." "Persistent, Portable, Structured Memory for AI" "Your AI Finally Remembers Who You Are"
Company Page Description (2,000 character limit)
Medium
Additional Info
``
``
Enovari gives AI persistent, portable, structured memory across every platform.
Today's AI assistants are brilliant but amnesiac. Every conversation starts from zero.
Every context is lost. Every preference forgotten. Enovari solves this.
Our platform provides:
-- Cross-session memory that persists across conversations and platforms
-- Structured knowledge storage with confidence scoring, provenance tracking, and semantic relationships
-- A persona system that lets AI maintain distinct identities with private memory namespaces
-- 140+ API integrations connecting AI memory to the tools you already use
-- Enterprise-grade architecture built on hybrid BM25+vector search with 15-signal retrieval
Whether you're a developer building AI applications, an enterprise deploying AI assistants,
or a power user who wants their AI to actually know them -- Enovari is the memory layer
that makes AI useful over time.
Built by a solo founder. Shipping real code. Not a pitch deck.
Try it: https://enovari.ai
``Phone
(add if available)
Industry
Software Development
Company size
1-10 employees
Headquarters
(your city, or "Remote")
Founded
2025/2026 (use actual year)
Specialties
AI Memory, Persistent Context, AI Infrastructure, MCP Protocol, Knowledge Graphs, AI Personas, Enterprise AI, Developer Tools
AI Memory, Persistent Context, AI Infrastructure, MCP Protocol, Knowledge Graphs, AI Personas, Enterprise AI, Developer Tools
Additional Info
Website: https://enovari.ai Phone: (add if available) Industry: Software Development Company size: 1-10 employees Headquarters: (your city, or "Remote") Founded: 2025/2026 (use actual year) Specialties: AI Memory, Persistent Context, AI Infrastructure, MCP Protocol, Knowledge Graphs, AI Personas, Enterprise AI, Developer Tools
Website: https://enovari.ai Phone: (add if available) Industry: Software Development Company size: 1-10 employees Headquarters: (your city, or "Remote") Founded: 2025/2026 (use actual year) Specialties: AI Memory, Persistent Context, AI Infrastructure, MCP Protocol, Knowledge Graphs, AI Personas, Enterprise AI, Developer Tools
Featured Section
Medium
Additional Info
Pin these to the top of the company page (add as you create them): 1. Product launch announcement (first post) 2. A technical deep-dive (shows credibility) 3. A customer testimonial or case study (social proof) 4. Link to enovari.ai with a compelling description
Pin these to the top of the company page (add as you create them): 1. Product launch announcement (first post) 2. A technical deep-dive (shows credibility) 3. A customer testimonial or case study (social proof) 4. Link to enovari.ai with a compelling description
Growing Company Page Followers
Medium
Additional Info
Invite all personal connections to follow the page Add the company page as your employer on your personal profile Share the company page creation as a personal post ("Just launched the official Enovari page...") Add LinkedIn follow button to enovari.ai website Add company page link to email signature Every personal post that performs well should tag @Enovari Encourage team members / advisors / beta users to follow and share Post consistently from the company page (3-5x/week minimum) Engage with comments on company page posts within 1 hour Use LinkedIn newsletters from the company page (boosts follower invites) Cross-promote on other channels (Twitter/X, email list, product in-app)
Invite all personal connections to follow the page Add the company page as your employer on your personal profile Share the company page creation as a personal post ("Just launched the official Enovari page...") Add LinkedIn follow button to enovari.ai website Add company page link to email signature Every personal post that performs well should tag @Enovari Encourage team members / advisors / beta users to follow and share Post consistently from the company page (3-5x/week minimum) Engage with comments on company page posts within 1 hour Use LinkedIn newsletters from the company page (boosts follower invites) Cross-promote on other channels (Twitter/X, email list, product in-app)
Company Page Content Pillars
Medium
Additional Info
Company page posts should feel distinct from personal profile posts. Use the company page for: Product updates and changelogs -- "Enovari v2.3 just shipped: here's what's new" Technical content and documentation highlights -- "How our 15-signal retrieval system works" Customer spotlights and use cases -- "How [User/Company] uses Enovari for X" Industry data and original research -- "We analyzed 10,000 AI sessions. Here's what we found about memory." Job postings and hiring (even for contractor roles or advisors -- it signals growth) Event announcements -- webinars, LinkedIn Lives, conference appearances > Tip: Company page posts are eligible for "Recommended" placement in the feed. LinkedIn's algorithm increasingly promotes company pages that post consistently (3+ times per week) and receive employee engagement. Having even 2-3 people who like/comment on company posts within the first 30 minutes dramatically increases distribution.
Company page posts should feel distinct from personal profile posts. Use the company page for: Product updates and changelogs -- "Enovari v2.3 just shipped: here's what's new" Technical content and documentation highlights -- "How our 15-signal retrieval system works" Customer spotlights and use cases -- "How [User/Company] uses Enovari for X" Industry data and original research -- "We analyzed 10,000 AI sessions. Here's what we found about memory." Job postings and hiring (even for contractor roles or advisors -- it signals growth) Event announcements -- webinars, LinkedIn Lives, conference appearances > Tip: Company page posts are eligible for "Recommended" placement in the feed. LinkedIn's algorithm increasingly promotes company pages that post consistently (3+ times per week) and receive employee engagement. Having even 2-3 people who like/comment on company posts within the first 30 minutes dramatically increases distribution.
▼
2. Personal Brand Strategy (Founder)
5 itemsWhy the Founder's Personal Brand Matters More Than the Company Page
Medium
Additional Info
On LinkedIn, people follow people, not logos. The founder's personal profile will generate 5-10x more engagement than the company page. This is especially true for: Solo founders (the story IS the brand) Technical products (credibility comes from the builder) AI/tech (the audience wants to learn from practitioners) The founder's LinkedIn profile IS the primary marketing channel. The company page is secondary. > Fact-check note: The "5-10x" figure is directionally correct and widely cited. LinkedIn's own data from their 2024 B2B Marketing Benchmark report showed that employee posts average 2-5x the engagement of company page posts, with founder/CEO posts often reaching 8-12x in the <50 employee bracket. The exact multiplier depends on your follower base and content quality, but the principle is well established.
On LinkedIn, people follow people, not logos. The founder's personal profile will generate 5-10x more engagement than the company page. This is especially true for: Solo founders (the story IS the brand) Technical products (credibility comes from the builder) AI/tech (the audience wants to learn from practitioners) The founder's LinkedIn profile IS the primary marketing channel. The company page is secondary. > Fact-check note: The "5-10x" figure is directionally correct and widely cited. LinkedIn's own data from their 2024 B2B Marketing Benchmark report showed that employee posts average 2-5x the engagement of company page posts, with founder/CEO posts often reaching 8-12x in the <50 employee bracket. The exact multiplier depends on your follower base and content quality, but the principle is well established.
Profile Optimization
Medium
Additional Info
The headline appears everywhere -- in search results, comments, connection requests. It must communicate value, not just a title. Bad: "Founder & CEO at Enovari" Good: "Building Enovari -- Persistent Memory for AI | Solo Founder | Shipping the infrastructure layer AI is missing" Also good: "Solo founder giving AI the memory it's been missing | Building Enovari (enovari.ai)" Professional but approachable High resolution, good lighting Face takes up 60-70% of the frame Solid or simple background No sunglasses, no group photos cropped 1584x396 px Include Enovari branding + one-line value prop Include URL: enovari.ai Should complement, not duplicate, the company page banner > Fact-check note (April 2026): The personal profile banner dimension is 1584x396 px. This is correct as of early 2026. LinkedIn recommends this size and will crop anything outside it. The safe zone for text is approximately the center 1200x300 px area due to profile photo overlap on the left side (desktop) and varying mobile crops. Structure it as a story, not a resume. Use this framework: ``
The headline appears everywhere -- in search results, comments, connection requests. It must communicate value, not just a title. Bad: "Founder & CEO at Enovari" Good: "Building Enovari -- Persistent Memory for AI | Solo Founder | Shipping the infrastructure layer AI is missing" Also good: "Solo founder giving AI the memory it's been missing | Building Enovari (enovari.ai)" Professional but approachable High resolution, good lighting Face takes up 60-70% of the frame Solid or simple background No sunglasses, no group photos cropped 1584x396 px Include Enovari branding + one-line value prop Include URL: enovari.ai Should complement, not duplicate, the company page banner > Fact-check note (April 2026): The personal profile banner dimension is 1584x396 px. This is correct as of early 2026. LinkedIn recommends this size and will crop anything outside it. The safe zone for text is approximately the center 1200x300 px area due to profile photo overlap on the left side (desktop) and varying mobile crops. Structure it as a story, not a resume. Use this framework: ``
I'm building Enovari because AI has a memory problem.
Every AI conversation starts from zero. Your assistant doesn't remember your preferences,
your projects, your decisions, or the context you spent an hour explaining yesterday.
This isn't a feature gap -- it's a fundamental architectural flaw.
Enovari fixes it.
We provide persistent, portable, structured memory for AI across every platform.
Not just "chat history" -- real structured knowledge with confidence scoring,
provenance tracking, semantic relationships, and a 15-signal hybrid retrieval system.
The technical details:
Cross-session memory that works across Claude, ChatGPT, and any MCP-compatible client
Knowledge graphs with 10,000+ nodes and 100,000+ typed edges
A persona system where AI maintains distinct identities with private memory
140+ API integrations
Enterprise-grade architecture, solo-founder speed
I'm a solo founder building this from scratch. No co-founder. No team of 50.
Just real code shipping to real users.
If you're building AI applications, deploying AI in your enterprise, or just tired
of your AI forgetting who you are -- I'd love to connect.
Try Enovari: https://enovari.ai
Topics I post about:
-- AI memory and persistent context
-- Solo founder journey (real talk, not hustle porn)
-- Technical deep dives into AI infrastructure
-- The future of AI-to-AI communication
-- Building in public
DM me about: partnerships, enterprise pilots, or if you just want to talk about
where AI memory is headed.
``
> Fact-check note: The About section character limit is 2,600 characters. This is correct as of 2025-2026. LinkedIn briefly tested a 2,000-character limit but reverted. Always verify in the editor, as LinkedIn occasionally adjusts limits without announcement.
Pin 3-6 pieces of content:
1. Your best-performing LinkedIn post
2. A link to enovari.ai
3. A technical blog post or article
4. A product demo video (if available)
5. Press coverage or podcast appearances
6. A "founding story" post
Current: Founder & CEO at Enovari (link to company page)
Write 2-3 sentences about what you're building and why
Previous roles: Include relevant technical/entrepreneurial experiencePositioning as a Thought Leader
Medium
Additional Info
1. Solo founder building AI infrastructure -- rare and compelling 2. "AI has a memory problem" -- a clear, resonant thesis that everyone who uses AI can relate to 3. Technical depth -- you can explain HOW the system works, not just WHAT it does 4. Building in public -- transparency about the journey, metrics, challenges 5. Vision for the machine web -- a big, original idea about AI-to-AI knowledge networks
1. Solo founder building AI infrastructure -- rare and compelling 2. "AI has a memory problem" -- a clear, resonant thesis that everyone who uses AI can relate to 3. Technical depth -- you can explain HOW the system works, not just WHAT it does 4. Building in public -- transparency about the journey, metrics, challenges 5. Vision for the machine web -- a big, original idea about AI-to-AI knowledge networks
Content Strategy: What to Post
Medium
Additional Info
40% Text-only posts (highest organic reach per LinkedIn's algorithm) 25% Image + text posts (screenshots, diagrams, whiteboard photos) 20% Carousel posts / PDF documents (high save rate, good for tutorials) 10% Video posts (product demos, talking head insights) 5% Articles / newsletters (long-form, lower reach but higher authority) > Fact-check note (April 2026): The format distribution is approximately correct. Text-only posts continue to have the highest average organic reach on LinkedIn. However, as of late 2025 and into 2026, LinkedIn has been giving increased weight to carousel/document posts and native video (especially short-form video under 90 seconds). Carousels may now rival or exceed text-only for total engagement in some niches. Newsletters get guaranteed push notifications to all subscribers, making them extremely high-value despite lower feed reach. The key update: LinkedIn now deprioritizes "engagement bait" style text posts more aggressively than in 2024, rewarding genuine expertise and originality.
40% Text-only posts (highest organic reach per LinkedIn's algorithm) 25% Image + text posts (screenshots, diagrams, whiteboard photos) 20% Carousel posts / PDF documents (high save rate, good for tutorials) 10% Video posts (product demos, talking head insights) 5% Articles / newsletters (long-form, lower reach but higher authority) > Fact-check note (April 2026): The format distribution is approximately correct. Text-only posts continue to have the highest average organic reach on LinkedIn. However, as of late 2025 and into 2026, LinkedIn has been giving increased weight to carousel/document posts and native video (especially short-form video under 90 seconds). Carousels may now rival or exceed text-only for total engagement in some niches. Newsletters get guaranteed push notifications to all subscribers, making them extremely high-value despite lower feed reach. The key update: LinkedIn now deprioritizes "engagement bait" style text posts more aggressively than in 2024, rewarding genuine expertise and originality.
Growing from 0 to 10K Followers
Medium
Additional Info
Phase 1: Foundation (0-500 followers) -- Weeks 1-6 Actions: Connect with everyone you know (aim for 500+ 1st connections) Post daily, 5 days per week minimum Comment on 20-30 posts per day from larger accounts in AI/tech Engage authentically -- add value in comments, not just "Great post!" Join 10-15 relevant LinkedIn groups and participate Turn on Creator Mode on your profile Add "Follow" as the default button (Creator Mode does this) Connect with every AI founder, developer, and decision-maker you can find > Fact-check note (April 2026): Creator Mode was officially retired by LinkedIn in early 2024. LinkedIn rolled Creator Mode features into all profiles. As of 2025-2026, ALL profiles now have the Follow button as default (you can switch it back to Connect in settings). The "Topics you post about" feature (hashtags displayed on your profile from Creator Mode) has also been removed. What remains: everyone now has access to LinkedIn Live, newsletters, and the follow-first profile button by default. Action: Do not look for a "Creator Mode" toggle -- it no longer exists. Instead, verify that your profile default action is set to "Follow" in Settings > Visibility > Profile default action button. Phase 2: Momentum (500-2,000 followers) -- Weeks 7-16 Actions: Continue daily posting, refine based on what's working Start carousel posts (these get saved and shared more) Engage with engagement pods or accountability groups (optional, use carefully) Begin LinkedIn newsletter (subscribers are notified of each edition) Guest appearances on LinkedIn Live or Audio Events Cross-promote LinkedIn content on Twitter/X, email, etc. Connect with journalists and analysts covering AI Share specific metrics: "We just hit X users" or "Our retrieval system scores Y" Phase 3: Authority (2,000-10,000 followers) -- Weeks 17-40 Actions: Start LinkedIn newsletter if you haven't (massive distribution advantage) Collaborate with other founders / thought leaders (co-posts, interviews) Publish data-driven posts (benchmarks, comparisons, original research) Create a signature series (e.g., "AI Memory Weekly" or "Solo Founder Diaries") Amplify top posts with LinkedIn ads ($5-10/day budget) Speak at virtual events and post clips Write LinkedIn articles for SEO-style long-form content Repurpose content: one idea becomes a post, a carousel, a video, and an article Commenting on big accounts is the #1 underrated growth tactic -- a good comment on a post with 50K views puts your name in front of 50K people Posting within 30 minutes of a major AI news story (GPT-5 announcement, etc.) rides the wave Tagging people thoughtfully (not spam-tagging) increases distribution Asking questions in posts increases comment count, which increases reach
Phase 1: Foundation (0-500 followers) -- Weeks 1-6 Actions: Connect with everyone you know (aim for 500+ 1st connections) Post daily, 5 days per week minimum Comment on 20-30 posts per day from larger accounts in AI/tech Engage authentically -- add value in comments, not just "Great post!" Join 10-15 relevant LinkedIn groups and participate Turn on Creator Mode on your profile Add "Follow" as the default button (Creator Mode does this) Connect with every AI founder, developer, and decision-maker you can find > Fact-check note (April 2026): Creator Mode was officially retired by LinkedIn in early 2024. LinkedIn rolled Creator Mode features into all profiles. As of 2025-2026, ALL profiles now have the Follow button as default (you can switch it back to Connect in settings). The "Topics you post about" feature (hashtags displayed on your profile from Creator Mode) has also been removed. What remains: everyone now has access to LinkedIn Live, newsletters, and the follow-first profile button by default. Action: Do not look for a "Creator Mode" toggle -- it no longer exists. Instead, verify that your profile default action is set to "Follow" in Settings > Visibility > Profile default action button. Phase 2: Momentum (500-2,000 followers) -- Weeks 7-16 Actions: Continue daily posting, refine based on what's working Start carousel posts (these get saved and shared more) Engage with engagement pods or accountability groups (optional, use carefully) Begin LinkedIn newsletter (subscribers are notified of each edition) Guest appearances on LinkedIn Live or Audio Events Cross-promote LinkedIn content on Twitter/X, email, etc. Connect with journalists and analysts covering AI Share specific metrics: "We just hit X users" or "Our retrieval system scores Y" Phase 3: Authority (2,000-10,000 followers) -- Weeks 17-40 Actions: Start LinkedIn newsletter if you haven't (massive distribution advantage) Collaborate with other founders / thought leaders (co-posts, interviews) Publish data-driven posts (benchmarks, comparisons, original research) Create a signature series (e.g., "AI Memory Weekly" or "Solo Founder Diaries") Amplify top posts with LinkedIn ads ($5-10/day budget) Speak at virtual events and post clips Write LinkedIn articles for SEO-style long-form content Repurpose content: one idea becomes a post, a carousel, a video, and an article Commenting on big accounts is the #1 underrated growth tactic -- a good comment on a post with 50K views puts your name in front of 50K people Posting within 30 minutes of a major AI news story (GPT-5 announcement, etc.) rides the wave Tagging people thoughtfully (not spam-tagging) increases distribution Asking questions in posts increases comment count, which increases reach
▼
4. LinkedIn Outreach Strategy
7 itemsIdentifying Target Audiences
Medium
LinkedIn Search Techniques
Medium
Additional Info
People > Title: "AI Engineer" + Location: United States People > Company: [target companies like Anthropic, OpenAI, Langchain, etc.] People > Industry: Technology, Information and Internet ``
People > Title: "AI Engineer" + Location: United States People > Company: [target companies like Anthropic, OpenAI, Langchain, etc.] People > Industry: Technology, Information and Internet ``
"AI engineer" OR "machine learning engineer" AND "memory" OR "context"
`
`
"CTO" OR "technical founder" AND "AI" OR "artificial intelligence"
`
`
"head of AI" OR "AI director" OR "chief AI officer"
`
`
"developer tools" OR "devtools" AND "AI" OR "LLM"
`
`
"AI agent" OR "AI assistant" AND "engineer" OR "developer"
`
`
"knowledge graph" OR "vector database" OR "RAG" AND "engineer"
`
`
"MCP" OR "model context protocol" AND "developer" OR "engineer"
`
`
"AI infrastructure" OR "ML platform" AND "founder" OR "CTO"
`
`
"prompt engineer" OR "AI ops" OR "LLMOps"
`
`
"conversational AI" OR "chatbot" AND "product manager" OR "lead"
`
`
"enterprise AI" AND "director" OR "VP" OR "head of"
`
`
"AI startup" AND "founder" OR "co-founder" AND NOT "stealth"
``
> Practical note on LinkedIn Boolean search: LinkedIn's free search supports basic Boolean (AND, OR, NOT, quotes for exact phrases, parentheses for grouping). However, free search results are limited to ~1,000 results and filtering is restricted. Known quirk: LinkedIn's free search sometimes ignores Boolean operators if the query is too complex. Keep queries to 2-3 operators for best results. For more advanced searches, LinkedIn Sales Navigator is dramatically more powerful. Also note: LinkedIn search quality varies -- results are influenced by your network, location, and LinkedIn's own relevance algorithm, not just Boolean logic.
Lead filters: Job title + company size + industry + geography
Account filters: Company revenue + headcount + tech stack
Saved searches with alerts for new matches
Dramatically more powerful than free LinkedIn search
ROI calculation: If Sales Navigator helps close even 5 extra users/month at $19.99, it pays for itself
> Fact-check note (April 2026): LinkedIn Sales Navigator pricing has changed. As of 2025-2026:
> - Sales Navigator Core: ~$99.99/month (billed annually at ~$79.99/mo)
> - Sales Navigator Advanced: ~$149.99/month (billed annually)
> - Sales Navigator Advanced Plus: custom enterprise pricing
>
> Prices vary by region and LinkedIn frequently runs promotions. A free 30-day trial is typically available for new users. The ROI math in the document is reasonable: at $19.99/mo per user, you need ~5 attributable sign-ups per month to justify the cost.
AI labs: Anthropic, OpenAI, Google DeepMind, Meta AI, Mistral, Cohere
AI infra: LangChain, LlamaIndex, Pinecone, Weaviate, Chroma, Replit
Enterprise AI users: Salesforce, HubSpot, Notion, Slack
AI-forward startups: anyone building AI agents, copilots, assistants
Dev tool companies: Vercel, Supabase, Railway, Fly.io
Consulting firms with AI practices: McKinsey, BCG, Deloitte, Accenture
AI agent/assistant companies: Dust, Fixie, Letta (formerly MemGPT), Mem0, Zep
MCP ecosystem companies: Any company building MCP servers or clients
AI coding tools: Cursor, Windsurf, GitHub Copilot team, Sourcegraph, Tabnine
AI data/infra platforms: Databricks, Snowflake, Scale AI, Weights & BiasesConnection Request Messages
Medium
Additional Info
LinkedIn allows 300 characters for connection request notes Personalization dramatically increases acceptance rates (40% vs 20%) Never pitch in the connection request -- just connect Reference something specific: their content, their company, their role > Fact-check note (April 2026): The 300-character limit for connection request notes is correct. However, it is worth noting that LinkedIn Premium subscribers can send connection requests with no note limit (up to ~2,000 characters). For free accounts, it is strictly 300 characters. The acceptance rate figures (40% personalized vs. 20% blank) are approximate averages from various LinkedIn outreach studies. Actual rates depend heavily on your profile quality, mutual connections, and the relevance of your message. Some studies show personalized requests at 50-60% acceptance vs. 20-30% for blank requests. ``
LinkedIn allows 300 characters for connection request notes Personalization dramatically increases acceptance rates (40% vs 20%) Never pitch in the connection request -- just connect Reference something specific: their content, their company, their role > Fact-check note (April 2026): The 300-character limit for connection request notes is correct. However, it is worth noting that LinkedIn Premium subscribers can send connection requests with no note limit (up to ~2,000 characters). For free accounts, it is strictly 300 characters. The acceptance rate figures (40% personalized vs. 20% blank) are approximate averages from various LinkedIn outreach studies. Actual rates depend heavily on your profile quality, mutual connections, and the relevance of your message. Some studies show personalized requests at 50-60% acceptance vs. 20-30% for blank requests. ``
Hi [Name] -- I saw your work on [specific project/post]. I'm building Enovari,
a persistent memory layer for AI systems. Would love to connect and exchange ideas
on AI infrastructure. - [Your Name]
`
`
Hi [Name] -- I'm a solo founder building AI memory infrastructure (enovari.ai).
Your work at [Company] is impressive. Would value connecting with a fellow builder.
[Your Name]
`
`
[Name] -- Really enjoyed your post about [topic]. I think a lot about AI memory
and persistent context. Would love to connect and continue the conversation.
`
`
Hi [Name] -- I'm researching how teams at [Company] think about AI persistence
and memory. Building in this space myself. Would love to connect. - [Your Name]
`
`
Hi [Name] -- Your post about context window limits hit home. I'm building the
memory layer that solves this problem (Enovari). Would love to connect and compare
notes. - [Your Name]
`
`
Hi [Name] -- Love what you're doing for the [Company] developer community.
I'm building AI memory infrastructure and would value your perspective.
Happy to connect? - [Your Name]
`
`
Hi [Name] -- Noticed we both follow the AI infrastructure space closely.
I'm building Enovari (persistent memory for AI). Always looking to connect
with people who care about this problem. - [Your Name]
`
`
Hi [Name] -- Saw your talk/panel at [Event]. Great insights on [topic].
I'm building in the AI memory space. Would love to connect. - [Your Name]
`
`
[Name] -- I see you're working with MCP/AI tool integrations. Building a
memory layer on MCP at Enovari. Would love to compare approaches.
`
`
Hi [Name] -- [Mutual Connection] suggested we connect. I'm building persistent
memory for AI (enovari.ai). Sounds like our work overlaps. - [Your Name]
``Post-Connection Engagement Sequence
Medium
Day 1 (connection accepted)
Do nothing. Or send a brief "Thanks for connecting!"
Days 2-7
Engage with their content. Like and comment on 2-3 of their posts.
Day 7-14
Send a value-first message:
Day 21+ (if warm)
Offer a demo, free trial, or early access:
Day 14-21 (if they replied)
Continue the conversation naturally. Share a relevant insight, article, or your own post. Only mention Enovari if it's directly relevant to what they said.
Continue the conversation naturally. Share a relevant insight, article, or your own post. Only mention Enovari if it's directly relevant to what they said.
If they never replied (Day 14+)
Do NOT send a follow-up pitch. Instead, continue engaging with their content. They will see your name and headline repeatedly. If they're interested, they'll reach out. Patience converts better than persistence.
Do NOT send a follow-up pitch. Instead, continue engaging with their content. They will see your name and headline repeatedly. If they're interested, they'll reach out. Patience converts better than persistence.
Additional Info
Do NOT immediately pitch after someone accepts your connection. Follow this sequence: ``
Do NOT immediately pitch after someone accepts your connection. Follow this sequence: ``
Hey [Name] -- been enjoying your posts about [topic].
Quick question: in your experience building [AI products/deploying AI],
how do you handle persistent context across sessions? It's something
I think about constantly as I build Enovari.
No pitch -- genuinely curious about your approach.
`
`
Based on our conversation, I think you'd find Enovari interesting.
Would you want to try it? Happy to set you up with [free trial /
extended trial / walkthrough]. No pressure either way.
``InMail Strategy (LinkedIn Premium / Sales Navigator)
Medium
Additional Info
InMails have a ~10-25% response rate when done well. You get 5-50/month depending on your plan. > Fact-check note (April 2026): InMail response rates vary widely. LinkedIn's own data suggests average InMail response rates of ~10-18%. Well-targeted, personalized InMails can reach 20-25%. Spray-and-pray InMails drop to 3-5%. Credit allocation: > - Premium Career: 5 InMail credits/month > - Premium Business: 15 InMail credits/month > - Sales Navigator Core: 50 InMail credits/month > - Recruiter Lite: 30 InMail credits/month > > Important: LinkedIn refunds InMail credits if the recipient responds within 90 days (accept or decline). This incentivizes quality messages. Unused credits roll over for up to 3 months (then expire). Subject line: Short, specific, curiosity-driven (under 40 characters) Body: 3-4 sentences max Always include a specific reason you're reaching out to THEM End with a low-commitment ask (not "Let's schedule a call") Send InMails on Tuesday-Thursday mornings (same peak times as posts) Do not use InMail templates/bulk sending -- personalization is everything Reference their content, company, or recent activity -- prove you did your homework Keep it under 400 words -- LinkedIn's own research shows shorter InMails get higher response rates Never attach files or include multiple links -- it looks like spam ``
InMails have a ~10-25% response rate when done well. You get 5-50/month depending on your plan. > Fact-check note (April 2026): InMail response rates vary widely. LinkedIn's own data suggests average InMail response rates of ~10-18%. Well-targeted, personalized InMails can reach 20-25%. Spray-and-pray InMails drop to 3-5%. Credit allocation: > - Premium Career: 5 InMail credits/month > - Premium Business: 15 InMail credits/month > - Sales Navigator Core: 50 InMail credits/month > - Recruiter Lite: 30 InMail credits/month > > Important: LinkedIn refunds InMail credits if the recipient responds within 90 days (accept or decline). This incentivizes quality messages. Unused credits roll over for up to 3 months (then expire). Subject line: Short, specific, curiosity-driven (under 40 characters) Body: 3-4 sentences max Always include a specific reason you're reaching out to THEM End with a low-commitment ask (not "Let's schedule a call") Send InMails on Tuesday-Thursday mornings (same peak times as posts) Do not use InMail templates/bulk sending -- personalization is everything Reference their content, company, or recent activity -- prove you did your homework Keep it under 400 words -- LinkedIn's own research shows shorter InMails get higher response rates Never attach files or include multiple links -- it looks like spam ``
Subject: AI memory question for you
Hi [Name],
I noticed your work on [specific thing] at [Company]. I'm building Enovari --
a persistent memory layer for AI systems (think: your AI finally remembers
everything across sessions and platforms).
We're solving the problem where AI assistants lose all context between
conversations. I'd love to know: is this a pain point your team has run into?
Either way, happy to share what we've learned building the retrieval system.
It's been a wild ride.
Best,
[Your Name]
`
`
Subject: persistent context for your AI stack
Hi [Name],
I saw [Company] is expanding its AI capabilities (congrats on [specific hire/announcement]).
Quick question: when your team deploys AI assistants internally, how do you handle
the fact that they lose all context between sessions?
We built Enovari to solve exactly this -- persistent memory that works across
any AI platform. Would love to hear how you're thinking about this problem.
[Your Name]
enovari.ai
`
`
Subject: how do you handle AI memory?
Hey [Name],
Your [repo/project/post] on [topic] caught my eye. Impressive work.
I'm building Enovari -- cross-platform persistent memory for AI.
Think: structured knowledge graphs with BM25+vector hybrid retrieval,
not just dumping chat logs into a vector DB.
If you're interested in AI memory architecture, I wrote about our
approach -- happy to share. No sales pitch, just nerd stuff.
[Your Name]
`
`
Subject: re: your post on [AI frustration]
Hi [Name],
Your post about [specific frustration with AI] resonated. This is literally
what I'm trying to fix.
I'm building Enovari -- persistent, structured memory for AI that works
across platforms. Your AI actually remembers who you are, what you're working on,
and what you told it last week.
Would you be interested in trying it? No cost, no strings.
[Your Name]
``Finding People Who Would Benefit from Enovari
Medium
Additional Info
Search posts containing: "AI context window" OR "AI forgets" OR "chat history" OR "AI memory" Look for people complaining about having to re-explain things to AI Monitor hashtags: #AIAgents #LLM #ChatGPT #Claude #AIAssistant Follow threads under major AI announcements (people discuss limitations) "Building AI agent" in posts or profiles "LangChain" OR "LlamaIndex" OR "MCP" in profiles (these devs need memory) People who follow Anthropic, OpenAI, LangChain company pages GitHub profiles linked from LinkedIn (check their repos for AI projects) "Evaluating AI tools" or "AI stack" in posts Job postings for AI roles (the company is investing in AI) Posts about enterprise AI deployment challenges CTOs writing about their tech stack decisions People who mention Mem0, Zep, MemGPT/Letta, or any AI memory competitor People who post about "AI memory" solutions and their limitations Comments on competitor product announcements expressing unmet needs Profiles mentioning "MCP," "Model Context Protocol," or "tool-use" Profiles with "RAG" or "retrieval augmented generation" (they understand the space and may want something better) People posting about vector databases who may be ready for structured memory Attendees of AI conferences (NeurIPS, ICML, AI Engineer Summit) Members of AI-focused Slack/Discord communities who also have LinkedIn presence Speakers at AI meetups and webinars
Search posts containing: "AI context window" OR "AI forgets" OR "chat history" OR "AI memory" Look for people complaining about having to re-explain things to AI Monitor hashtags: #AIAgents #LLM #ChatGPT #Claude #AIAssistant Follow threads under major AI announcements (people discuss limitations) "Building AI agent" in posts or profiles "LangChain" OR "LlamaIndex" OR "MCP" in profiles (these devs need memory) People who follow Anthropic, OpenAI, LangChain company pages GitHub profiles linked from LinkedIn (check their repos for AI projects) "Evaluating AI tools" or "AI stack" in posts Job postings for AI roles (the company is investing in AI) Posts about enterprise AI deployment challenges CTOs writing about their tech stack decisions People who mention Mem0, Zep, MemGPT/Letta, or any AI memory competitor People who post about "AI memory" solutions and their limitations Comments on competitor product announcements expressing unmet needs Profiles mentioning "MCP," "Model Context Protocol," or "tool-use" Profiles with "RAG" or "retrieval augmented generation" (they understand the space and may want something better) People posting about vector databases who may be ready for structured memory Attendees of AI conferences (NeurIPS, ICML, AI Engineer Summit) Members of AI-focused Slack/Discord communities who also have LinkedIn presence Speakers at AI meetups and webinars
Weekly Outreach Cadence
Medium
Additional Info
To avoid LinkedIn restrictions while maintaining consistent outreach: > Important limits: LinkedIn restricts accounts that send too many connection requests too quickly. The generally safe threshold is 80-100 connection requests per week for accounts with 500+ connections. New accounts or accounts with low acceptance rates should stay under 50/week. If your acceptance rate drops below 30%, slow down -- LinkedIn may restrict your account. See Section 12 for details on restrictions.
To avoid LinkedIn restrictions while maintaining consistent outreach: > Important limits: LinkedIn restricts accounts that send too many connection requests too quickly. The generally safe threshold is 80-100 connection requests per week for accounts with 500+ connections. New accounts or accounts with low acceptance rates should stay under 50/week. If your acceptance rate drops below 30%, slow down -- LinkedIn may restrict your account. See Section 12 for details on restrictions.
▼
5. LinkedIn Groups
3 itemsRelevant Groups to Join
Medium
Additional Info
> Fact-check note on group sizes (April 2026): LinkedIn group member counts fluctuate constantly and many groups have renamed, merged, or become inactive over time. The group names and approximate sizes listed above are estimates based on the most popular groups in these categories as of late 2025. LinkedIn does not provide a public directory of all groups sorted by size. Before joining, search for the exact group name on LinkedIn and verify it is (a) still active, (b) has recent posts, and (c) is not overrun with spam. Many large LinkedIn groups (100K+ members) have very low engagement because they are poorly moderated. A 20K-member group with active daily discussions is more valuable than a 500K-member group full of link spam. > > Known issue: Several group names listed above are generic enough that multiple groups with similar names may exist. When searching, look for the largest and most actively moderated version.
> Fact-check note on group sizes (April 2026): LinkedIn group member counts fluctuate constantly and many groups have renamed, merged, or become inactive over time. The group names and approximate sizes listed above are estimates based on the most popular groups in these categories as of late 2025. LinkedIn does not provide a public directory of all groups sorted by size. Before joining, search for the exact group name on LinkedIn and verify it is (a) still active, (b) has recent posts, and (c) is not overrun with spam. Many large LinkedIn groups (100K+ members) have very low engagement because they are poorly moderated. A 20K-member group with active daily discussions is more valuable than a 500K-member group full of link spam. > > Known issue: Several group names listed above are generic enough that multiple groups with similar names may exist. When searching, look for the largest and most actively moderated version.
Group Engagement Strategy
Medium
Additional Info
1. Never drop links to Enovari in groups. This gets you banned instantly. 2. Give value first. Answer questions, share insights, help people. 3. Establish expertise over 2-4 weeks before mentioning anything about your product. 4. When you do mention Enovari, make it contextual: "We ran into this exact problem while building our AI memory system..." 5. Post original insights, not promotional content. 6. Engage with other posts -- don't just post your own. Week 1-2: Just comment on others' posts. Add value. Be helpful. Week 3-4: Share an insight or lesson learned (no product mention). Week 5+: Occasionally reference your work at Enovari when contextually relevant. Ongoing: 3:1 ratio -- 3 pure-value contributions for every 1 that mentions your product. "I've been researching AI memory systems. Here's what I found about [specific technical topic]..." Answer a question someone asked about AI architecture / persistent context Share a benchmark or data point: "We tested X retrieval approaches and here's what performed best..." React to AI news with a unique take Ask genuine questions: "How is everyone handling cross-session context in their AI deployments?"
1. Never drop links to Enovari in groups. This gets you banned instantly. 2. Give value first. Answer questions, share insights, help people. 3. Establish expertise over 2-4 weeks before mentioning anything about your product. 4. When you do mention Enovari, make it contextual: "We ran into this exact problem while building our AI memory system..." 5. Post original insights, not promotional content. 6. Engage with other posts -- don't just post your own. Week 1-2: Just comment on others' posts. Add value. Be helpful. Week 3-4: Share an insight or lesson learned (no product mention). Week 5+: Occasionally reference your work at Enovari when contextually relevant. Ongoing: 3:1 ratio -- 3 pure-value contributions for every 1 that mentions your product. "I've been researching AI memory systems. Here's what I found about [specific technical topic]..." Answer a question someone asked about AI architecture / persistent context Share a benchmark or data point: "We tested X retrieval approaches and here's what performed best..." React to AI news with a unique take Ask genuine questions: "How is everyone handling cross-session context in their AI deployments?"
Group Posting Rules (Common Across Most Groups)
Medium
Additional Info
Most LinkedIn groups have these rules: No self-promotion or direct advertising No link-only posts Posts must be relevant to the group topic No repetitive posting of the same content Engage respectfully; no flame wars Some groups require admin approval before posts go live Share your knowledge, not your product link When someone asks about AI memory/context, answer thoroughly and mention "I've been working on this problem" (people will check your profile) Your profile does the selling -- make sure it's optimized Share blog posts from enovari.ai/blog (educational content, not sales pages) > Practical note on LinkedIn Groups in 2025-2026: LinkedIn Groups have declined significantly in utility compared to 2018-2020. Many large groups are overrun with spam and have low engagement. LinkedIn has deprioritized groups in the algorithm -- group posts rarely appear in the main feed. However, groups remain useful for: > - Discovery: Finding people to connect with (browse active group members) > - Credibility signal: Being a member of relevant groups appears on your profile > - Niche communities: Smaller, well-moderated groups (5K-30K members) often have better engagement than mega-groups > - Direct access: You can message group members directly even without a 1st-degree connection > > Strategy adjustment: Treat groups primarily as a source of connection targets rather than a content distribution channel. Browse group discussions to find people asking questions you can answer, then connect with them and engage on their main-feed posts instead.
Most LinkedIn groups have these rules: No self-promotion or direct advertising No link-only posts Posts must be relevant to the group topic No repetitive posting of the same content Engage respectfully; no flame wars Some groups require admin approval before posts go live Share your knowledge, not your product link When someone asks about AI memory/context, answer thoroughly and mention "I've been working on this problem" (people will check your profile) Your profile does the selling -- make sure it's optimized Share blog posts from enovari.ai/blog (educational content, not sales pages) > Practical note on LinkedIn Groups in 2025-2026: LinkedIn Groups have declined significantly in utility compared to 2018-2020. Many large groups are overrun with spam and have low engagement. LinkedIn has deprioritized groups in the algorithm -- group posts rarely appear in the main feed. However, groups remain useful for: > - Discovery: Finding people to connect with (browse active group members) > - Credibility signal: Being a member of relevant groups appears on your profile > - Niche communities: Smaller, well-moderated groups (5K-30K members) often have better engagement than mega-groups > - Direct access: You can message group members directly even without a 1st-degree connection > > Strategy adjustment: Treat groups primarily as a source of connection targets rather than a content distribution channel. Browse group discussions to find people asking questions you can answer, then connect with them and engage on their main-feed posts instead.
▼
6. Content Ideas: 30+ Post Templates
10 itemsFounding Story Posts
Medium
Additional Info
Post 1: The Origin Story Hook: Opens with the "solo founder, no co-founder" pattern interrupt that performs extremely well on LinkedIn. Why it works for Enovari: The solo founder building infrastructure angle is rare, authentic, and relatable. It combines underdog narrative with technical credibility. Full draft: ``
Post 1: The Origin Story Hook: Opens with the "solo founder, no co-founder" pattern interrupt that performs extremely well on LinkedIn. Why it works for Enovari: The solo founder building infrastructure angle is rare, authentic, and relatable. It combines underdog narrative with technical credibility. Full draft: ``
I'm a solo founder building AI memory infrastructure.
No co-founder. No VC. No team of engineers.
Just me, a thesis that AI needs persistent memory,
and an obsessive need to prove it works.
Here's how it started:
I was using Claude for a complex project. I'd spent two hours
giving it context about my architecture, my preferences,
my constraints. The conversation was productive.
Then my session ended.
Next day, I opened a new conversation. Brand new.
Zero context. I had to explain everything again.
From scratch. Like we'd never met.
That's when it hit me: this isn't a minor annoyance.
This is a fundamental flaw in how AI works.
LLMs have no memory. Every session is a blank slate.
Every user is a stranger. Every conversation starts from zero.
So I started building the memory layer.
6 months later, Enovari stores [X] memories
across [Y] sessions with [Z]% accuracy.
It works across Claude, ChatGPT, and any MCP-compatible client.
Not "chat history." Real structured knowledge with
confidence scoring and semantic relationships.
The hardest part wasn't the technology.
It was believing a solo founder could build
infrastructure that matters.
What's a bet you've made on yourself that
others thought was crazy?
#SoloFounder #BuildInPublic #AIMemory
`
Post 2: The "Why I Quit" Story
Hook: "I quit [X] to build an AI memory platform. People thought I was insane." Taps into career-risk narrative that LinkedIn loves.
Full draft:
`
I quit [previous role/stable income] to build
an AI memory platform.
People thought I was insane.
"AI companies have billions in funding."
"You can't compete alone."
"Memory? That's a feature, not a company."
Here's what they were wrong about:
1. AI companies have billions but they're focused
on models, not memory. Nobody owns the persistence layer.
2. Solo founders can move faster than teams of 50.
I shipped a complete memory system in months,
not quarters.
3. Memory isn't a feature. It's infrastructure.
Just like databases aren't a "feature" of applications.
Memory sits underneath everything else.
4. The market is creating itself. Every AI user
who's ever had to re-explain their project
is a potential customer.
5. Being solo means every decision is fast.
No committees. No consensus-building. Ship, learn, iterate.
Today: [specific milestone or metric]
The people who told me I was crazy?
Some of them are now asking for early access.
What's the best "bad" advice you've ever ignored?
#StartupLife #AIInfrastructure #SoloFounder
`
Post 3: The Hardest Day
Hook: "The hardest day of building Enovari wasn't technical." Subverts expectation (people expect technical challenges from a tech founder).
Full draft:
`
The hardest day of building Enovari wasn't technical.
It was the Tuesday in [month] when I realized I'd been
building the wrong feature for three weeks.
Not wrong technically. The code worked fine.
Wrong strategically. Nobody needed it.
I'd built an elegant solution to a problem
I imagined my users had.
But when I talked to actual users?
They wanted something completely different.
Three weeks of work. Deleted.
Here's what I learned:
1. Your first instinct about what users want
is almost always wrong. Talk to them BEFORE building.
2. Deleting code is not failure. It's iteration.
The willingness to throw away work is a superpower.
3. Solo founders have no buffer. When you build
the wrong thing, there's nobody else to blame
and nobody else to pick up the slack.
That deleted feature taught me more about
building products than any successful ship.
Every founder has a "I almost quit" day.
What was yours?
#FounderLife #BuildInPublic #Resilience
``Technical / Educational Posts
Medium
Additional Info
Post 4: AI's Memory Problem (Core Thesis) Hook: "Your AI assistant is brilliant. And it has amnesia." Two short sentences that create instant cognitive dissonance. Why this is Enovari's signature post: This should be the first post you publish. It establishes the core thesis. Repost a variation of this every 6-8 weeks. Full draft: ``
Post 4: AI's Memory Problem (Core Thesis) Hook: "Your AI assistant is brilliant. And it has amnesia." Two short sentences that create instant cognitive dissonance. Why this is Enovari's signature post: This should be the first post you publish. It establishes the core thesis. Repost a variation of this every 6-8 weeks. Full draft: ``
Your AI assistant is brilliant.
And it has amnesia.
Every single conversation starts from zero.
Every preference. Forgotten.
Every context. Gone.
Every hour you spent explaining your project. Wasted.
This is not a minor inconvenience.
It is a fundamental architectural flaw.
LLMs were designed for single-turn reasoning.
Nobody built the memory layer.
So I did.
Enovari gives AI persistent, structured memory
across every session and every platform.
Not "chat history."
Real knowledge. With confidence scores.
With provenance. With relationships.
AI without memory is a genius with amnesia.
It's time to fix that.
What's the most frustrating thing your AI
keeps forgetting?
#ArtificialIntelligence #AIMemory #Innovation
`
Post 5: Technical Deep Dive
Hook: "Most people think AI 'memory' = storing chat logs. That's like saying human memory = a tape recorder." Analogy-driven hook that reframes understanding.
Why it works: Establishes Enovari's technical differentiation in accessible language. The numbered list format drives dwell time and saves.
Full draft:
`
Most people think AI "memory" = storing chat logs.
That's like saying human memory = a tape recorder.
Real AI memory needs:
1. STRUCTURE -- Not raw text. Typed knowledge units
with topics, confidence, provenance, and taxonomy.
2. RETRIEVAL -- Not keyword search. Hybrid BM25 +
vector search with 15 ranking signals.
3. RELATIONSHIPS -- Not flat storage. Knowledge graphs
with typed, weighted, bidirectional edges.
4. PERSISTENCE -- Not session-scoped. Memory that
survives across conversations, platforms, and models.
5. IDENTITY -- Not one-size-fits-all. Personas with
private memory namespaces and behavioral profiles.
This is what we built at Enovari.
And it actually works.
Which of these surprises you most?
#AIArchitecture #MachineLearning #DevTools
`
Post 6: "Everyone Gets This Wrong"
Hook: "3 things everyone gets wrong about AI memory." List + contrarian framing is high-engagement on LinkedIn.
Full draft:
`
3 things everyone gets wrong about AI memory:
1. "Just increase the context window."
Context windows are RAM, not memory.
You don't solve memory by giving someone
a bigger desk. You solve it with a filing system.
2. "RAG is memory."
RAG retrieves documents. Memory retrieves
knowledge -- with confidence, provenance,
relationships, and temporal awareness.
3. "Memory is a feature, not a product."
That's like saying databases are a feature.
Memory is infrastructure. It sits under
everything else.
I'm building the infrastructure layer.
That's Enovari.
What would you add to this list?
#AI #LLM #TechDebate
`
Post 7: MCP Protocol Explainer
Hook: "MCP (Model Context Protocol) is the most important AI standard nobody is talking about." Bold claim drives curiosity clicks.
Full draft:
`
MCP (Model Context Protocol) is the most important
AI standard nobody is talking about.
Here's why it matters:
MCP is an open protocol that lets AI models
talk to external tools and data sources
through a standardized interface.
Think of it like USB for AI.
Before USB, every device had a proprietary connector.
Printers, keyboards, cameras -- all different.
USB created one standard. Everything just worked.
MCP does the same for AI. One protocol that lets
any AI client connect to any compatible service.
Why this is huge:
-- No vendor lock-in. Your AI memory works with
Claude, ChatGPT, or any MCP client.
-- No custom integrations. Build once, connect everywhere.
-- Composability. AI can use multiple tools in one
conversation through a single standard.
Enovari is built on MCP. It means any AI client
that speaks MCP can access persistent memory
through our platform.
No custom integrations.
No vendor lock-in.
One protocol. Every AI.
If you're building AI applications,
MCP should be on your radar.
Questions? I've been deep in this for months.
#MCP #AIInfrastructure #OpenStandards
`
Post 7b: Knowledge Graph Explainer (NEW)
Hook: "Your AI stores information in flat text. That's like storing a library by dumping all the pages into one pile."
Full draft:
`
Your AI stores information in flat text.
That's like storing a library by dumping
all the pages into one pile.
When we built Enovari's memory system,
we didn't just store text. We built
a knowledge graph.
What does that mean in practice?
Every piece of knowledge has:
-- A topic and category
-- A confidence score (how sure are we?)
-- A provenance trail (where did this come from?)
-- Typed relationships to other knowledge
-- Temporal metadata (when was this true?)
When AI retrieves a memory, it doesn't just
find a matching string.
It traverses relationships.
It weighs confidence.
It considers recency.
It understands context.
The result: AI that doesn't just "remember"
-- it UNDERSTANDS what it remembers.
This is the difference between a tape recorder
and a brain.
What questions do you have about knowledge graphs
in AI? I'll answer in the comments.
#KnowledgeGraph #AIArchitecture #AIMemory
``Insight / Opinion Posts
Medium
Additional Info
Post 8: Industry Prediction Full draft: ``
Post 8: Industry Prediction Full draft: ``
Prediction: By 2028, "memoryless AI" will be
as unacceptable as "offline software."
Today, we tolerate AI that forgets everything
between sessions.
In 2 years, users will refuse to use an AI
that doesn't remember them.
Just like we refuse to use software
that doesn't sync to the cloud.
Memory will be table stakes.
The companies building memory infrastructure
now will own that layer.
Agree or disagree?
#FutureOfAI #ArtificialIntelligence #Prediction
`
Post 9: Hot Take on AI News
Template with Enovari-specific guidance:
`
[React to the latest AI announcement/news]
Everyone is talking about [the news].
But nobody is asking the real question:
[Your unique angle related to AI memory]
Here's my take:
[3-4 paragraphs with your analysis,
tying it back to the memory problem]
What do you think?
#AI #[relevant hashtag]
`
Example: Reacting to a new model release:
`
[Company] just released [Model Name].
Bigger context window. Better reasoning. Faster responses.
But it still forgets everything the moment
you close the tab.
Every model release follows the same pattern:
more intelligence, zero persistence.
It's like building faster cars with no fuel tank.
Impressive in bursts. Useless for journeys.
The next breakthrough in AI won't come from
a bigger model.
It'll come from a model that remembers.
What matters more to you: a smarter AI
or one that knows who you are?
#AI #[ModelName] #AIMemory
`
Post 10: Contrarian View
`
Unpopular opinion:
Most AI startups are building the wrong thing.
They're competing to build better models.
Better chatbots. Better copilots.
But none of them are solving the real problem:
AI has no memory.
Every "AI assistant" is a goldfish
with a PhD.
The next breakthrough in AI won't be
a bigger model.
It will be a model that remembers.
Fight me in the comments.
#AI #Startups #HotTake
`
Post 11: The Machine Web Vision
`
I think the next internet won't be built for humans.
Not in a dystopian way. In an infrastructure way.
The World Wide Web is a presentation layer
-- HTML, CSS, JavaScript -- designed for human eyes
reading rendered pages.
But AI doesn't need rendered pages.
AI needs structured, queryable knowledge.
Imagine an internet where:
-- Every "website" is a queryable knowledge node
-- Every "link" is a typed, weighted relationship
-- Every "search" is a structured query, not keyword matching
-- Every piece of information has provenance and confidence
This isn't science fiction.
Pieces of this exist TODAY.
At Enovari, we've built structured memory with
10,000+ knowledge nodes and 100,000+ typed connections.
It's a single node on a future network.
What happens when there are millions?
#FutureOfAI #Web3 #Innovation
`
Post 11b: The "AI Agents Need Memory" Post (NEW)
Hook: "Everyone is building AI agents. Almost nobody is giving them memory." Timely take on the agent trend.
`
Everyone is building AI agents.
Almost nobody is giving them memory.
Think about what we're doing:
We're deploying autonomous AI systems that make
decisions, take actions, and interact with users...
...and then wiping their brain clean
after every interaction.
An agent without memory is an employee
with amnesia who shows up every morning
having forgotten everything from yesterday.
You'd never hire that person.
But that's what we're building.
AI agents need:
-- Memory of past interactions with each user
-- Memory of their own decisions and outcomes
-- Memory of what worked and what didn't
-- Memory of the context they're operating in
This is what I'm building at Enovari.
Persistent memory that makes agents
actually useful over time.
If you're building AI agents,
this is the missing piece.
What's the biggest memory-related challenge
in your agent stack?
#AIAgents #ArtificialIntelligence #AIMemory
``Behind-the-Scenes / Build in Public
Medium
Additional Info
Post 12: Metrics Update ``
Post 12: Metrics Update ``
Enovari monthly update -- [Month] numbers:
-- [X] active users (up/down from last month)
-- [X] memories stored
-- [X]% retrieval accuracy
-- [X] API integrations live
-- $[X] MRR
What went well:
[2-3 bullets]
What went wrong:
[2-3 bullets -- be honest]
What I'm building next:
[2-3 bullets]
Solo founder transparency.
No vanity metrics.
Just real numbers.
#BuildInPublic #SaaS #StartupMetrics
`
Post 13: Tool Stack / How I Build
Full Enovari-specific draft:
`
My solo founder tool stack for building
an AI memory platform:
Code: TypeScript + Python
Database: PostgreSQL + vector extensions
Search: Hybrid BM25 + vector (custom-built)
Hosting: [your choices -- Vercel/Railway/AWS/etc.]
AI: Claude (primary), GPT-4 (testing)
Design: Figma + Canva
Analytics: PostHog / Plausible
Marketing: LinkedIn (you're reading it) + email
Payments: Stripe
Monitoring: [your choices]
Total monthly cost: $[X]
The entire Enovari platform -- 15-signal retrieval,
knowledge graphs, persona system, 140+ integrations --
is built and maintained by one person.
You don't need a team of 50
to build real infrastructure.
You need the right tools and
an unreasonable amount of focus.
What's in your build stack?
#SoloFounder #TechStack #BuildInPublic
`
Post 14: Lessons Learned
`
Things I've learned building an AI startup alone
that nobody warned me about:
1. Talking to users is terrifying and essential.
2. Your first architecture will be wrong. Ship it anyway.
3. Nobody cares about your product. They care about their problem.
4. Marketing takes 10x longer than you think.
5. The loneliest moment is when the code works and nobody knows.
Which of these hits hardest?
#FounderLessons #StartupLife #SoloFounder
`
Post 14b: Revenue Milestone Post (NEW)
Template for when you hit milestones:
`
$[X] MRR.
Not from a launch day spike.
Not from a viral tweet.
From [X] people paying $19.99/month because
the product actually solves their problem.
Here's the breakdown:
-- [X]% organic (LinkedIn, search, word of mouth)
-- [X]% referral (users telling other users)
-- [X]% content (blog posts, newsletter)
What I've learned about growing a
$19.99/mo product:
1. Enterprise pricing gets all the glory,
but consumer SaaS teaches you to build
products people genuinely love.
2. Every dollar of MRR from a real user
is worth 100x the "potential revenue"
in a pitch deck.
3. Solo founder + real revenue + real users
= the best position in tech right now.
Next milestone: $[X] MRR.
#BuildInPublic #SaaS #MRR
``Engagement / Question Posts
Medium
Additional Info
Post 15: Simple Question ``
Post 15: Simple Question ``
Quick question for AI builders:
When your AI agent needs to "remember" something
from a previous conversation, what do you do?
A) Stuff it in the system prompt
B) RAG from a vector database
C) Store in a custom database
D) Hope the context window is big enough
E) Cry
Genuinely curious how people solve this today.
#AI #DevTools #Poll
`
Post 16: Fill-in-the-blank
`
The biggest problem with AI assistants today is _________.
Wrong answers only.
(Just kidding. I want the real ones too.)
#AI #ArtificialIntelligence
`
Post 17: "This or That"
`
AI hot takes -- where do you stand?
Bigger context windows vs. Better memory systems?
Open source vs. Closed source AI?
AI agents vs. AI copilots?
Prompt engineering vs. Fine-tuning?
Drop your answers below.
I'll share mine in the comments.
#AI #TechDebate
`
Post 17b: "Day in the Life" Engagement Post (NEW)
`
What does a solo AI founder's day actually look like?
6:30 AM -- Coffee. Check metrics. Existential dread (brief).
7:00 AM -- Code. Deep work on retrieval system.
10:00 AM -- User feedback review. Adjust roadmap.
11:00 AM -- Marketing (writing this post).
12:00 PM -- Lunch. Read AI papers.
1:00 PM -- More code. Bug fixes. New features.
3:00 PM -- Support tickets. User conversations.
4:00 PM -- LinkedIn engagement. Comments. Outreach.
5:00 PM -- Architecture planning for tomorrow.
6:00 PM -- "Done" for the day (narrator: he was not done).
9:00 PM -- "Just one more feature..." (every night).
The truth about solo founding:
it's not 80 hours of coding.
It's 80 hours of EVERYTHING.
What does your workday look like?
#SoloFounder #FounderLife #BuildInPublic
``Social Proof & Validation Posts
Medium
Additional Info
Post 18: First Customer / User Story Full draft with Enovari-specific framing: ``
Post 18: First Customer / User Story Full draft with Enovari-specific framing: ``
Our first user said something that
stopped me in my tracks:
"[Actual quote from a user about their
experience with Enovari]"
Here's the context:
They were building an AI assistant for their team.
Every time someone on their team started a new
conversation, the assistant had no idea who they were,
what project they were working on, or what had been
decided in previous sessions.
They tried stuffing context into system prompts.
They tried RAG with a vector database.
They tried manual "memory" via saved prompts.
None of it worked well.
Then they tried Enovari.
Within a week, their AI assistant remembered:
-- Each team member's role and preferences
-- Active project details and decisions
-- Past conversations and outcomes
-- Technical constraints and requirements
"It's like the AI actually joined the team,"
they said.
Building for real users > building for pitch decks.
#CustomerSuccess #BuildInPublic #AI
`
Post 19: Benchmark / Data Post
`
We just ran [benchmark name] on Enovari
against [alternatives/baseline].
Results:
[Table or list of specific metrics]
The takeaway:
[1-2 sentences about what this means]
Full methodology in the comments.
#AIBenchmarks #DataDriven #AIMemory
`
Post 20: Integration / Partnership Announcement
`
Enovari now works with [Platform/Tool].
What this means:
[2-3 bullet points about what users can do]
This brings our total integrations to [X].
Our thesis: AI memory should work everywhere
you work. Not just one platform.
What integration should we build next?
#ProductUpdate #AI #DevTools
``Thought-Provoking / Viral-Optimized Posts
Medium
Additional Info
Post 21: Pattern Interrupt ``
Post 21: Pattern Interrupt ``
I have no co-founder.
No investors.
No team.
No office.
No safety net.
I have:
-- A thesis about AI memory
-- 200,000+ lines of working code
-- A product that actual humans use
-- The stubbornness of a mule
That's enough.
#SoloFounder #Startups #Grit
`
Post 22: "Before and After"
This is one of Enovari's highest-potential post formats -- the concrete demo of what the product does.
`
AI conversation WITHOUT memory:
User: "I'm working on a React project."
AI: "Tell me about your React project."
[next session]
User: "Can you help with my project?"
AI: "What project? Tell me about it."
[next session]
User: "Remember my React app?"
AI: "I don't have memory of previous conversations."
AI conversation WITH Enovari:
User: "Can you help with my project?"
AI: "Sure -- picking up where we left off with your
React dashboard. Last time we fixed the auth flow
and you wanted to add data visualization next.
Your preferred stack is TypeScript + D3.
Want to start with the chart component?"
That's the difference.
Memory isn't optional. It's everything.
#AI #UX #ProductDesign
`
Post 23: Analogy Post
`
Imagine hiring a brilliant consultant.
Every morning they arrive with no memory
of the previous day.
No idea who you are.
No idea what you discussed.
No idea what they recommended.
You'd fire them immediately.
But that's exactly what every AI
assistant does today.
We wouldn't tolerate amnesia in a human hire.
Why do we tolerate it in AI?
#ArtificialIntelligence #FutureOfWork
`
Post 24: List Post
`
10 things AI should remember about you
(but currently doesn't):
1. Your name and role
2. Your current projects
3. Your tech stack preferences
4. Your communication style
5. Your past decisions and why you made them
6. Your team structure
7. Your deadlines
8. The context from last conversation
9. What worked and what didn't
10. What you explicitly asked it to remember
Enovari handles all 10.
What would you add to this list?
#AI #Personalization #Memory
`
Post 24b: "The Real Competition" Post (NEW)
`
People ask me: "Who's your competition?"
My real answer:
Not Mem0.
Not Zep.
Not another AI memory startup.
My competition is the system prompt.
90% of AI builders "solve" memory by
stuffing context into the system prompt.
It works at small scale.
It's a nightmare at real scale.
System prompts have:
-- Token limits (you run out of space)
-- No structure (just a wall of text)
-- No confidence scoring (everything is equally "true")
-- No relationships (flat, not graph)
-- No cross-platform portability (locked to one client)
Real memory needs infrastructure.
Not a longer system prompt.
That's what Enovari is.
How are you handling AI memory today?
System prompt? Vector DB? Something else?
#AIMemory #DevTools #AIInfrastructure
``Carousel Post Ideas
Medium
Post 25
"The 5 Layers of AI Memory" (architectural explainer)
Post 26
"Solo Founder Survival Guide: 10 Lessons" (founder content)
Post 27
"AI Memory vs. RAG vs. Context Windows: A Comparison" (educational)
Post 28
"How to Give Your AI Agent Persistent Memory" (tutorial)
Post 29
"The History of Memory: From Human Brain to AI" (storytelling)
Post 30
"Building an AI Startup With $0 in Funding" (founder journey)
Additional Info
Slide outline: 1. Cover: "The 5 Layers of AI Memory -- What most AI systems are missing" 2. Layer 1: Capture -- How knowledge enters the system (not just chat logs) 3. Layer 2: Structure -- Typed knowledge units with metadata, confidence, provenance 4. Layer 3: Store -- Knowledge graphs with typed relationships and edges 5. Layer 4: Retrieve -- Hybrid BM25 + vector search with multi-signal ranking 6. Layer 5: Apply -- Contextual recall that adapts to the current conversation 7. "Most AI stops at Layer 1" -- comparison diagram 8. "Enovari implements all 5 layers" -- summary 9. CTA: Follow for more AI architecture content / visit enovari.ai Slide outline: 1. Cover: "10 Lessons From Building an AI Startup Alone" 2. Lesson 1: Your first architecture will be wrong. Ship it anyway. 3. Lesson 2: Talk to users before you build features. 4. Lesson 3: Marketing is not optional. Start on Day 1. 5. Lesson 4: The loneliest days are the most productive. 6. Lesson 5: $1 of real revenue > $1M of hypothetical revenue. 7. Lesson 6: Build in public. The accountability is the feature. 8. Lesson 7: Your product is not your baby. Kill features ruthlessly. 9. Lesson 8: Consistency beats intensity. Show up every day. 10. Lesson 9: Nobody is coming to save you. That's the superpower. 11. Lesson 10: The unfair advantage of one person: speed. 12. CTA: Follow for the real story of building Enovari. Slide outline: 1. Cover: "AI Memory vs. RAG vs. Context Windows -- What's the Difference?" 2. Context Windows: What they are (token limit per session) 3. Context Windows: Limitations (session-scoped, no structure, expensive to fill) 4. RAG: What it is (retrieve documents to augment prompts) 5. RAG: Limitations (document-level, no confidence scoring, no relationships) 6. Persistent Memory: What it is (structured knowledge that persists across sessions) 7. Persistent Memory: Advantages (structure, confidence, relationships, cross-platform) 8. Side-by-side comparison table 9. "When to use which" decision flowchart 10. CTA: Enovari provides the memory layer. Visit enovari.ai
Slide outline: 1. Cover: "The 5 Layers of AI Memory -- What most AI systems are missing" 2. Layer 1: Capture -- How knowledge enters the system (not just chat logs) 3. Layer 2: Structure -- Typed knowledge units with metadata, confidence, provenance 4. Layer 3: Store -- Knowledge graphs with typed relationships and edges 5. Layer 4: Retrieve -- Hybrid BM25 + vector search with multi-signal ranking 6. Layer 5: Apply -- Contextual recall that adapts to the current conversation 7. "Most AI stops at Layer 1" -- comparison diagram 8. "Enovari implements all 5 layers" -- summary 9. CTA: Follow for more AI architecture content / visit enovari.ai Slide outline: 1. Cover: "10 Lessons From Building an AI Startup Alone" 2. Lesson 1: Your first architecture will be wrong. Ship it anyway. 3. Lesson 2: Talk to users before you build features. 4. Lesson 3: Marketing is not optional. Start on Day 1. 5. Lesson 4: The loneliest days are the most productive. 6. Lesson 5: $1 of real revenue > $1M of hypothetical revenue. 7. Lesson 6: Build in public. The accountability is the feature. 8. Lesson 7: Your product is not your baby. Kill features ruthlessly. 9. Lesson 8: Consistency beats intensity. Show up every day. 10. Lesson 9: Nobody is coming to save you. That's the superpower. 11. Lesson 10: The unfair advantage of one person: speed. 12. CTA: Follow for the real story of building Enovari. Slide outline: 1. Cover: "AI Memory vs. RAG vs. Context Windows -- What's the Difference?" 2. Context Windows: What they are (token limit per session) 3. Context Windows: Limitations (session-scoped, no structure, expensive to fill) 4. RAG: What it is (retrieve documents to augment prompts) 5. RAG: Limitations (document-level, no confidence scoring, no relationships) 6. Persistent Memory: What it is (structured knowledge that persists across sessions) 7. Persistent Memory: Advantages (structure, confidence, relationships, cross-platform) 8. Side-by-side comparison table 9. "When to use which" decision flowchart 10. CTA: Enovari provides the memory layer. Visit enovari.ai
LinkedIn Newsletter Ideas
Medium
Additional Info
"The Memory Layer" -- AI memory insights, weekly "Solo Founder Diaries" -- behind the scenes of building Enovari "AI Infrastructure Weekly" -- broader AI infra coverage 1. "Why AI Has an Amnesia Problem (And Why Nobody Is Fixing It)" 2. "The Technical Architecture of AI Memory Systems" 3. "Month 1 Building Enovari: Numbers, Lessons, Failures" 4. "RAG Is Not Memory: Why the Industry Is Confused" 5. "The Machine Web: An Internet Built for AI" 6. "How We Built a 15-Signal Retrieval System" 7. "Every AI Tool I Tried Before Building My Own" 8. "What Enterprise AI Gets Wrong About Context" (See Section 9 for full newsletter strategy.)
"The Memory Layer" -- AI memory insights, weekly "Solo Founder Diaries" -- behind the scenes of building Enovari "AI Infrastructure Weekly" -- broader AI infra coverage 1. "Why AI Has an Amnesia Problem (And Why Nobody Is Fixing It)" 2. "The Technical Architecture of AI Memory Systems" 3. "Month 1 Building Enovari: Numbers, Lessons, Failures" 4. "RAG Is Not Memory: Why the Industry Is Confused" 5. "The Machine Web: An Internet Built for AI" 6. "How We Built a 15-Signal Retrieval System" 7. "Every AI Tool I Tried Before Building My Own" 8. "What Enterprise AI Gets Wrong About Context" (See Section 9 for full newsletter strategy.)
Additional Post Ideas (Quick-fire)
Medium
Additional Info
31. Share a screenshot of your code/terminal with a lesson 32. React to a competitor's launch with a thoughtful analysis 33. Share a book/paper that influenced your approach to AI memory 34. Post about a bug that taught you something important 35. "AMA: Ask me anything about AI memory systems" 36. Share a user's feature request and your response 37. Post about what you'd do differently if starting over 38. Repost an AI news article with your unique commentary 39. Share a "day in the life" of a solo founder 40. Post a short demo video of Enovari in action 41. Share a "failure resume" -- things that didn't work and what you learned 42. Post a "myth vs. reality" comparison about AI memory 43. React to an academic paper on memory/retrieval systems 44. Share the specific metrics that matter to you and why 45. Post about a feature you chose NOT to build and why 46. Share a conversation with a user that changed your roadmap 47. "What I wish I knew before starting an AI company" 48. Post a 60-second screen recording of Enovari in action 49. Share your reading list for AI infrastructure 50. "The most underrated AI paper nobody read" (with your commentary)
31. Share a screenshot of your code/terminal with a lesson 32. React to a competitor's launch with a thoughtful analysis 33. Share a book/paper that influenced your approach to AI memory 34. Post about a bug that taught you something important 35. "AMA: Ask me anything about AI memory systems" 36. Share a user's feature request and your response 37. Post about what you'd do differently if starting over 38. Repost an AI news article with your unique commentary 39. Share a "day in the life" of a solo founder 40. Post a short demo video of Enovari in action 41. Share a "failure resume" -- things that didn't work and what you learned 42. Post a "myth vs. reality" comparison about AI memory 43. React to an academic paper on memory/retrieval systems 44. Share the specific metrics that matter to you and why 45. Post about a feature you chose NOT to build and why 46. Share a conversation with a user that changed your roadmap 47. "What I wish I knew before starting an AI company" 48. Post a 60-second screen recording of Enovari in action 49. Share your reading list for AI infrastructure 50. "The most underrated AI paper nobody read" (with your commentary)
▼
7. LinkedIn Ads (Future Phase)
6 itemsShould Enovari Run LinkedIn Ads?
Medium
Additional Info
LinkedIn Ads are the most expensive social media ads. The economics are challenging for a low-price consumer/prosumer product. Here's the math: > Fact-check note (April 2026): These LinkedIn Ads benchmarks are approximately correct but lean toward the higher end. More precise ranges based on 2025 data: > - CPC: $5-14 for B2B tech targeting (can go higher for senior titles) > - CPM: $30-90 (varies dramatically by audience size and targeting specificity) > - CTR: 0.35-0.65% for Sponsored Content (LinkedIn's own benchmark is ~0.44%) > - Conversion rate: 2-5% is reasonable for landing page conversions > - Cost per acquisition: $100-400+ for SaaS is typical on LinkedIn > > The document's conclusion that LinkedIn Ads are poor unit economics for a $19.99/mo product is correct. At $200 CAC and $19.99/mo, payback period is ~10 months assuming no churn. This only works if annual retention is very high (>80%). Verdict: LinkedIn Ads are NOT recommended as a primary acquisition channel at $19.99/mo. Amplifying organic content that's already performing well ($5-10/day) Retargeting website visitors (cheap, high-intent) Promoting lead magnets (free tools, whitepapers, webinars) B2B enterprise pipeline (if you launch enterprise pricing at $99-999/mo) Building brand awareness in a targeted professional audience
LinkedIn Ads are the most expensive social media ads. The economics are challenging for a low-price consumer/prosumer product. Here's the math: > Fact-check note (April 2026): These LinkedIn Ads benchmarks are approximately correct but lean toward the higher end. More precise ranges based on 2025 data: > - CPC: $5-14 for B2B tech targeting (can go higher for senior titles) > - CPM: $30-90 (varies dramatically by audience size and targeting specificity) > - CTR: 0.35-0.65% for Sponsored Content (LinkedIn's own benchmark is ~0.44%) > - Conversion rate: 2-5% is reasonable for landing page conversions > - Cost per acquisition: $100-400+ for SaaS is typical on LinkedIn > > The document's conclusion that LinkedIn Ads are poor unit economics for a $19.99/mo product is correct. At $200 CAC and $19.99/mo, payback period is ~10 months assuming no churn. This only works if annual retention is very high (>80%). Verdict: LinkedIn Ads are NOT recommended as a primary acquisition channel at $19.99/mo. Amplifying organic content that's already performing well ($5-10/day) Retargeting website visitors (cheap, high-intent) Promoting lead magnets (free tools, whitepapers, webinars) B2B enterprise pipeline (if you launch enterprise pricing at $99-999/mo) Building brand awareness in a targeted professional audience
When to Start LinkedIn Ads
Medium
LinkedIn Ad Types
Medium
Additional Info
> Update (April 2026): LinkedIn has introduced Thought Leader Ads (the ability to sponsor posts from personal profiles, not just company pages). This is highly relevant for Enovari -- you can boost your founder's best-performing personal posts as ads, which get significantly higher engagement than traditional company page ads because they look organic. Available in Campaign Manager under Sponsored Content > "Browse existing content" and selecting employee posts.
> Update (April 2026): LinkedIn has introduced Thought Leader Ads (the ability to sponsor posts from personal profiles, not just company pages). This is highly relevant for Enovari -- you can boost your founder's best-performing personal posts as ads, which get significantly higher engagement than traditional company page ads because they look organic. Available in Campaign Manager under Sponsored Content > "Browse existing content" and selecting employee posts.
Targeting Options
Medium
Additional Info
LinkedIn has the best B2B targeting of any platform: Job title (exact or similar) Job function (Engineering, IT, Product, etc.) Seniority level (VP, Director, Manager, etc.) Company name (target specific companies) Company size (1-10, 11-50, 51-200, 201-500, 500+) Company industry Skills (AI, Machine Learning, Python, etc.) Groups (target members of specific LinkedIn groups) Education (school, degree, field of study) Interests (inferred from behavior) Location (country, region, city) Matched Audiences (website retargeting, email list upload, lookalikes) Audience 1: AI Developers Job titles: AI Engineer, ML Engineer, Software Engineer Skills: Artificial Intelligence, Machine Learning, Python, LLM Seniority: IC to Senior Company size: 11-5000 Audience 2: AI Decision Makers Job titles: CTO, VP Engineering, Head of AI, Director of Engineering Industry: Technology, Software Company size: 51-5000 Seniority: Director+ Audience 3: Enterprise AI Buyers Job function: Information Technology Skills: AI Strategy, Digital Transformation Seniority: VP+ Company size: 500+ Audience 4: Website Retargeting People who visited enovari.ai but didn't sign up Exclude existing customers Show them testimonials, demos, or trial offers
LinkedIn has the best B2B targeting of any platform: Job title (exact or similar) Job function (Engineering, IT, Product, etc.) Seniority level (VP, Director, Manager, etc.) Company name (target specific companies) Company size (1-10, 11-50, 51-200, 201-500, 500+) Company industry Skills (AI, Machine Learning, Python, etc.) Groups (target members of specific LinkedIn groups) Education (school, degree, field of study) Interests (inferred from behavior) Location (country, region, city) Matched Audiences (website retargeting, email list upload, lookalikes) Audience 1: AI Developers Job titles: AI Engineer, ML Engineer, Software Engineer Skills: Artificial Intelligence, Machine Learning, Python, LLM Seniority: IC to Senior Company size: 11-5000 Audience 2: AI Decision Makers Job titles: CTO, VP Engineering, Head of AI, Director of Engineering Industry: Technology, Software Company size: 51-5000 Seniority: Director+ Audience 3: Enterprise AI Buyers Job function: Information Technology Skills: AI Strategy, Digital Transformation Seniority: VP+ Company size: 500+ Audience 4: Website Retargeting People who visited enovari.ai but didn't sign up Exclude existing customers Show them testimonials, demos, or trial offers
Recommended Ad Strategy (When Ready)
Medium
Additional Info
Phase 1: Content Amplification ($150/month) Boost your top 3 organic posts per month Budget: $5/day on each for 10 days Target: AI Developers audience Goal: Followers + awareness Phase 2: Retargeting ($300/month) Install LinkedIn Insight Tag on enovari.ai Retarget visitors with demo videos and testimonials Budget: $10/day Goal: Sign-ups from warm traffic Phase 3: Lead Generation ($500-1000/month) Only if enterprise tier exists ($99+/mo) Lead Gen Forms with whitepaper or free tool download Target: AI Decision Makers audience Goal: Enterprise demo requests
Phase 1: Content Amplification ($150/month) Boost your top 3 organic posts per month Budget: $5/day on each for 10 days Target: AI Developers audience Goal: Followers + awareness Phase 2: Retargeting ($300/month) Install LinkedIn Insight Tag on enovari.ai Retarget visitors with demo videos and testimonials Budget: $10/day Goal: Sign-ups from warm traffic Phase 3: Lead Generation ($500-1000/month) Only if enterprise tier exists ($99+/mo) Lead Gen Forms with whitepaper or free tool download Target: AI Decision Makers audience Goal: Enterprise demo requests
Minimum Budget Recommendation
Medium
Bottom line for Enovari today
Spend $0-300/month on ads. Invest the real effort in organic content. When you have enterprise pricing ($99-999/mo), increase ad spend significantly.
Spend $0-300/month on ads. Invest the real effort in organic content. When you have enterprise pricing ($99-999/mo), increase ad spend significantly.
▼
9. LinkedIn Newsletters
3 itemsWhy LinkedIn Newsletters Are a Priority Channel
Medium
Push notifications
Every subscriber receives a notification AND an email when you publish. No algorithm gatekeeping.
SEO
LinkedIn Newsletter articles are indexed by Google. They rank well for long-tail keywords.
Authority building
A newsletter positions you as a regular publisher, not just a poster.
Compounding
Subscribers accumulate over time. Unlike posts (which decay), your newsletter list only grows.
Subscriber invites
When you create a newsletter, LinkedIn lets you send a one-time invite to all your followers/connections. This is extremely powerful for bootstrapping subscriber count.
When you create a newsletter, LinkedIn lets you send a one-time invite to all your followers/connections. This is extremely powerful for bootstrapping subscriber count.
Additional Info
LinkedIn Newsletters are one of the most underutilized distribution tools on the platform. Key advantages: Push notifications: Every subscriber receives a notification AND an email when you publish. No algorithm gatekeeping. Subscriber invites: When you create a newsletter, LinkedIn lets you send a one-time invite to all your followers/connections. This is extremely powerful for bootstrapping subscriber count. SEO: LinkedIn Newsletter articles are indexed by Google. They rank well for long-tail keywords. Authority building: A newsletter positions you as a regular publisher, not just a poster. Compounding: Subscribers accumulate over time. Unlike posts (which decay), your newsletter list only grows. > Fact-check note (April 2026): LinkedIn Newsletters were opened to all members in 2022 (previously restricted to Creator Mode, which has since been retired). As of 2026, any member can create a newsletter. LinkedIn sends an email notification to subscribers for each edition -- this is confirmed and is the primary distribution advantage. The one-time follower invite when creating a newsletter is also confirmed (LinkedIn prompts you to invite all connections/followers). This invite typically converts 5-15% of your connections into subscribers.
LinkedIn Newsletters are one of the most underutilized distribution tools on the platform. Key advantages: Push notifications: Every subscriber receives a notification AND an email when you publish. No algorithm gatekeeping. Subscriber invites: When you create a newsletter, LinkedIn lets you send a one-time invite to all your followers/connections. This is extremely powerful for bootstrapping subscriber count. SEO: LinkedIn Newsletter articles are indexed by Google. They rank well for long-tail keywords. Authority building: A newsletter positions you as a regular publisher, not just a poster. Compounding: Subscribers accumulate over time. Unlike posts (which decay), your newsletter list only grows. > Fact-check note (April 2026): LinkedIn Newsletters were opened to all members in 2022 (previously restricted to Creator Mode, which has since been retired). As of 2026, any member can create a newsletter. LinkedIn sends an email notification to subscribers for each edition -- this is confirmed and is the primary distribution advantage. The one-time follower invite when creating a newsletter is also confirmed (LinkedIn prompts you to invite all connections/followers). This invite typically converts 5-15% of your connections into subscribers.
Newsletter Setup
Medium
Frequency
Biweekly to start. Weekly once you have a content rhythm.
Cover image
1920x1080 px. Branded with Enovari visual identity.
Publish day
Tuesday or Wednesday morning (peak engagement days)
Description
"The unfiltered story of building an AI startup alone. Real numbers, real challenges, real lessons. Following the journey of Enovari from zero to [goal]."
"The unfiltered story of building an AI startup alone. Real numbers, real challenges, real lessons. Following the journey of Enovari from zero to [goal]."
Recommendation
Start with "The Memory Layer." It directly serves your product positioning and attracts the right audience. You can always launch a second newsletter later.
Start with "The Memory Layer." It directly serves your product positioning and attracts the right audience. You can always launch a second newsletter later.
Additional Info
Name: "The Memory Layer" Description: "Weekly insights on AI memory, persistent context, and the infrastructure that makes AI useful over time. Written by a solo founder building Enovari (enovari.ai)." Frequency: Biweekly to start. Weekly once you have a content rhythm. Cover image: 1920x1080 px. Branded with Enovari visual identity. Publish day: Tuesday or Wednesday morning (peak engagement days) Name: "Solo Founder Diaries" Description: "The unfiltered story of building an AI startup alone. Real numbers, real challenges, real lessons. Following the journey of Enovari from zero to [goal]." This angle casts a wider net (all founders, not just AI people) but dilutes the AI memory positioning.
Name: "The Memory Layer" Description: "Weekly insights on AI memory, persistent context, and the infrastructure that makes AI useful over time. Written by a solo founder building Enovari (enovari.ai)." Frequency: Biweekly to start. Weekly once you have a content rhythm. Cover image: 1920x1080 px. Branded with Enovari visual identity. Publish day: Tuesday or Wednesday morning (peak engagement days) Name: "Solo Founder Diaries" Description: "The unfiltered story of building an AI startup alone. Real numbers, real challenges, real lessons. Following the journey of Enovari from zero to [goal]." This angle casts a wider net (all founders, not just AI people) but dilutes the AI memory positioning.
Newsletter Content Strategy
Medium
Additional Info
``` # [Catchy Title]
``` # [Catchy Title]
▼
The TL;DR
0 items
▼
The Big Idea
0 items
▼
From the Build
0 items
▼
What I'm Reading
0 items
▼
One Question
1 itemsGrowing Newsletter Subscribers
Medium
On launch
Use LinkedIn's built-in invite to notify all connections/followers
In every post
Add "Subscribe to my newsletter 'The Memory Layer' for weekly AI memory insights" as a CTA
In comments
When someone asks a great question, say "I actually covered this in my newsletter -- [link]"
Cross-promotion
Mention the newsletter in your email signature, on enovari.ai, in other social channels
Quality content
The best growth hack is consistently publishing content worth subscribing to
Additional Info
On launch: Use LinkedIn's built-in invite to notify all connections/followers In every post: Add "Subscribe to my newsletter 'The Memory Layer' for weekly AI memory insights" as a CTA In comments: When someone asks a great question, say "I actually covered this in my newsletter -- [link]" Cross-promotion: Mention the newsletter in your email signature, on enovari.ai, in other social channels Quality content: The best growth hack is consistently publishing content worth subscribing to
On launch: Use LinkedIn's built-in invite to notify all connections/followers In every post: Add "Subscribe to my newsletter 'The Memory Layer' for weekly AI memory insights" as a CTA In comments: When someone asks a great question, say "I actually covered this in my newsletter -- [link]" Cross-promotion: Mention the newsletter in your email signature, on enovari.ai, in other social channels Quality content: The best growth hack is consistently publishing content worth subscribing to
▼
10. LinkedIn Live & Audio Events
3 itemsLinkedIn Live
Medium
Additional Info
LinkedIn Live is a live video broadcasting feature that sends notifications to your followers and gets algorithmic priority in the feed. Live video gets 7x more reactions and 24x more comments than native video (LinkedIn's own published data) It positions you as a thought leader willing to go unscripted Live viewers can become immediate connections and followers Replay stays on your profile as regular video content > Fact-check note (April 2026): The "7x more reactions and 24x more comments" stat was published by LinkedIn in 2020-2021. The exact multiplier has likely decreased as more people go live (more competition), but LinkedIn Live still receives significant algorithmic boost and notification priority compared to pre-recorded video. LinkedIn Live was opened to all members with 150+ followers in 2023 (no longer requires application). Format 1: Monthly "AI Memory AMA" 30-minute live session Open Q&A about AI memory, persistent context, building with MCP Promotes Enovari organically through the conversation Record and repurpose clips as short posts Format 2: "Building in Public" live coding/demo Show yourself building a feature live Walk through the Enovari architecture Extremely authentic and differentiated (few founders do this) Format 3: Expert interview Invite an AI engineer, CTO, or researcher for a 20-minute conversation Their audience sees the notification, expanding your reach Builds relationship with the guest (potential customer/advocate) LinkedIn Live uses third-party streaming tools (StreamYard, Restream, OBS) LinkedIn does not have a native "go live" button -- you need a streaming tool that connects via RTMP StreamYard (free tier available) is the easiest option Good microphone and lighting are essential (laptop webcam is fine, but audio quality matters most) Promote the live event 3-5 days in advance with a LinkedIn post Create a LinkedIn Event for the live session (people can RSVP and get reminders) Go live for 20-45 minutes (too short feels rushed, too long loses viewers) Have a co-host or moderator for Q&A management Pin a comment with a link to enovari.ai during the stream Repurpose the recording: clip 3-5 highlights as standalone posts
LinkedIn Live is a live video broadcasting feature that sends notifications to your followers and gets algorithmic priority in the feed. Live video gets 7x more reactions and 24x more comments than native video (LinkedIn's own published data) It positions you as a thought leader willing to go unscripted Live viewers can become immediate connections and followers Replay stays on your profile as regular video content > Fact-check note (April 2026): The "7x more reactions and 24x more comments" stat was published by LinkedIn in 2020-2021. The exact multiplier has likely decreased as more people go live (more competition), but LinkedIn Live still receives significant algorithmic boost and notification priority compared to pre-recorded video. LinkedIn Live was opened to all members with 150+ followers in 2023 (no longer requires application). Format 1: Monthly "AI Memory AMA" 30-minute live session Open Q&A about AI memory, persistent context, building with MCP Promotes Enovari organically through the conversation Record and repurpose clips as short posts Format 2: "Building in Public" live coding/demo Show yourself building a feature live Walk through the Enovari architecture Extremely authentic and differentiated (few founders do this) Format 3: Expert interview Invite an AI engineer, CTO, or researcher for a 20-minute conversation Their audience sees the notification, expanding your reach Builds relationship with the guest (potential customer/advocate) LinkedIn Live uses third-party streaming tools (StreamYard, Restream, OBS) LinkedIn does not have a native "go live" button -- you need a streaming tool that connects via RTMP StreamYard (free tier available) is the easiest option Good microphone and lighting are essential (laptop webcam is fine, but audio quality matters most) Promote the live event 3-5 days in advance with a LinkedIn post Create a LinkedIn Event for the live session (people can RSVP and get reminders) Go live for 20-45 minutes (too short feels rushed, too long loses viewers) Have a co-host or moderator for Q&A management Pin a comment with a link to enovari.ai during the stream Repurpose the recording: clip 3-5 highlights as standalone posts
LinkedIn Audio Events
Medium
Additional Info
LinkedIn Audio Events (similar to Twitter Spaces/Clubhouse) are live audio rooms. They're lower-effort than video and feel more conversational. Lower barrier to entry than video (no camera setup needed) Listeners can join from mobile while multitasking Intimate, conversational format builds deeper connections LinkedIn sends notifications to followers when you start an audio event 1. "AI Memory Office Hours" -- Weekly/biweekly 30-minute session where anyone can ask questions about AI memory, MCP, or building AI infrastructure. 2. "Founder Fireside" -- Monthly conversation with another AI/tech founder. Casual, unscripted, real talk. 3. "AI News Roundup" -- React to the week's biggest AI news live with your audience. Real-time commentary. Audio events work best with 2+ speakers (solo audio can feel like a podcast without the polish) Invite 2-3 people to co-host before going live (ensures you have initial conversation partners) Keep it under 45 minutes Promote 24-48 hours in advance
LinkedIn Audio Events (similar to Twitter Spaces/Clubhouse) are live audio rooms. They're lower-effort than video and feel more conversational. Lower barrier to entry than video (no camera setup needed) Listeners can join from mobile while multitasking Intimate, conversational format builds deeper connections LinkedIn sends notifications to followers when you start an audio event 1. "AI Memory Office Hours" -- Weekly/biweekly 30-minute session where anyone can ask questions about AI memory, MCP, or building AI infrastructure. 2. "Founder Fireside" -- Monthly conversation with another AI/tech founder. Casual, unscripted, real talk. 3. "AI News Roundup" -- React to the week's biggest AI news live with your audience. Real-time commentary. Audio events work best with 2+ speakers (solo audio can feel like a podcast without the polish) Invite 2-3 people to co-host before going live (ensures you have initial conversation partners) Keep it under 45 minutes Promote 24-48 hours in advance
LinkedIn Events (Non-Live)
Medium
Additional Info
You can create LinkedIn Events for any occasion -- webinars, product launches, meetups, office hours. Events get their own page and attendees receive notifications. Enovari product launch announcements Webinar: "How to Give Your AI Agent Persistent Memory" Meetup (virtual or local): "AI Infrastructure Builders" Monthly "Ask Me Anything" sessions
You can create LinkedIn Events for any occasion -- webinars, product launches, meetups, office hours. Events get their own page and attendees receive notifications. Enovari product launch announcements Webinar: "How to Give Your AI Agent Persistent Memory" Meetup (virtual or local): "AI Infrastructure Builders" Monthly "Ask Me Anything" sessions
▼
11. Collaborative Articles & Creator Mode
2 itemsCollaborative Articles
Medium
Additional Info
LinkedIn Collaborative Articles are AI-generated article starters on professional topics where experts are invited to contribute their perspectives. Contributing to these articles earns you a "Top Voice" badge on your profile. The "Top Voice" or "Community Top Voice" badge appears on your profile and in search results, signaling expertise Your contribution appears under your name with a link to your profile It's high-authority, low-effort content (you're adding a perspective, not writing a full article) LinkedIn actively promotes Collaborative Articles in search results and the feed > Fact-check note (April 2026): LinkedIn Collaborative Articles launched in early 2023 and have been actively expanded. Contributing quality perspectives can earn a "Community Top Voice" badge in a specific skill area (e.g., "Top Voice in Artificial Intelligence"). This badge appears on your profile for as long as you remain active in contributions. LinkedIn has tightened criteria -- you need to contribute consistently (multiple articles per month) and receive positive reactions on your contributions to earn/maintain the badge. The badge has real credibility value: it appears in connection requests and comments, increasing trust. 1. Go to linkedin.com/pulse/discover 2. Search for topics: "Artificial Intelligence," "Machine Learning," "AI Infrastructure," "Knowledge Management" 3. Find articles where you can add expert commentary 4. Write 3-5 sentence contributions that demonstrate genuine expertise 5. Contribute to 5-10 articles per week for 2-4 weeks 6. LinkedIn will evaluate your contributions and award the badge if quality and consistency are high enough Reference your experience building Enovari Share specific data points or technical insights Disagree respectfully with the article's premise when warranted (this stands out) Avoid generic advice -- be specific and opinionated Artificial Intelligence Machine Learning Natural Language Processing Software Architecture AI Infrastructure (if available) Knowledge Management
LinkedIn Collaborative Articles are AI-generated article starters on professional topics where experts are invited to contribute their perspectives. Contributing to these articles earns you a "Top Voice" badge on your profile. The "Top Voice" or "Community Top Voice" badge appears on your profile and in search results, signaling expertise Your contribution appears under your name with a link to your profile It's high-authority, low-effort content (you're adding a perspective, not writing a full article) LinkedIn actively promotes Collaborative Articles in search results and the feed > Fact-check note (April 2026): LinkedIn Collaborative Articles launched in early 2023 and have been actively expanded. Contributing quality perspectives can earn a "Community Top Voice" badge in a specific skill area (e.g., "Top Voice in Artificial Intelligence"). This badge appears on your profile for as long as you remain active in contributions. LinkedIn has tightened criteria -- you need to contribute consistently (multiple articles per month) and receive positive reactions on your contributions to earn/maintain the badge. The badge has real credibility value: it appears in connection requests and comments, increasing trust. 1. Go to linkedin.com/pulse/discover 2. Search for topics: "Artificial Intelligence," "Machine Learning," "AI Infrastructure," "Knowledge Management" 3. Find articles where you can add expert commentary 4. Write 3-5 sentence contributions that demonstrate genuine expertise 5. Contribute to 5-10 articles per week for 2-4 weeks 6. LinkedIn will evaluate your contributions and award the badge if quality and consistency are high enough Reference your experience building Enovari Share specific data points or technical insights Disagree respectfully with the article's premise when warranted (this stands out) Avoid generic advice -- be specific and opinionated Artificial Intelligence Machine Learning Natural Language Processing Software Architecture AI Infrastructure (if available) Knowledge Management
Creator Mode (Legacy Notes)
Medium
Additional Info
> Important update (April 2026): LinkedIn officially retired Creator Mode in early 2024. All Creator Mode features (Follow button as default, access to LinkedIn Live, access to Newsletters, profile topics) were either rolled out to all users or deprecated. You do NOT need to enable Creator Mode -- it no longer exists as a toggle. > > What was part of Creator Mode and is now available to everyone: > - Follow as default profile button (configurable in Settings) > - LinkedIn Live access (available to all users with 150+ followers) > - Newsletter creation (available to all users) > - LinkedIn Audio Events (available to all users) > > What was removed: > - Profile topic hashtags (the 5 hashtags displayed under your name) > - Featured "Talks about" section on profile > > Action item: If any other documents or checklists in the Enovari marketing folder reference "turn on Creator Mode," update them to reflect that this setting no longer exists.
> Important update (April 2026): LinkedIn officially retired Creator Mode in early 2024. All Creator Mode features (Follow button as default, access to LinkedIn Live, access to Newsletters, profile topics) were either rolled out to all users or deprecated. You do NOT need to enable Creator Mode -- it no longer exists as a toggle. > > What was part of Creator Mode and is now available to everyone: > - Follow as default profile button (configurable in Settings) > - LinkedIn Live access (available to all users with 150+ followers) > - Newsletter creation (available to all users) > - LinkedIn Audio Events (available to all users) > > What was removed: > - Profile topic hashtags (the 5 hashtags displayed under your name) > - Featured "Talks about" section on profile > > Action item: If any other documents or checklists in the Enovari marketing folder reference "turn on Creator Mode," update them to reflect that this setting no longer exists.
▼
12. LinkedIn Etiquette & Avoiding Restrictions
6 itemsLinkedIn Account Restrictions: What Gets You Flagged
Medium
Additional Info
LinkedIn actively monitors for spam, automation, and inauthentic behavior. If your account gets restricted, you may lose the ability to send connection requests, messages, or post content for days or weeks. Severe violations can result in permanent bans.
LinkedIn actively monitors for spam, automation, and inauthentic behavior. If your account gets restricted, you may lose the ability to send connection requests, messages, or post content for days or weeks. Severe violations can result in permanent bans.
How to Avoid Looking Like a Bot
Medium
Additional Info
LinkedIn's anti-spam systems are designed to detect non-human behavior. Even if you're doing everything manually, certain patterns can flag you. Vary your activity timing. Don't send 20 connection requests at exactly 9:00 AM every day. Spread activity throughout the day with natural gaps. Personalize every message. Even small variations (mentioning their name, company, or a recent post) signal human behavior. Identical messages are the #1 bot signal. Don't max out daily limits. If LinkedIn allows ~100 connection requests/week, aim for 50-70. Leave headroom. Engage before requesting. Like or comment on someone's post before sending a connection request. This creates a natural interaction trail. Mix your activities. Don't just send connection requests. Alternate between posting, commenting, viewing profiles, reading articles, and messaging. A real person does all of these; a bot does one thing in bulk. Use the mobile app. Some activity from the mobile app (which LinkedIn can verify) reduces bot suspicion. Warm up new accounts slowly. If your account is new or has been dormant, start with 5-10 connection requests/day and increase gradually over 2-3 weeks. Respond to messages promptly. A pattern of sending messages but never responding to replies looks automated.
LinkedIn's anti-spam systems are designed to detect non-human behavior. Even if you're doing everything manually, certain patterns can flag you. Vary your activity timing. Don't send 20 connection requests at exactly 9:00 AM every day. Spread activity throughout the day with natural gaps. Personalize every message. Even small variations (mentioning their name, company, or a recent post) signal human behavior. Identical messages are the #1 bot signal. Don't max out daily limits. If LinkedIn allows ~100 connection requests/week, aim for 50-70. Leave headroom. Engage before requesting. Like or comment on someone's post before sending a connection request. This creates a natural interaction trail. Mix your activities. Don't just send connection requests. Alternate between posting, commenting, viewing profiles, reading articles, and messaging. A real person does all of these; a bot does one thing in bulk. Use the mobile app. Some activity from the mobile app (which LinkedIn can verify) reduces bot suspicion. Warm up new accounts slowly. If your account is new or has been dormant, start with 5-10 connection requests/day and increase gradually over 2-3 weeks. Respond to messages promptly. A pattern of sending messages but never responding to replies looks automated.
What Gets You Banned (Hard Rules)
Medium
Additional Info
These will almost certainly result in account restriction or ban: 1. Using third-party automation tools that violate LinkedIn's Terms of Service (Dux-Soup, PhantomBuster, LinkedHelper, Octopus CRM, etc.). LinkedIn actively detects these through browser fingerprinting and API monitoring. 2. Scraping profile data for use outside LinkedIn. 3. Creating fake profiles or impersonating others. 4. Sending unsolicited sexual/harassing messages. 5. Bulk connection requests to people with no relevance (e.g., sending 500 requests to random people). 6. Selling connection requests or engagement (paying for likes/comments from engagement pods). 7. Repeatedly posting content that gets flagged/removed by LinkedIn.
These will almost certainly result in account restriction or ban: 1. Using third-party automation tools that violate LinkedIn's Terms of Service (Dux-Soup, PhantomBuster, LinkedHelper, Octopus CRM, etc.). LinkedIn actively detects these through browser fingerprinting and API monitoring. 2. Scraping profile data for use outside LinkedIn. 3. Creating fake profiles or impersonating others. 4. Sending unsolicited sexual/harassing messages. 5. Bulk connection requests to people with no relevance (e.g., sending 500 requests to random people). 6. Selling connection requests or engagement (paying for likes/comments from engagement pods). 7. Repeatedly posting content that gets flagged/removed by LinkedIn.
LinkedIn Etiquette (Unwritten Rules)
Medium
Additional Info
These aren't official rules, but violating them damages your reputation and effectiveness: Always include a note explaining why you're connecting (even though it's optional) Never pitch in the connection request message Don't send connection requests to people you have zero relevance to If someone declines or ignores your request, do NOT send again Don't connect and immediately unfollow (people notice) Never send a sales pitch as your first message after connecting Don't send the same message template to many people (they compare notes) Don't add people to email lists without permission just because you connected Keep messages concise (under 200 words for cold outreach) Don't follow up more than once if someone doesn't reply. Silence IS a response. Never send voice messages or video messages as cold outreach (feels invasive) Don't tag people who didn't ask to be tagged (especially more than 3-5 per post) Don't comment "Great post!" with no substance just to be seen Don't copy/paste other people's posts as your own (plagiarism is rampant on LinkedIn and people notice) Don't use pods/engagement groups where you trade likes/comments (LinkedIn detects these patterns and reduces reach for all participants) Don't post more than once per day (the second post cannibalizes the first) Don't immediately delete and repost content because it's not performing well Don't use clickbait without delivering on the promise Don't post exclusively self-promotional content (the rule of thumb is 80% value, 20% promotion) Read the group rules before posting Lurk for a week before contributing Never DM group members with sales pitches Don't post the same content across multiple groups on the same day
These aren't official rules, but violating them damages your reputation and effectiveness: Always include a note explaining why you're connecting (even though it's optional) Never pitch in the connection request message Don't send connection requests to people you have zero relevance to If someone declines or ignores your request, do NOT send again Don't connect and immediately unfollow (people notice) Never send a sales pitch as your first message after connecting Don't send the same message template to many people (they compare notes) Don't add people to email lists without permission just because you connected Keep messages concise (under 200 words for cold outreach) Don't follow up more than once if someone doesn't reply. Silence IS a response. Never send voice messages or video messages as cold outreach (feels invasive) Don't tag people who didn't ask to be tagged (especially more than 3-5 per post) Don't comment "Great post!" with no substance just to be seen Don't copy/paste other people's posts as your own (plagiarism is rampant on LinkedIn and people notice) Don't use pods/engagement groups where you trade likes/comments (LinkedIn detects these patterns and reduces reach for all participants) Don't post more than once per day (the second post cannibalizes the first) Don't immediately delete and repost content because it's not performing well Don't use clickbait without delivering on the promise Don't post exclusively self-promotional content (the rule of thumb is 80% value, 20% promotion) Read the group rules before posting Lurk for a week before contributing Never DM group members with sales pitches Don't post the same content across multiple groups on the same day
Content That Performs Poorly on LinkedIn
Medium
Additional Info
Understanding what NOT to post is as important as knowing what to post. 1. Pure product announcements without a story or insight. "We just launched feature X!" gets minimal engagement unless framed as a lesson or insight. 2. Links to external content without native LinkedIn text. A bare URL with "check this out" gets buried by the algorithm. 3. Overly formal/corporate tone. LinkedIn rewards personality. Writing like a press release kills engagement. 4. Long posts with no white space. Dense paragraphs without line breaks get scrolled past. 5. Humble-bragging. "So grateful to be featured in [prestigious outlet]" without adding real insight. People see through it. 6. Motivational quotes without personal context. "Success is a journey, not a destination" adds no value. 7. Reshares without commentary. Clicking the reshare button without adding your own take gets near-zero reach. 8. Multi-image posts where images don't tell a clear story. Random screenshots without context confuse viewers. 9. Content with no hook. Starting with a boring first line means nobody clicks "See more." 10. Content that talks AT the audience instead of engaging them. Posts with no question, no invitation to comment, no conversational element. 11. Negative/complaining content. "Why does everything in tech suck?" Negativity without constructive insight gets low engagement and hurts your brand. 12. Content that's too niche for your audience. A post about Enovari's database indexing strategy might fascinate 10 engineers and lose everyone else. Balance depth with accessibility. Misleading claims (especially about income/results) Controversial political/religious content Content that attacks specific people or companies by name Sexually suggestive content Content that promotes illegal activity Fake testimonials or fabricated data
Understanding what NOT to post is as important as knowing what to post. 1. Pure product announcements without a story or insight. "We just launched feature X!" gets minimal engagement unless framed as a lesson or insight. 2. Links to external content without native LinkedIn text. A bare URL with "check this out" gets buried by the algorithm. 3. Overly formal/corporate tone. LinkedIn rewards personality. Writing like a press release kills engagement. 4. Long posts with no white space. Dense paragraphs without line breaks get scrolled past. 5. Humble-bragging. "So grateful to be featured in [prestigious outlet]" without adding real insight. People see through it. 6. Motivational quotes without personal context. "Success is a journey, not a destination" adds no value. 7. Reshares without commentary. Clicking the reshare button without adding your own take gets near-zero reach. 8. Multi-image posts where images don't tell a clear story. Random screenshots without context confuse viewers. 9. Content with no hook. Starting with a boring first line means nobody clicks "See more." 10. Content that talks AT the audience instead of engaging them. Posts with no question, no invitation to comment, no conversational element. 11. Negative/complaining content. "Why does everything in tech suck?" Negativity without constructive insight gets low engagement and hurts your brand. 12. Content that's too niche for your audience. A post about Enovari's database indexing strategy might fascinate 10 engineers and lose everyone else. Balance depth with accessibility. Misleading claims (especially about income/results) Controversial political/religious content Content that attacks specific people or companies by name Sexually suggestive content Content that promotes illegal activity Fake testimonials or fabricated data
Recovery: What to Do If You Get Restricted
Medium
Additional Info
If LinkedIn restricts your account: 1. Stop all outreach activity immediately. Don't try to "push through" a restriction. 2. Review the restriction message. LinkedIn usually tells you what triggered it and how long the restriction lasts. 3. Wait the full restriction period. Restrictions typically last 1-7 days for first offenses. 4. When restrictions lift, reduce activity. Come back at 50% of your previous activity level. 5. If permanently restricted, appeal. Use linkedin.com/help/linkedin/ask/TS-AARL to submit an appeal. Be honest about what happened. 6. Keep posting. Account restrictions usually only affect messaging/connection features, not posting. Continue your content cadence.
If LinkedIn restricts your account: 1. Stop all outreach activity immediately. Don't try to "push through" a restriction. 2. Review the restriction message. LinkedIn usually tells you what triggered it and how long the restriction lasts. 3. Wait the full restriction period. Restrictions typically last 1-7 days for first offenses. 4. When restrictions lift, reduce activity. Come back at 50% of your previous activity level. 5. If permanently restricted, appeal. Use linkedin.com/help/linkedin/ask/TS-AARL to submit an appeal. Be honest about what happened. 6. Keep posting. Account restrictions usually only affect messaging/connection features, not posting. Continue your content cadence.
3. LinkedIn Algorithm & Content Mechanics
3.1 How the LinkedIn Algorithm Works (2025-2026)
LinkedIn's feed algorithm evaluates posts in stages:
Stage 1: Quality Filter (0-60 minutes)
- Is this spam? Low quality? Self-promotional link? If yes, suppressed.
- Posts with external links get ~40-50% less reach than native content.
- Text-only posts and carousels pass this filter most easily.
- LinkedIn shows the post to a small slice of your network (estimated 5-15% of followers, varying by account age and history).
- Measures: dwell time (how long people look at it), reactions, comments, shares, clicks on "See more."
- Dwell time is the most important early signal. Longer posts that make people stop scrolling win.
- If the test audience engaged well, LinkedIn expands to more followers + their networks.
- Comments are weighted heavily for expansion (estimated 8-15x more impactful than likes, though LinkedIn does not publish exact multipliers).
- Shares to feed (not DMs) trigger further expansion.
- Top posts continue getting impressions for 5-14 days.
- Most posts decay after 48-72 hours.
- Evergreen posts can resurface if someone comments days later.
- Text-only posts with a strong hook (highest average organic reach)
- Carousel / PDF documents (high dwell time + saves)
- Single image + text (if the image adds value)
- Video (native upload, not YouTube links)
- Polls (high engagement but low quality audience)
- Articles (low reach but good for SEO and authority)
- Posts with external links (lowest reach -- LinkedIn penalizes outbound links)
- High dwell time (long-form posts, carousels)
- Comments (especially within the first hour)
- Saves (bookmarks)
- Shares (reposts with commentary)
- Profile views generated by the post
- Engagement from outside your network (2nd and 3rd degree connections)
- "See more" click rate (people expanding the post to read the full text)
- Time spent reading comments (indicates quality discussion)
- External links in the post body (put them in the first comment instead)
- Editing a post within the first hour
- Tagging more than 5 people
- Using more than 5 hashtags
- Posting more than once per day (cannibalization)
- Engagement bait ("Like if you agree, comment if you don't")
- Posting and disappearing -- not responding to comments
- Posting identical or near-identical content to what you've posted before
- Using LinkedIn's "Celebrate an occasion" templates (extremely low reach)
- Resharing/reposting without adding commentary (LinkedIn deprioritizes empty reshares)
- Post at the same time every day -- LinkedIn rewards consistency
- If targeting West Coast tech, 7 AM PT (10 AM ET) works well
- International audiences: consider UTC scheduling
- Your analytics (once you have 30+ posts) will tell you YOUR best times
- Tuesday and Wednesday are statistically the two strongest days across nearly all studies. Prioritize your best content for those days.
- Avoid posting between 10 PM and 6 AM ET -- even if you're a night owl, schedule for morning
- Contrarian statement: "AI doesn't need more parameters. It needs memory."
- Surprising number: "Our AI remembered 10,000 facts across 47 sessions with 94% accuracy."
- Story opener: "Last Tuesday at 2 AM, I almost shut down my company."
- Question: "Why does your AI forget everything the moment you close the tab?"
- Bold claim: "The next billion-dollar AI infrastructure company won't build models. It'll build memory."
- Pattern interrupt: "I'm a solo founder. No co-founder. No team. Here's what that actually looks like."
- Relatable pain: "I explained my project to ChatGPT for the 47th time yesterday. Same project. Same context. From scratch. Again."
- Micro-story with tension: "A CTO at a Fortune 500 company told me my product was impossible. Three weeks later, he asked for enterprise pricing."
- Counter-intuitive insight: "The most important AI feature isn't intelligence. It's memory. Here's why."
- Vulnerable admission: "I've been building this for a year and I still don't know if it will work. But here's what I DO know."
- Use short paragraphs (1-2 sentences max)
- Use line breaks liberally (white space = dwell time)
- Use dashes, arrows, or bullets for lists
- Include a personal angle or specific detail
- Optimal length: 800-1,300 characters for text posts (long enough for dwell time, short enough to finish)
- For carousel-accompanying text: 200-400 characters (the carousel itself provides the dwell time)
- End with a question to drive comments
- Or a clear takeaway / lesson
- Or a call to action (follow for more, check link in comments)
- Do NOT end with a link -- put links in the first comment
- Effective CTA formulas for Enovari: - "What's the most frustrating thing your AI keeps forgetting?" (invites stories) - "Am I wrong? Tell me in the comments." (invites debate) - "If you're building AI agents, DM me -- I have something that might help." (invites DMs without being salesy) - "Follow me for more on AI memory and solo founder life." (simple and direct)
- Each swipe counts as dwell time
- High save rate (people bookmark for later)
- They look professional and authoritative
- They're shareable as standalone assets
- 8-12 slides is the sweet spot
- First slide = eye-catching title (this is the "hook")
- Last slide = CTA (follow me, visit enovari.ai, share this)
- One idea per slide
- Large text (readable on mobile)
- Consistent branding (Enovari colors/logo on each slide)
- File format: PDF upload (LinkedIn converts to carousel)
- "5 Reasons Your AI Forgets Everything (And How to Fix It)"
- "The Architecture Behind AI Memory Systems"
- "Solo Founder Toolkit: Tools I Use to Build Enovari"
- "AI Memory vs. AI Chat History: What's the Difference?"
- "How MCP Protocol Is Changing AI Infrastructure"
- "RAG vs. Persistent Memory: A Visual Comparison" (side-by-side diagrams)
- "The 7 Types of Knowledge an AI Should Remember" (with examples for each)
- "How to Give Your AI Agent Long-Term Memory in 5 Steps" (tutorial format)
- #ArtificialIntelligence (largest AI hashtag)
- #AIMemory (niche, own this one)
- #BuildInPublic
- #SoloFounder
- #AIInfrastructure
- #MachineLearning
- #DevTools
- #SaaS
- #StartupLife
- #TechFounder
- #LLM
- #GenerativeAI
- #EnterpriseAI
- #FutureOfWork
- #AIAgents
- #KnowledgeGraph
- #MCP
- [ ] Create LinkedIn company page with all optimized content
- [ ] Upload logo and banner images
- [ ] Optimize personal profile (headline, about, featured, banner)
- [ ] Verify profile default action is set to "Follow" (Settings > Visibility)
- [ ] Connect with 100 people in AI/tech
- [ ] Join 10-15 LinkedIn groups
- [ ] Write and publish first 3 posts
- [ ] Install LinkedIn Insight Tag on enovari.ai
- [ ] Add LinkedIn follow button to enovari.ai
- [ ] Publish 5 posts (1 per weekday)
- [ ] Comment on 20+ posts daily from AI/tech leaders
- [ ] Send 25 personalized connection requests
- [ ] Engage in 3+ groups
- [ ] Create first carousel post
- [ ] Invite all personal connections to follow company page
- [ ] Maintain 5 posts/week cadence
- [ ] Continue daily commenting (20+/day)
- [ ] Send 25 connection requests/week
- [ ] Publish first LinkedIn article
- [ ] Track: impressions, followers, engagement rate
- [ ] Analyze top-performing post, create more like it
- [ ] Engage in groups 3x/week
- [ ] Launch LinkedIn newsletter (see Section 9)
- [ ] Create 2 carousel posts per week
- [ ] Start DM outreach sequence with warm connections
- [ ] Post first product demo or video
- [ ] Collaborate with 1-2 other AI founders on content
- [ ] Aim for 500+ followers on personal profile
- [ ] First company page newsletter edition
- [ ] Track: which content types drive website traffic
- [ ] First LinkedIn Live or Audio Event (see Section 10)
- [ ] Begin boosting top posts ($5-10/day)
- [ ] Aim for 1,000+ followers on personal profile
- [ ] Aim for 250+ followers on company page
- [ ] Create a signature content series
- [ ] Start tracking LinkedIn-attributed sign-ups
- [ ] Refine outreach based on what messages get responses
- [ ] Publish 2 long-form articles
- [ ] Goal: 1-2 posts breaking 10K impressions
- [ ] Contribute to 3-5 Collaborative Articles (see Section 11)
- [ ] Evaluate LinkedIn Live as recurring format
- [ ] Professional headshot (recent, high-quality)
- [ ] Custom banner with Enovari branding (1584x396 px)
- [ ] Headline includes value proposition (not just "Founder")
- [ ] About section tells a story (not a resume)
- [ ] Featured section has 3-6 pinned items
- [ ] Experience includes Enovari with description
- [ ] Profile default action set to "Follow" (Settings > Visibility)
- [ ] Custom URL (linkedin.com/in/yourname)
- [ ] Contact info includes enovari.ai
- [ ] Skills section includes AI, Machine Learning, etc.
- [ ] 500+ connections
- [ ] Profile is set to "Open to" appropriate options
- [ ] Newsletter created and first edition published
- [ ] Contributing to Collaborative Articles (targeting "Top Voice" badge)
- [ ] LinkedIn Insight Tag installed on enovari.ai
- LinkedIn native analytics (for both personal and company pages)
- Shield App (free tier) -- personal profile analytics
- LinkedIn Sales Navigator (paid, but essential for outreach)
- Canva (free tier) -- carousel and banner design
- Buffer or Hootsuite (free tier) -- scheduling posts
- Total impressions (are you growing reach?)
- Engagement rate (impressions vs. reactions+comments)
- Follower growth rate
- Profile views (are people clicking through?)
- Website clicks from LinkedIn (tracked via UTM parameters)
- Connection request acceptance rate
- DM response rate
- Top-performing post and why
- Content format comparison (text vs. carousel vs. video)
- Newsletter subscriber growth and open rates
- LinkedIn-attributed sign-ups on enovari.ai
- Conversion funnel: impression > profile view > website click > sign-up
- Outreach efficiency: messages sent > replies > conversations > conversions
- Anthropic, OpenAI (how major AI companies communicate)
- LangChain, LlamaIndex (AI infrastructure positioning)
- Pinecone, Weaviate (vector database / AI infra marketing)
- Notion, Slack (developer/productivity tool marketing)
- Vercel, Supabase (developer-focused company page strategies)
- Search for people with "AI" in their headline who post 3+ times per week
- Engage with their content consistently to build relationship before reaching out
- Look for creators with 10K-50K followers (large enough to have reach, small enough to notice your comments)
Stage 2: Test Audience (1-4 hours)
Stage 3: Expansion (4-48 hours)
Stage 4: Viral or Decay (48+ hours)
Fact-check note (April 2026): The stage-based model described here is a widely used simplification and is directionally accurate. LinkedIn does not publicly document exact algorithm mechanics. Key corrections and updates:
> - Test audience percentage: The commonly cited "5-8%" figure is an estimate from third-party analyses (e.g., Richard van der Blom's annual algorithm research). It likely varies significantly based on account history, follower count, and content type. LinkedIn's own engineering blog describes the process as "collaborative filtering" rather than a fixed percentage test. A more accurate statement is "a small initial audience, typically your most engaged connections."
- Comments vs. likes weighting: The "10x" figure is an approximation from third-party experiments (notably Just Connecting's 2024 algorithm study). LinkedIn has not confirmed a specific multiplier. The principle is correct: comments signal deeper engagement and are weighted more heavily than reactions. The ratio may be closer to 8-15x depending on comment quality and length.
- External link penalty: The 40-50% reach reduction for posts with external links is consistent with multiple third-party studies (Hootsuite, Buffer, and independent creators all report 30-50% lower reach). This remains accurate in 2026. LinkedIn's stated reason is that they want to keep users on-platform.
- 2025-2026 algorithm shift: LinkedIn announced in mid-2024 a shift toward "knowledge and advice" content over "engagement bait." Posts that demonstrate genuine expertise, provide actionable advice, or share original professional knowledge are now actively boosted. Viral-optimized content (hot takes with no substance, controversy for controversy's sake) is being deprioritized. This is a meaningful change from 2023 when engagement bait still performed well.
- New signal -- "meaningful comments": LinkedIn now evaluates comment quality. Short comments ("Great!") and emoji-only comments carry much less weight than substantive replies (5+ words that add to the conversation).
3.2 What Gets Visibility
Highest reach formats (ranked):
Update (April 2026): LinkedIn has been actively promoting short-form native video (under 90 seconds) since late 2025, similar to Reels/TikTok format. A dedicated video feed tab was added to the mobile app in 2025. Short native video may now rank higher than single image + text for some accounts. Polls have been deprioritized -- LinkedIn reduced poll distribution in late 2024 after an oversaturation of low-effort polls. Articles now get slightly better distribution if published as a Newsletter (because newsletters trigger push notifications to subscribers).
Algorithm signals that boost reach:
Algorithm killers:
Fact-check note: The "editing a post within the first hour kills reach" claim is widely repeated but has mixed evidence. Some creators report no impact, others report significant drops. The safest practice is to proofread carefully before publishing. If you must edit, small typo fixes (not content changes) appear to have minimal impact. The "more than 5 hashtags" threshold is also debated -- LinkedIn's own guidance says 3-5 is ideal, but the penalty for 6-8 is unclear. Stick with 3-5 to be safe.
3.3 Best Posting Times
Optimal posting windows (US time zones):
| Day | Best Time | Why | |||
| Tuesday | 7:00-8:30 AM ET | Highest weekday engagement | |||
| Wednesday | 7:00-8:30 AM ET | Strong mid-week activity | |||
| Thursday | 7:00-8:30 AM ET | Decision-makers are active | |||
| Monday | 8:00-9:00 AM ET | Slightly lower but still good | |||
| Friday | 7:00-8:00 AM ET | Acceptable, drops off after noon | |||
| Saturday/Sunday | Avoid for B2B | Unless targeting international audiences | |||
| Metric | Month 1 Target | Month 3 Target | Month 6 Target | ||
| Personal followers | 300 | 1,000 | 3,000 | ||
| Company followers | 100 | 300 | 1,000 | ||
| Posts published | 20 | 60 (cumulative) | 120 (cumulative) | ||
| Avg impressions/post | 500 | 2,000 | 5,000 | ||
| Engagement rate | 3% | 5% | 5%+ | ||
| Website clicks from LinkedIn | 50 | 200 | 500 | ||
| Sign-ups from LinkedIn | 5 | 20 | 50 | ||
| Connection requests sent | 100 | 300 | 500 | ||
| Meaningful conversations started | 10 | 30 | 60 | ||
| Newsletter subscribers | -- | 200 | 1,000 | ||
| LinkedIn Live viewers (if running) | -- | 20 | 50+ | ||
| Week | Mon | Tue | Wed | Thu | Fri |
| 1 | Insight post | Technical deep-dive | Founder story | Hot take/opinion | Engagement question |
| 2 | Industry trend | Carousel (tutorial) | Behind the scenes | News commentary | Poll/question |
| 3 | Bold prediction | Technical explainer | Lesson learned | Contrarian view | Community shoutout |
| 4 | Metrics update | Product update | Personal reflection | Guest collaboration | Week in review |
| Name | Relevance | What to Learn | |||
| Harrison Chase (LangChain) | AI infrastructure founder | How he positions technical products on LinkedIn | |||
| Dedy Kredo (Mem0) | Direct competitor space | What messaging resonates for AI memory products | |||
| Jerry Liu (LlamaIndex) | AI data framework founder | Technical thought leadership style | |||
| Amjad Masad (Replit) | Developer tools founder | Building in public at scale | |||
| Guillaume Lample (Mistral) | AI lab founder | How AI researchers communicate on LinkedIn | |||
| Any YC/Techstars AI founder | Peer founders | Observe what gets traction in the AI startup space | |||
| Feature | Status | Access | Notes | ||
| Personal profile posts | Active | All users | Primary content channel | ||
| Company page posts | Active | Page admins | Secondary channel | ||
| LinkedIn Newsletter | Active | All users | High-value -- push notifications to subscribers | ||
| LinkedIn Articles | Active | All users | Long-form, good for SEO | ||
| LinkedIn Live | Active | 150+ followers | Requires third-party streaming tool | ||
| LinkedIn Audio Events | Active | All users | Similar to Twitter Spaces | ||
| LinkedIn Events | Active | All users | RSVPs and reminders | ||
| Collaborative Articles | Active | Invited or opt-in | Earn "Top Voice" badge | ||
| Carousel/Document posts | Active | All users | PDF upload or multi-image | ||
| Native video | Active | All users | Up to 10 minutes, short-form prioritized | ||
| Polls | Active (deprioritized) | All users | Reduced algorithmic distribution | ||
| Creator Mode | Retired (2024) | N/A | Features rolled into all profiles | ||
| LinkedIn Stories | Retired (2021) | N/A | Discontinued | ||
| Cover Story (profile video) | Limited | Some users | 30-second profile intro video | ||
| Scheduled posts | Active | All users | Native scheduling in post composer | ||
| Thought Leader Ads | Active | Campaign Manager | Sponsor personal profile posts as ads |