CARTOGRAPH
Cartograph — Go-To-Market Plan
Company: Cartograph (company id retro-games) · Date: 2026-10-01
Stage: concept-stage company — this is the launch plan; the app is scoped, not yet built, and signup is not open.
All numbers, prices and targets below are hypotheses to validate, not commitments.
1. Positioning
For people who collect physical retro games, Cartograph is the collector's catalog app that turns a photo into a verified, dated, sourced record of what you own and what it's worth — because spreadsheets lose editions, price guides go stale, and nothing else treats a loose cart, a CIB and a sealed copy as different objects with different markets.
Positioning anchors
- Verify, don't guess. Edition confirmation (region, print, label variant) is mandatory — a first-print isn't a reprint.
- Dated & sourced. Every estimate shows its date and source; we never show a bare number.
- Tactile. The shelf feels like your shelf — the archive is the product, not a form.
We are not: a marketplace (yet), a ROM/emulation anything, a generic price-guide site, a grading service.
2. First launch segment
Collectors who already maintain an inventory somewhere painful (spreadsheet, Notes app, PriceCharting lists) — concentrated, reachable, and they feel the problem weekly. Beachhead consoles: NES/SNES/N64/Game Boy/Genesis/PS1, matching the pilot catalog.
Secondary reachable segments we design for but don't launch at: flea-market buyers (speed case), retro stores (batch case), haul/shelf creators (showcase case). The accuracy-validation pilot recruits from all four.
3. Channel experiments (cheap, measurable, sequential)
| Channel | Experiment | Hypothesis | Metric |
|---|---|---|---|
| Retro-game subreddits & Discords | "Scan accuracy challenge": post 20-item accuracy results; recruit pilot testers | Accuracy demos convert skeptics better than feature lists | Pilot signups per post; match-rate replies |
| YouTube/TikTok haul & shelf creators | Give 5 mid-size creators pilot builds; they film real shelf digitization | Watching a shelf build itself is the strongest ad | View→site CTR, pilot signups via creator codes |
| Retro stores & flea markets | "Counter pilot": store scans intake for a week, we publish anonymized accuracy report | Stores are credibility engines + data engines | Items scanned/store/week; correction rate |
| SEO | "What is my [game] worth?" long-tail pages w/ dated sample estimates | High-intent queries exist at scale | Signup rate on estimate pages |
| Retro conventions (PRGE, Midwest Gaming Classic…) | Booth-adjacent scan station: "bring a cart, leave with a record" | Physical demo converts on the spot | Scans/day, pilot conversion on-site |
Kill criteria per experiment set in advance: e.g., creator posts that yield <1% engaged CTR get cut after 2 posts.
4. Example creative (concepts, not final assets)
- "The dupe check" — 15s vertical video: buyer at flea market holds up a cart, scans, app replies "You already own this. NTSC-U, CIB, Good." Tag: Stop buying the same game twice.
- "Loose ≠ CIB ≠ Sealed" — carousel showing the same title's three different markets; last card: Cartograph knows the difference.
- "Shelf tour" — creator pans a physical shelf while the app's shelf renders it in sync.
- Accuracy receipt — a plain monospace receipt: "247 items scanned · 94.3% auto-confirmed · 11 corrections logged." Radical honesty as marketing.
5. First-100-user plan
- Founding 25 — hand-recruited from retro Discords/subreddits + personal network; 20-item minimum onboarding task; weekly 15-min calls.
- Pilot batch 26–75 — creator-driven signups + 2–3 retro stores running the counter pilot; in-app "flag a wrong match" wired to our triage queue.
- Pilot batch 76–100 — convention/flea-market recruits; waitlist opens (honest queue, no fake counter).
Each cohort gated on the previous one's accuracy metrics — see §7.
6. Pricing & cost hypotheses
- Free during pilot. Post-pilot hypothesis: free tier ≤100 items/20 scans per month; Pro ~$4.99/mo or ~$39.99/yr. Store tier later.
- Cost drivers per user: match-service inference (~fractions of a cent/scan at scale), catalog/pricing data license (likely the biggest fixed cost), CloudKit sync (near-zero at pilot scale). StoreKit takes 15% (small-business program) — 30% only above threshold.
- Validation: run Van Westendorp-style pricing questions in pilot calls + a paywall-dry-run screen after week 4 (measuring taps, not charging).
7. Metrics & validation gates
North-star for pilot: verified-scans-per-active-collector-per-week (items scanned where user confirmed edition — the product's core loop).
| Metric | Green (continue) | Red (fix before scaling) |
|---|---|---|
| Auto-match accuracy on pilot consoles | ≥90% top-1, ≥97% after user confirm | <85% top-1 → expand catalog before marketing spend |
| Scan→shelf completion | ≥80% | <60% → flow too slow, fix UX |
| D7 retention (weekly collectors) | ≥40% | <20% → collection habit not forming |
| Duplicate-detection save recall | users report it prevented ≥1 re-buy | feature invisible → resurface |
| Pilot→paid intent (dry-run) | ≥15% tap "subscribe" on dry-run | <5% → re-price or re-scope before building billing |
| Store pilot: scans per store per week | ≥100 | <30 → store persona deferred |
Gates: Gate A — accuracy ≥90% on pilot console set before any paid acquisition. Gate B — D7 ≥40% in batch 2 before App Store public launch. Gate C — pricing dry-run ≥15% before building StoreKit integration.
8. Risks & honest limits
- Catalog/pricing licensing is the critical path; no live estimates ship until a real source exists. (Landing page only ever shows labeled samples.)
- Edition ambiguity (variants without distinguishing marks) — mitigated by always-confirm UX and "needs a second photo" prompts (back label, spine code).
- Collector trust — one fake-looking estimate poisons the brand; that's why every number carries date + source.