Derek Johnson
Verdict PURSUE WITH CAVEATS
There is a real gap (80% of jobs not on traditional boards is roughly accurate), real persona pain (early-career invisibility into local market), and real differentiation (geographic-first + financial reality) — but building a job board is one of the most capital-intensive product types in software, and four personas at launch is two too many.
Real wedge, viable persona, dangerous scope.
- 01Narrow to ONE city + ONE persona for 12 months — Atlanta + Transitional Explorer (music majors, AI-anxious creatives) — and validate the Discovery Ratio at 60%+ with 100 paying users at $9/mo before expansion [+11 → 69]
- 02Land 1 college partnership (Spelman, Morehouse, GA Tech career services) for white-label distribution — they pay $10-25K/year, give you 5K+ engaged early-career users, and you skip CAC for the first 1000 users [+13 → 71]
- 03Drop the 4-persona ambition entirely; ship the financial calculators first (free, viral, no jobs data needed), use that to build email list of 10K early-career grads, then layer the map ON the audience [+9 → 67]
Same domain, same research, same vendor pain. Three nearby ideas with their own confidence estimates derived from the analysis above.
College Career-Services White Label
64%B2B SaaS sold to college career-services offices.
Colleges have $5-15K/year career-services budgets, are desperate to demonstrate post-graduation outcomes (US News rankings depend on it), and bring a captive audience of 5-25K students per institution. You skip individual CAC entirely. One signed school = 18 months of runway.
Validate: One pilot with one school: ~$8K + 4 months sales cycle
Risk: Career-services procurement is slow (12-month sales cycles) and is becoming a crowded category (Handshake dominant; Symplicity entrenched).
Take-Home Calculator Standalone
52%Just the financial-reality calculators, free, viral.
The Take-Home + Commute + Roommate calculators are inherently shareable on TikTok/Instagram ("watch your $75K offer become $52K after taxes/commute"). No jobs data needed. Free user acquisition engine. Layer the job map on this audience later.
Validate: 4-week sprint + 100 TikTok posts: ~$2K
Risk: Calculators are commoditized; one TikTok-native creator can rebuild this in a weekend.
Atlanta-Only Hidden Map
67%One city, one persona, prove the model.
Geographic-first means the product LOOKS the same at city scale and at national scale. Better to prove the Discovery Ratio metric in one market (Atlanta has the right combo of dense employment + diverse industries + accessible network for the founder). Expand city-by-city. Smaller TAM but defensible per-city moat.
Validate: MVP map + 50 paying users at \$9/mo: ~$15K, ~6 months
Risk: City-by-city expansion is slow; reaching $1M ARR requires 10+ cities, each with its own expansion cost.
Add context the analysis missed, change a constraint, or disagree with a specific conclusion. The verdict will re-evaluate, and you will see what moved — and what did not.
Included in your $29. Two rounds max — use them wisely.
A real wedge (the hidden job market is genuinely 60-80% invisible to traditional boards) with strong persona insight (the "music major in Atlanta wondering where her skills apply" is a vivid, underserved customer) — but priced as four products at once. The crawler-agent infrastructure is real engineering capex (6-12 months pre-revenue). The survival path requires brutal narrowing: one city, one persona, one acquisition channel for 12 months. The financial-reality calculators are the most-shareable atom and may be the right wedge before the map.
Gap: Owns the surface category. $1.5B in annual profit. If WhenIGraduate works, Indeed builds the map UX in 12 months and crushes it. The defense is depth in financial tools they don't care about.
Gap: Dominant in college career services. 1,400 partner schools. Closed network model. The white-label angle WhenIGraduate could pursue is exactly Handshake's territory.
Gap: Has the AI augmentation score (LinkedIn AI Skills Insights). Has the skills taxonomy (LinkedIn Skills Graph). Has 1B users. The "AI score per job" feature WhenIGraduate plans, LinkedIn already ships.
Gap: Owns the financial-calculator category at scale. Their take-home tool is free, well-ranked in Google, and they own the SEO. To win calculator-led acquisition you have to outrank them.
Gap: City-specific job + company-info platform. Closest analog to "city-first" approach. Established in 22 cities. Profitable. The model works; the competition is real.
The "hidden job market" claim is real and well-supported. Closing this gap is genuine value — but the gap exists because crawling /careers pages at scale is hard, not because nobody noticed.
https://www.bls.gov/opub/mlr/Demonstrates the college-career-services market's economics. WhenIGraduate would be entering Handshake's territory if it pursues the white-label angle.
https://www.crunchbase.com/organization/joinhandshakeFor a $9/mo subscription, payback is 5-8 months IF retention holds. Tight but workable. Free-to-paid conversion will need to clear 3% to be sustainable.
https://andrewchen.com/cac-by-vertical/- 01
Four personas at launch is two too many. Which ONE is the wedge for the first 12 months — and what tells you the other three should wait?
- 02
Crawling 50K+ company career pages weekly, parsing them into a unified skills taxonomy, is 6-12 months of engineering before users see value. What's the bridge product that captures audience while you build it?
- 03
If the financial calculators (take-home, commute, roommate) are the most shareable atom, why not ship those first and use them to build a 10K+ email list before building the map?
- 04
When LinkedIn rolls out city-based discovery + AI augmentation scoring (already partly shipped), what protects WhenIGraduate? "We're prettier" is not a reason. What's the structural advantage that survives a LinkedIn product update?
- 05
White-label to colleges OR direct-to-consumer freemium — which sales motion are you actually building? Both means neither gets adequate engineering and sales attention.
You have one of the strongest persona insights in this entire portfolio. The next 60 days are about brutal scope discipline — one city, one persona, one channel.
- Pick ONE city + ONE persona
Atlanta + Transitional Explorer (music major, AI-anxious creative). Write a one-paragraph profile of the exact user. Every product decision for 12 months serves this user.
2 hours - Cold-call 10 music/creative grads in Atlanta
Find them via Spelman, Morehouse, Atlanta College of Art, GSU film program. Ask: "What jobs exist for you that you can't find on Indeed?" Listen for what they say in their own words.
6-8 hours - Ship the Take-Home calculator (free, no jobs data)
No-code, can be live in 1 weekend. Drive 1,000 visitors via TikTok ("watch your $75K offer become $52K"). Test if calculator-led acquisition actually works at this audience before building the map.
$200 + 2 weekends
- Map 200 Atlanta companies' /careers pages
Manual spreadsheet, no crawler agent yet. List: company, address, current openings, posted within 30 days. This is the proof of the Discovery Ratio (vs. Indeed's Atlanta listings).
40 hours = 1 dedicated week - Pitch 1 Atlanta college career-services office
Spelman, Morehouse, or Georgia State. Pitch: free pilot for one semester, white-label "Spelman Hidden Map," they cover marketing to students. Even if rejected, the conversation reshapes what you're selling.
8 hours + 4 weeks waiting - Launch the Atlanta map MVP for Transitional Explorers
Curated 200 companies, the Take-Home calculator, the AI-augmentation tag (scraped from public LinkedIn data). Ship to 50 hand-recruited beta users from your cold-call list.
$3-5K + 6 weeks engineering
- Write the kill criterion
"If by [date 90 days out] the Atlanta MVP doesn't have 100 weekly active users with 60%+ Discovery Ratio (jobs on the map not on Indeed), I shut down or pivot."
15 min - Cap the build budget
No more than $25K total spend before reaching the 100-user milestone. The crawler agents and skills taxonomy can wait until Atlanta is proven manually.
30 min planning
The pitched scope sits in the middle of every axis. The defensible position is bottom-left (DTC + one city) before any expansion to enterprise or multi-city.
SAMPLE REPORT · v1.3.2 · ILLUSTRATIVE — NOT A LIVE RUN · 7 SOURCES SHOWN
HONESTY-AUDIT PROMPT SHA256:7f867a14952f878d… · SAMPLE REPORT