CognitiveHire Content Strategy

# CognitiveHire Content Strategy
> Version 1.0 | March 2026 | Built for launch

---

STAGE 1: AUDIT

Current Assets

What exists:
- Grand Diomande consulting brand (granddiomande.com) — active, Vercel-hosted
- 50+ iOS apps shipped (39 on TestFlight)
- 112K+ AI interactions across 11 domains in Supabase
- Technical blog posts drafted but not published (cognitive twin, CALC, Vantage)
- Strong technical credibility: ML, distributed systems, multi-agent coordination, iOS, creative production
- Guinean-American heritage + N'Ko language AI work — authentic differentiator
- No dedicated CognitiveHire content presence yet

What's missing:
- Platform-specific content presence for CognitiveHire
- A consistent publishing cadence
- Demo content (video, interactive twin interface)
- An audience that knows this product exists

Assessment:
This is a greenfield launch with strong underlying credibility. The asset gap is distribution, not substance. Mohamed has proof. The work is to make that proof visible to the right people.

---

STAGE 2: IDEATE

Six divergent paths evaluated before selection.

### Path A: Educational
Premise: "Here's what your AI usage data actually reveals about you"

Teaches the audience something they didn't know. Positions Mohamed as the person who figured this out first. Low defensiveness from the audience. Works well for cold traffic who don't know CognitiveHire yet.

Core angle: Your AI usage history is a behavioral fingerprint. Here's how to read it.

Sample hook: "I analyzed 112,000 AI interactions. Your prompts tell me more about how you think than your resume ever could."

---

### Path B: Story-Driven
Premise: "I let an AI twin do my job interview. Here's what happened."

High narrative tension. Personal stakes. Immediately differentiates from thought leadership fluff. Works across LinkedIn, Twitter, YouTube. People share stories, not arguments.

Core angle: First-person account of building and deploying a cognitive twin for hiring contexts.

Sample hook: "I sent my AI twin to a technical screening. It answered better than I would have. That's either a problem or an opportunity, depending on who you are."

---

### Path C: Data/Authority
Premise: "I analyzed 112K AI interactions. Here's what I found."

Specificity builds trust. Numbers are shareable. Sets up Mohamed as the person who has done the empirical work everyone else is theorizing about. Best for LinkedIn and long-form blog.

Core angle: What 11 domains of AI usage actually reveal about cognitive patterns, decision-making styles, and problem-solving approaches.

Sample hook: "112,000 AI interactions. 11 domains. 2 years. Here's what your prompting style actually says about you."

---

### Path D: Provocative/Contrarian
Premise: "Your resume is worthless. Your ChatGPT history isn't."

High friction, high share rate. Will generate disagreement, which is good for reach. Best for Twitter/X. Needs careful execution to avoid being click-bait without substance.

Core angle: Traditional hiring selects for performance on a made-up task (the interview). AI interaction history shows actual thinking.

Sample hook: "You've been optimizing the wrong document. Nobody cares about your resume anymore. They should be asking for your AI interaction history."

---

### Path E: Demo-First
Premise: "Talk to my cognitive twin right now. Then decide if you'd hire me."

Highest differentiation. Puts the product in front of the audience before the pitch. Risk: requires the twin interface to be polished enough for public exposure. Best used once product is ready.

Core angle: Skip the explanation. Experience it.

Sample hook: "I built an AI twin trained on 2 years of my actual thinking. You can talk to it right now at [link]. Come back when you're done."

---

### Path F: Industry Analysis
Premise: "The $200B recruiting industry is solving the wrong problem"

Appeals to investors and HR-tech adjacent audience. Positions CognitiveHire inside a market narrative. Good for fundraising context. Lower virality potential than personal story paths.

Core angle: The recruiting industry is optimizing interview performance, not job performance. These are not the same thing.

Sample hook: "Recruiting spends $200B/year finding people who are good at interviewing. That's not the same as finding people who are good at working."

---

STAGE 3: CURATE

Evaluation Matrix

Audience-Path Match

Tech hiring managers frustrated with AI-generated resumes:
Best served by B, D, C in that order. They feel the pain. Story and provocation meet them where the frustration lives. Data validates what they already suspect.

Senior engineers who want to showcase thinking depth:
Best served by A, C, E. They want a tool that works, not a pitch. Education and data show them how it works. Demo closes.

HR-tech/AI investors:
Best served by F, C, B. Market framing first, data to support, story to humanize.

Selected Primary Paths (in order of deployment)

1. D (Provocative) for initial attention and reach, Twitter/X first, then LinkedIn
2. B (Story-driven) for depth and trust-building, LinkedIn anchor content
3. C (Data/Authority) for credibility foundation, blog + LinkedIn long-form
4. A (Educational) for sustained organic growth, ongoing series
5. E (Demo-first) at product readiness, not before

Path F feeds investor conversations directly. Not a content priority for public channels but worth one long-form piece for the website.

---

STAGE 4: ARCHITECTURE

Content Pillars

Pillar 1: The Question Thesis
The idea that questions reveal more than answers. Every post in this pillar argues that the question someone asks is a window into how they actually think. This is the philosophical core of CognitiveHire.

Topics: prompt quality as cognitive signal, question formulation as problem decomposition, what bad prompts cost teams.

Pillar 2: What the Data Shows
Findings from 112K+ interactions. Specific, empirical, unexpected. This pillar builds authority and gives people something to share.

Topics: domain specialization patterns, interaction volume vs. quality ratios, what a 4.3

Pillar 3: The Interview Problem
Why traditional hiring is broken specifically in an AI world. Not generic "recruiting is broken" takes. Specific, current, tied to CognitiveHire's thesis.

Topics: AI-generated resumes removing signal, interview performance vs. work performance gap, what hiring managers actually need to know but can't find out.

Pillar 4: Builder Stories
First-person accounts of building and using CognitiveHire. Honest. Specific. Mohamed is the first user and the most credible case study.

Topics: technical decisions made, what broke, what worked, what the twin got right vs. wrong, the experience of being evaluated by your own thinking.

Pillar 5: The Practical Edge
For engineers and technical hiring managers who want to use or evaluate CognitiveHire. How-to content. Setup. Interpretation. Use cases.

Topics: how to read a cognitive profile, what domains to look for in a hire, how to build your own interaction corpus, the difference between volume and depth.

---

Format Matrix

---

Voice Guide

This is Mohamed's voice. Not a brand voice. His.

What it sounds like:
- Short sentences. Declarative. No hedging.
- Specific numbers, not round ones. "112,000" not "over 100,000."
- Admits what doesn't work yet. Honesty is the differentiator.
- Never corporate. Never "excited to announce."
- Gets to the point in sentence one. No context-setting preamble.
- First person, builder's perspective. "I built" not "we developed."

What to avoid:
- Em dashes. Use commas or periods instead.
- "Leverage," "delve," "craft," "seamless," "game-changer," "cutting-edge," "holistic," "synergy."
- Vague claims. "Massive scale" means nothing. "112,000 interactions across 11 domains" means something.
- Passive voice on key claims. "I found" beats "it was found."
- Overpromising. CognitiveHire is early. Say so.

Voice checkpoint (run every post through these):
1. Would Mohamed say this out loud to a peer, not an audience?
2. Is there at least one specific number or fact?
3. Does it get to the actual point by line 2?
4. Is there anything a PR department would be proud of? (If yes, cut it.)

---

Repurposing Chain

Long-form blog post (2,000 words)
        |
        +-- LinkedIn long-form post (condensed to 600-800 words, hook rewritten)
        |           |
        |           +-- LinkedIn observation post (one key insight, 150-200 words)
        |
        +-- Twitter/X thread (10-12 tweets, one insight per tweet)
        |           |
        |           +-- 3-5 standalone single-tweet takes extracted from the thread
        |
        +-- YouTube script (takes the narrative structure, adds demo or visual layer)
        |
        +-- Email newsletter section (reader version: more context, more personal)

One piece of foundational work generates 8-12 deployable content units. Batch the writing. Distribute over 2-3 weeks per piece.

---

Production Workflow

Weekly rhythm:

Monday: Write 1 long-form piece (blog or LinkedIn article). This is the week's anchor.
Tuesday: Break the anchor into Twitter thread + 2-3 standalone tweets. Schedule them.
Wednesday: Write 3 short LinkedIn observation posts for the rest of the week.
Thursday: Record any video content if applicable. No more than 1 video per week at launch.
Friday: Review what performed. Note hooks and formats that got traction.

Batch cycle:
Every 2 weeks, run a 2-hour content batch session. Write 8-10 short posts in one sitting. Schedule them out. This prevents daily writing pressure and maintains quality.

Asset dependencies:
- Twitter/X: no dependencies, publish immediately
- LinkedIn posts: no dependencies, schedule 24-48 hours out
- YouTube: requires screen recording or camera setup. Batch record 2-3 videos in one session.
- Blog: requires final edit pass + CognitiveHire site to be live for links to resolve

---

STAGE 5: SCHEDULE

Week 1: Establish the Thesis (March 31 to April 6)

Goal: Plant the core idea. Get the contrarian take in front of hiring managers. No product pitch yet.

---

Week 2: The Story (April 7-13)

Goal: Make it personal. Humanize the thesis. Give people a narrative to follow.

---

Week 3: The Problem Frame (April 14-20)

Goal: Make the pain vivid for hiring managers. Show them the gap.

---

Week 4: The Build (April 21-27)

Goal: Show the work. Builders follow builders. Make the technical credibility visible.

---

STAGE 6: PUBLISH

Pre-Publish Checklist

Before any post goes live:

• [ ] Voice check: no em dashes, no AI-isms, no corporate tone
• [ ] Specificity check: at least one concrete number or named thing
• [ ] Hook check: point reached by line 2 or earlier
• [ ] Platform-native check: does this read like content from this platform, not pasted from another?
• [ ] CTA check: is there one? Is it appropriate to the content's job (build trust vs. drive action)?
• [ ] Link check: any links working and not sending to broken pages?
• [ ] Tag check: any names mentioned? Tag them on LinkedIn if relevant.

---

Platform-Specific Adaptation Notes

LinkedIn:
- First 2 lines determine whether someone hits "see more." Make them do work.
- LinkedIn rewards posting directly, not linking out. Embed the insight, then link at the bottom if needed.
- No bullet lists in the first 3 lines. They collapse to "..." in the preview.
- Formatting: line breaks over paragraphs. Single sentences. White space reads as confidence.
- Best posting times: Tuesday-Thursday, 7-9 AM or 12-1 PM EST.
- Comment on the post yourself within the first 30 minutes. Adds a thread, boosts early signals.
- Example adapted hook for LinkedIn:
> "I have 112,000 AI interactions in a database. Every one of them is timestamped, domain-tagged, and linked to an outcome.
> Your resume has none of that.
> Here's what I can tell about how you think from the first one."

Twitter/X:
- Thread format: first tweet stands alone as a complete take. Don't make it a teaser.
- Number the tweets only if there are more than 6 and order matters.
- Last tweet in a thread: one sentence. Action or reflection. Not a summary.
- Engagement bait questions at the end of threads work. Don't pretend they don't.
- Best times: 8-10 AM EST, 7-9 PM EST.
- Example thread opener:
> "Your resume is a document you control.
> Your AI interaction history is a document you can't fake.
> The recruiting industry hasn't figured out the difference yet. (thread)"

YouTube:
- First 30 seconds of every video must hook on a specific claim, not a topic.
- Thumbnail text: 3 words max. Specific number or bold claim.
- No intro music. Start talking.
- For demo videos: screen share at full resolution, cursor visible, narration real-time. Don't script it to death. Authentic demos > polished ones.
- For founder story videos: face on camera. Single location. No background effects. Plain background is fine.
- Description: paste the full post text from LinkedIn as the description body. SEO benefit, minimal extra work.
- End screen: one CTA. "Subscribe" or "Try CognitiveHire." Not both.

Blog/Site:
- Discovery-first openings. Start with the specific moment, not the conclusion.
- Headers should be specific claims, not topic labels. "What I Found at 4.3
- Include one architecture table or data breakdown per post. These get shared.
- End every post with a concrete state of the build: what works, what doesn't, what's next. Honesty here earns repeat readers.
- Target 1,200-2,000 words. Long enough to rank, short enough to read.

---

Cross-Platform Sequencing

Post order for maximum reach on a single piece of content:

1. Twitter/X thread drops first (highest organic reach, fastest feedback)
2. LinkedIn long-form drops same day or next morning (different audience, different format)
3. Blog article drops 3-5 days later (more polished, more context, references the earlier posts)
4. YouTube drops week 2 (builds on established interest, gives visual learners an entry point)
5. Email (once list is built): curated digest every 2 weeks, links to 3-4 best pieces

---

STAGE 7: ANALYZE

Week 1 Success Metrics

Primary signals (these matter):
- LinkedIn: 500+ impressions per post on average. At least one post crossing 1,000.
- Twitter/X: At least one thread with 50+ retweets or quotes.
- Inbound DMs or comments from people in the target audience (hiring managers, engineers, investors). Even 3-5 is meaningful in week 1.
- Email list: 50+ signups if CTA is present (even just "DM me if you want early access")

Secondary signals (context, not decisions):
- LinkedIn follower count change
- Twitter/X follower count change
- Blog page views

Week 1 questions to answer:
- Which pillar generated the most engagement?
- Did the contrarian takes (Path D) outperform the educational takes (Path A)?
- Are hiring managers or engineers more responsive?
- What was the best-performing hook format?

---

Month 1 Success Metrics

Audience built:
- LinkedIn: 200+ new followers or connections from target audience
- Twitter/X: 100+ new followers from target audience
- Email: 200+ signups (if capture mechanism is live)

Content performance:
- At least 2 posts crossing 5,000 LinkedIn impressions
- At least one thread going beyond the immediate follower base (shares, quote-tweets)
- Blog: 500+ unique visits across all posts combined
- YouTube: 500+ views on at least one video if channel is live

Product signal:
- 50+ people asking about CognitiveHire or requesting access
- At least 5 conversations with potential users or investors initiated from content
- At least one inbound from a hiring manager wanting to use the platform

Content operations:
- Publishing cadence maintained: 4 LinkedIn posts/week, daily Twitter/X presence
- Repurposing chain tested and functional (1 blog post feeding 5+ pieces)
- Batch workflow established

---

Month 3 Success Metrics

Audience scale:
- LinkedIn: 1,000+ followers from target audience
- Twitter/X: 500+ engaged followers
- Email: 1,000+ subscribers
- YouTube: 200+ subscribers and at least one video with 1,000+ views

Traction signals:
- Consistent inbound from content (at least 20 leads/month)
- 2-3 pieces ranking on Google for specific search terms (cognitive hiring, AI interaction history, AI-based hiring)
- At least one piece cited or shared by someone with an existing audience in HR-tech or AI
- Podcast or newsletter feature request

Content quality signals:
- Average LinkedIn engagement rate above 3
- Return readers on blog (20
- Twitter/X threads generating reply conversations, not just likes

Business conversion:
- CognitiveHire waitlist or beta: 500+ signups
- At least 2-3 companies in active conversations about piloting the platform
- Content directly attributable to at least one substantive investor conversation

Pivot triggers:
- If educational content (Pillar 1, 5) consistently outperforms provocative content (Pillar 3): shift tone, less contrarian.
- If hiring managers aren't engaging but engineers are: adjust ICP focus. Engineers are users first, buyers second. Build the user base.
- If LinkedIn outperforms Twitter/X by 5x or more: reduce Twitter frequency, increase LinkedIn depth.
- If YouTube demo video outperforms all written content: shift to video-first with text as support.

---

Appendix: Copy Examples

LinkedIn Post Examples (ready to publish or adapt)

Pillar 2 (Data/Authority) example:

I have 112,000 AI interactions in a database.

Timestamped. Domain-tagged. Pattern-indexed.

Here's one thing I found: the people who ask the most specific questions also produce the most reliable systems. Not a correlation I expected. The data is pretty clear about it.

Your prompting style is a cognitive fingerprint. You're leaving it everywhere. Most companies haven't thought to look at it.

CognitiveHire does.

More on what the data shows this week.

Pillar 3 (Interview Problem) example:

Last month I could have written myself a perfect resume in 15 minutes using AI.

So could every other engineer.

That's not a complaint. That's the problem hiring managers are now dealing with. The resume as a signal is gone.

What's left?

The questions you asked to build the thing. Those are harder to fake.

That's what CognitiveHire is built on.

Pillar 4 (Builder Stories) example:

The first time I ran CognitiveHire on my own interaction history, I expected to see my strongest domains.

I expected iOS. Multi-agent systems. ML.

I found those. But I also found a pattern I hadn't noticed: every time I worked on N'Ko language AI, my question quality went up significantly. The questions got more precise. The follow-ups got more structured.

I care about that project differently than I care about other projects. The data shows it.

That's the kind of signal CognitiveHire surfaces. Not what you list. What you actually care about.

---

Twitter/X Thread Examples

Thread: "The Resume Is Broken" (8 tweets)

1/ Your resume is a document you control.
Your AI interaction history is a document you can't fake.
The recruiting industry hasn't figured out the difference yet.

2/ I have 112,000 AI interactions in a database.
Every conversation is timestamped, domain-tagged, and linked to actual output.
No recruiter has ever asked to see it.

3/ Why not?
They don't have a way to read it.
That's the gap CognitiveHire is built to close.

4/ Traditional interviews test your ability to explain what you know under pressure.
That's not what your job will ask you to do.
It never was.

5/ AI interaction history tests something else:
How do you decompose a problem?
What do you do when you're stuck?
How does your question quality change as a project gets harder?

6/ The question you ask when you're stuck is more revealing than anything on your resume.
It's also something you can't rehearse.

7/ I built CognitiveHire because I'm tired of hiring processes that test interview performance.
I want to evaluate work performance.
These are not the same thing.

8/ If you're a hiring manager dealing with AI-generated resumes and wondering what signal is left:
this is what's left.
DM me or follow for more.

---

YouTube Video Concepts

Video 1: "What a Cognitive Twin Profile Actually Looks Like" (5 min, demo)
- Opens on screen share of actual CognitiveHire output
- No intro, no music, immediate narration: "This is what 2 years of AI interactions turns into."
- Walks through 3 profile sections: domain distribution, question quality over time, peak cognitive engagement periods
- Ends with: "This is what I'd show a hiring manager instead of a resume."
- Thumbnail: "My AI Profile" with a visual of the data output

Video 2: "I Built a System That Turns 2 Years of AI Conversations Into a Hiring Profile" (10 min, founder story)
- Face on camera. Plain background. No graphics.
- Opens with the specific frustration moment: "I was looking at a stack of resumes and they all said the same things."
- Walks through the idea, the first prototype, what worked, what didn't
- Shows the data (screen share) at the 5-minute mark
- Ends with the thesis stated plainly: "The question is worth more than the answer."
- Thumbnail: Mohamed on camera, text "112,000 Conversations"

Video 3: "The 3 Things I Can Tell About a Candidate From Their AI History" (7 min, educational)
- Whiteboard or screen-annotated walkthrough
- Specific, numbered, actionable
- Each section has a real example from Mohamed's own data
- Opens with: "I've analyzed 112,000 AI interactions. Here are the 3 patterns that actually predict how someone thinks."
- Thumbnail: "3 Signals" with numbered list visual

---

Last updated: March 2026
Next review: April 27 (end of 4-week launch calendar)