~ / AI / The 100k Database Applied to Me

The 100k Database Playbook, Applied to My Situation

This is a brutally honest, self-directed application of Fraser's 100k Database playbook to my specific context: software engineer, northern France, currently unemployed, 15,243 LinkedIn connections, chronic focus problems, past startup failure (Valyent), and a desperate need to make money before having to take a job.

The Diagnosis: Why I Haven't Made Money Yet

Fraser made $1,182 in his first 2 weeks with a Google Sheet. I've been vibecoding PHP runtimes in Rust for fun. Here's the gap:

The Valyent lesson I keep citing but never applying: "We spent a year engineering instead of looking for customers." I'm doing it again. Every single time. Palmframe, Hypercode, Wikistral, Hypercafe, PHP.rs — all engineering, zero customers.

My Unfair Advantages (That I'm Not Using)

1. 15,243 LinkedIn connections. Fraser had zero audience when he started. I have 15,000 people, 2,400+ of whom are founders/CEOs, 1,000+ CTOs/tech leads, 6,000+ engineers. This is a gold mine I've never touched.
2. Deep French tech network. Scaleway (76), OVHcloud (68), Mistral AI (67), Qonto (25), Decathlon Digital (19), lemlist (14), Ledger (16), BeReal (11). These aren't randos — these are real companies where I know real people.
3. I can build things fast. Fraser paid $150 for his web app. I can vibecode an entire platform in a weekend. This is an advantage ONLY if I first validate that people will pay.
4. I'm in France. The French tech market is underserved by English-language DaaS products. Language barrier = moat.
5. I understand vibecoding. Jonathan Serra is already working with vibecoders. This is a real, emerging market I know personally.
6. I already have AI market research infrastructure. My /ai section has 30+ deep analyses. I understand markets better than most founders.

The 3 Most Realistic DaaS Ideas for Me

Applying Fraser's criteria: (1) B2B, (2) I have domain knowledge, (3) I can populate it, (4) I can reach buyers through my existing network.

Idea 1: French Tech Startup Database

A curated, enriched database of every French tech startup: founder contacts, funding stage, tech stack, team size, hiring status. Updated monthly.

Who buys it?

• Dev agencies & freelancers (479 in my network) — they need to prospect startups that will hire them
• SaaS companies selling to startups (devtools, infra, fintech) — Datadog, Supabase, lemlist-type companies
• Recruiters — who's hiring, who's the CTO, what's the email
• VCs & accelerators (68 in my network: OSS Ventures, Hexa, etc.) — deal flow

Why me?

I already have 11,000+ unique companies in my LinkedIn. I can scrape Station F directories, French Tech listings, Crunchbase France, LinkedIn company pages. I'm French, I know the ecosystem. No one is doing this well in French.

MVP data points

Company name, founder name(s), founder email, founder LinkedIn, company URL, tech stack (if detectable), funding stage, number of employees, city, industry vertical.

Price

€149/mo (cheaper than ZoomInfo, more niche, in French). One-time download: €297.

First 10 customers

The 479 freelancers in my LinkedIn. They need leads. I have leads. It's that simple.

Idea 2: Vibecoder Service Provider Database

A curated database of vibecoders, their tech stacks, portfolios, pricing, and availability. Plus: a directory of vibecoded apps with what stack was used and who built them.

Who buys it?

• Non-technical founders who need to find vibecoders to build their MVP
• Agencies looking to subcontract to vibecoders
• Jonathan Serra-type consultants who help vibecoders consolidate — they need to find vibecoders to sell to
• Tool companies (Cursor, Lovable, Bolt, Replit) wanting to understand their user base

Why me?

I'm literally vibecoding right now. I know the tools, the stacks, the community. This market is brand new — nobody has organized it yet. First mover advantage.

Risk

Market may be too nascent. Could be hard to monetize at $150+/mo. Better as a lead magnet / community play initially.

Idea 3: French Tech Conference Attendee Database (Fraser's Exact Model)

Database of attendees from French tech conferences: VivaTech, Web2Day, DevFest, France Digitale Day, etc. Enriched with direct contact info.

Who buys it?

• Sponsors & exhibitors who paid €5K-50K for a booth and want to reach people they didn't meet
• SaaS companies who couldn't attend but want the contacts
• Recruiters hunting for talent at these events

Why me?

I attend some of these events. I can access event apps (Fraser's pro tip: you don't need to attend, just pay for the event app). French conferences are under-exploited compared to US ones.

Risk

Legal risk in France (GDPR is stricter than in Fraser's case). C&D from conference organizers. Needs careful positioning — "likely attendees" rather than "verified attendees."

My Recommendation to Myself: Idea 1

The French Tech Startup Database wins because:

1. I already have the seed data (11,000+ companies in my LinkedIn export)
2. I have the buyers in my network (479 freelancers, 68 VCs, hundreds of SaaS company employees)
3. I have the technical ability to enrich data fast (scraping, APIs, AI)
4. Nobody does this well for France specifically
5. It's B2B — low chargeback, high stickiness
6. I can use it myself (to find customers for whatever I build next)
7. It doesn't require me to attend conferences or buy event app access

The Execution Plan: 14 Days to First Revenue

Days 1-3: Build the Google Sheet

DO NOT WRITE ANY CODE. NOT ONE LINE.

1. Create a Google Sheet with columns:
   Startup Name | Founder(s) | Founder Email | Founder LinkedIn |
   Company URL | City | Industry | Funding Stage |
   Team Size | Tech Stack | Last Updated

2. Populate it from my OWN LinkedIn export:
   - Filter connections by "Founder" / "Co-Founder" / "CEO" titles
   - That's ~2,400 people with companies
   - I already have their LinkedIn URLs
   - Start with the top 500 (most recent connections first)

3. Enrich the first 200 leads:
   - Use Apollo.io free tier (250 credits/mo) for emails
   - Use company websites for tech stack (or BuiltWith free)
   - Use LinkedIn company pages for team size + funding

4. Quality check: verify 20 random entries manually

OUTPUT: Google Sheet with 200-500 enriched French tech startups

Days 4-5: Validate with My Network (Fraser's "Accidental Share" Method)

1. Share the Google Sheet (view-only, no download) with
   30-50 freelancers from my LinkedIn connections.

   Message (LinkedIn DM, not email):
   "Hey [NAME] — I put together a database of French tech
   startups with founder contacts. Thought it might be useful
   for prospecting. Here's the link: [GOOGLE SHEET URL]
   Let me know what you think!"

2. Track who views the sheet (Google Sheets shows active viewers)

3. When they stop viewing, follow up:
   "Hey, did you find the list useful? Anything missing
   that would make it more valuable for you?"

4. Listen to feedback. Do NOT sell yet.
   - What data points do they want?
   - What industries are they looking for?
   - Would they pay for this if it was updated monthly?

5. If 5+ people say "yes I'd pay for this":
   Create a Stripe link at €149/mo and share it.

   "Cool — I'm thinking of keeping it updated monthly
   with new startups, verified emails, tech stack data.
   €149/month if you want ongoing access.
   Here's the link: [STRIPE]"

Days 6-10: Cold Outreach to My Own Network

1. Segment my LinkedIn connections:

   SEGMENT A (highest priority): Freelancers (479 connections)
   → They need clients. My database IS their client list.
   → Message angle: "I compiled a prospecting list of
      French startups with founder emails"

   SEGMENT B: Founders of devtools/SaaS (selling to startups)
   → They need to reach startup founders. I have them.
   → Message angle: "I have a database of 500+ French
      startup founders with direct emails. Thought it
      could help with your outreach"

   SEGMENT C: VCs / Accelerators (OSS Ventures, Hexa, etc.)
   → They need deal flow. I have emerging startups listed.
   → Message angle: "I'm building a French startup database.
      Would this be useful for deal sourcing?"

   SEGMENT D: Recruiters and HR
   → They need to reach founders/CTOs at startups
   → Message angle: "Database of French tech startups
      with CTO/founder contacts"

2. Send 20-30 LinkedIn DMs per day (not automated, personal)

3. Use Fraser's golden rule: do NOT sell in the first message.
   Get a reply first. The sale happens in the conversation.

Days 11-14: Iterate Based on Feedback

IF 5+ subscribers:
  → Keep going. Add 100 new startups per week.
  → Start enriching with Clay or Findymail for better data.
  → Consider building a simple search web app (I can
     vibecode this in 2 hours, but ONLY after 10+ subscribers).

IF 0 subscribers but positive feedback:
  → Adjust pricing (try €99/mo or €49/mo)
  → Try one-time download at €197
  → Add more data points people asked for
  → Keep sharing the free sheet to build demand

IF 0 subscribers and no interest:
  → Pivot to Idea 2 or 3
  → Do NOT spend more than 2 weeks on a dead idea
  → The whole point of the Google Sheet MVP is that
     it costs nothing to fail

The Anti-Alexis Patterns (Rules to Follow)

1. NO CODE for 14 days. Google Sheet only. If I catch myself opening VS Code to "just build a quick prototype," I stop. Fraser made $87k ARR from a Google Sheet. I can survive 2 weeks without touching code.
2. NO new project ideas for 14 days. No Hypercafe. No PHP.rs. No observability prototypes. No pentesting tools. One idea, 14 days, full commitment.
3. 20 LinkedIn DMs per day, every day. This is the work. Not coding. Not research. Not reading books about startups. Sending messages to real people.
4. NO building in public on Twitter. Fraser got 0 subscribers from 14k Twitter followers. I will too. LinkedIn is where my buyers are.
5. Charge money on day 5. Not day 30. Not "when it's ready." Day 5. The Stripe link goes out on day 5, period.
6. Track revenue, not vanity metrics. I don't care about views, likes, followers, GitHub stars. The only number that matters: € in Stripe.
7. If it doesn't work in 14 days, pivot immediately. Don't romanticize the idea. Don't "give it more time." Fraser moved fast. So should I.

The Math: What €149/mo Looks Like

Fraser's costs: $500/mo for outreach team + $0.15/lead enrichment. My costs would be even lower because I can automate enrichment with my own scripts. 30 subscribers would already be more than enough to live on in northern France.

The Upsell Ladder (After Validation)

From Fraser's playbook, applied to my context:

1. Base product: French startup database — €149/mo
2. One-time download: For people who won't subscribe — €297
3. Custom list building: "Give me startups in [specific vertical] with [specific criteria]" — €500-1,000 one-time
4. Outreach-as-a-service: I'll email the startups FOR you using my data — €1,500/mo
5. Spin-off databases: French AI startups only, French fintech only, French devtools only — each €99/mo
6. The web app (later): Searchable, filterable, with saved searches and alerts — €199/mo premium tier

Why This Is Different From Everything Else I've Done

Start: Tomorrow Morning

07:00 — Create the Google Sheet. Set up columns.
08:00 — Export my LinkedIn connections CSV to a working format.
09:00 — Filter for founders/co-founders/CEOs. That's ~2,400 rows.
10:00 — Take the first 300. Add to the sheet.
11:00 — Start enriching with Apollo free tier (emails).
14:00 — Enrich company data (URL, team size, city, industry).
16:00 — Quality check 30 random entries.
17:00 — Sheet is live. Share with 10 freelancers from my network.
17:30 — Go for a walk. Let them look at it.
20:00 — Check who viewed it. Follow up with anyone who did.

Day 1 output: 300 enriched leads in a Google Sheet,
shared with 10 real potential customers.

No code. No Rust. No PHP. No blog post about it. Just a spreadsheet and a Stripe link.

Self-directed analysis based on The 100k Database by Fraser (@iamfra5er) and my own LinkedIn network analysis.