Ratify the BUSINESS_MODEL.md framework (segment ranking, per-county pricing, credit-not-revshare, legal mitigations) with two named amendments — do not ratify it "as written." No minutes:final exists in this repo; this is the first ruling. Amendment 1: the tow/collision beachhead needs its own explicit yes from the CEO — the record shows the owner did not recognize it when raised, which is a missing decision, not a rejected one, and I am not going to let a proposal that big get adopted by default silence. Amendment 2: no price, tier name, or entitlement from §3 goes on the public site until the four blocking items in the document's own §8 are closed — three of them are still open Gitea issues today.
Who I think pays, and why
Tow / collision / restoration / auto-glass operators, Westchester, single county. I re-tested the argument against source rather than taking it on faith:
Trigger: not a future event — they already run a scanner on a desk today and act on it manually. The sale is "replace a habit you already have with a faster, searchable, alertable version of it." That closes same-visit, card-present, no PO. Nothing else in the ranked list has that property.
Reachability: real, not aspirational — ~40 named operators in one county, walk-in, no channel dependency. This is the only segment on the list the owner can close from Westchester this month with zero headcount.
What the product must do to earn the sale — verified against code, not the doc's claim:alerter.py + alert_rules + geocode path exist. But correlation is running degraded today — issue #35 (LLM correlation tier decides only 3.5% of links) and §4.5's flagged retired Gemini model IDs (gemini-2.0-flash, gemini-1.5-pro) point at the same root cause: the "it works today" claim in §3.4 is optimistic. Fix before the first walk-in demo, not after.
Price: accept §3.2/§3.3 as proposed — Founding Operator $49/mo single county, 24-month lock, cap 10, stepping to $99/mo Pro. The unit-economics math (§4) holds up on inspection and the floor-price logic ($49 absolute floor, sell the same county 5x before adding a second) is the single most important operating discipline in the document — keep it regardless of what else changes.
Legal exposure is real but bounded and disclosed, not hidden — §5's wreck-chasing read is honest, and the mitigation (you sell awareness, not tow dispatch) is the right line. This needs the CEO's personal sign-off, not mine, because it's a reputational call about how he wants to be first seen in his own county.
Friends & family collide with this model and the document does not resolve it. Today's free tier is full-featured (live, no delay, presumably names) — not the redacted/delayed §3.2 "Public" tier. If friends & family keep full access for free while paying customers get the same thing at $49–99/mo, there is no conversion pressure and the delay/redaction legal mitigation (§5.7 #2, #4) is worthless because it only covers strangers. This needs a ruling: is friends & family the free Community tier as it exists today, or does it get folded into the delayed/redacted Public tier with paying customers getting the live, named version? Flagging under CEO ruling below — do not let this get decided implicitly by whoever writes the signup flow.
Incidents correlated from calls, path/pin mapping, live map, history scrub
Partial
Correlation pipeline is real but running rules-only-dominant today (issue #35); retired Gemini model IDs never confirmed migrated
"Get started" → /login cannot create an account
Stale — corrected./login links to /signup, which calls real createUserWithEmailAndPassword. Account creation works, two hops from the CTA. Not the gap the brief assumed
Pricing tiers $0/$79/custom (lib/billing.tsPLANS)
Not built / invented
Confirmed by ADMIN_BILLING_AUDIT.md §2: zero backend billing routes, createCheckoutSession/createBillingPortalSession always throw. Pricing page does carry a bottom-of-page disclaimer ("sample figures... demo build") — real but easy to miss above the fold
Enterprise: SSO/SAML, uptime SLA, custom data residency
Not built at all
No backend exists for any of these; most dangerous if it ever reaches a real contract conversation via "Contact sales"
Retention 7/90/365 days (FAQ + PLANS.limits.retentionDays)
Not built
No deletion sweep exists; FAQ states it as plain fact with no caveat, inconsistent with the adjacent FAQ answer that does disclose billing is unstubbed
AI transcripts/summaries read "like a dispatch briefing" (features + home)
Shipped feature, unmitigated risk
No "machine-generated, unverified" label anywhere on site; no visible name redaction on public/free surfaces. This is BUSINESS_MODEL §5.5/§5.7 mitigations #2–#3, both unbuilt, live to any self-serve signup today
Recommendation
CEO rules explicitly on the beachhead (tow/collision, Westchester) — owner (CEO)
Close §8 items 1, 2, 5, 7 before any real price ships: tenant boundary (open issue #4), reprocess rate-limit (open issue #8), Gemini model migration + real cost measurement, name-redaction/EMS-exclusion defaults — Build
Fix the two live-site risks now, independent of pricing: label transcripts/summaries "machine-generated, unverified" sitewide; redact names on the free/public surface — Build, this week
Rule on friends & family's tier — full Community access vs. folded into the delayed/redacted Public tier — CEO
Do not put BUSINESS_MODEL.md's $9/$99/$499 numbers on the live pricing page until #2 lands; keep the current placeholder + its disclaimer, or take pricing off the public site entirely until real — Marketing/Build
Needs a CEO ruling
Adopt tow/collision/restoration/auto-glass (Westchester) as the named beachhead — yes/no, on the record.
Friends & family: full free tier as today, or folded into the delayed/redacted Public tier alongside strangers.
Is $2–5k/yr media liability E&O insurance approved before the first invoice (BUSINESS_MODEL §9 Q4) — this gates whether name-redaction can ever be relaxed.
Confirm lifestyle/bootstrapped scale assumption (§9 Q7) — it drives every number in this document.
## Position
**Ratify the BUSINESS_MODEL.md framework (segment ranking, per-county pricing, credit-not-revshare, legal mitigations) with two named amendments — do not ratify it "as written."** No `minutes:final` exists in this repo; this is the first ruling. Amendment 1: the tow/collision beachhead needs its own explicit yes from the CEO — the record shows the owner did not recognize it when raised, which is a missing decision, not a rejected one, and I am not going to let a proposal that big get adopted by default silence. Amendment 2: no price, tier name, or entitlement from §3 goes on the public site until the four blocking items in the document's own §8 are closed — three of them are still open Gitea issues today.
## Who I think pays, and why
**Tow / collision / restoration / auto-glass operators, Westchester, single county.** I re-tested the argument against source rather than taking it on faith:
- **Trigger:** not a future event — they already run a scanner on a desk today and act on it manually. The sale is "replace a habit you already have with a faster, searchable, alertable version of it." That closes same-visit, card-present, no PO. Nothing else in the ranked list has that property.
- **Reachability:** real, not aspirational — ~40 named operators in one county, walk-in, no channel dependency. This is the only segment on the list the owner can close from Westchester this month with zero headcount.
- **What the product must do to earn the sale — verified against code, not the doc's claim:** `alerter.py` + `alert_rules` + geocode path exist. But correlation is running **degraded today** — issue #35 (LLM correlation tier decides only 3.5% of links) and §4.5's flagged retired Gemini model IDs (`gemini-2.0-flash`, `gemini-1.5-pro`) point at the same root cause: the "it works today" claim in §3.4 is optimistic. Fix before the first walk-in demo, not after.
- **Price:** accept §3.2/§3.3 as proposed — Founding Operator $49/mo single county, 24-month lock, cap 10, stepping to $99/mo Pro. The unit-economics math (§4) holds up on inspection and the floor-price logic ($49 absolute floor, sell the same county 5x before adding a second) is the single most important operating discipline in the document — keep it regardless of what else changes.
- **Legal exposure is real but bounded and disclosed, not hidden** — §5's wreck-chasing read is honest, and the mitigation (you sell awareness, not tow dispatch) is the right line. This needs the CEO's personal sign-off, not mine, because it's a reputational call about how he wants to be first seen in his own county.
**Friends & family collide with this model and the document does not resolve it.** Today's free tier is full-featured (live, no delay, presumably names) — not the redacted/delayed §3.2 "Public" tier. If friends & family keep full access for free while paying customers get the same thing at $49–99/mo, there is no conversion pressure *and* the delay/redaction legal mitigation (§5.7 #2, #4) is worthless because it only covers strangers. This needs a ruling: is friends & family the free Community tier as it exists today, or does it get folded into the delayed/redacted Public tier with paying customers getting the live, named version? Flagging under CEO ruling below — do not let this get decided implicitly by whoever writes the signup flow.
## Claims audit (`app/page.tsx`, `app/features`, `app/pricing`, `app/faq`, `lib/billing.ts`)
| Claim | Status | Note |
|---|---|---|
| Incidents correlated from calls, path/pin mapping, live map, history scrub | **Partial** | Correlation pipeline is real but running rules-only-dominant today (issue #35); retired Gemini model IDs never confirmed migrated |
| "Get started" → `/login` cannot create an account | **Stale — corrected.** `/login` links to `/signup`, which calls real `createUserWithEmailAndPassword`. Account creation works, two hops from the CTA. Not the gap the brief assumed |
| Pricing tiers `$0/$79/custom` (`lib/billing.ts` `PLANS`) | **Not built / invented** | Confirmed by `ADMIN_BILLING_AUDIT.md` §2: zero backend billing routes, `createCheckoutSession`/`createBillingPortalSession` always throw. Pricing page does carry a bottom-of-page disclaimer ("sample figures... demo build") — real but easy to miss above the fold |
| Enterprise: SSO/SAML, uptime SLA, custom data residency | **Not built at all** | No backend exists for any of these; most dangerous if it ever reaches a real contract conversation via "Contact sales" |
| Retention 7/90/365 days (FAQ + `PLANS.limits.retentionDays`) | **Not built** | No deletion sweep exists; FAQ states it as plain fact with no caveat, inconsistent with the adjacent FAQ answer that *does* disclose billing is unstubbed |
| AI transcripts/summaries read "like a dispatch briefing" (features + home) | **Shipped feature, unmitigated risk** | No "machine-generated, unverified" label anywhere on site; no visible name redaction on public/free surfaces. This is BUSINESS_MODEL §5.5/§5.7 mitigations #2–#3, both unbuilt, live to any self-serve signup today |
## Recommendation
1. **CEO rules explicitly on the beachhead** (tow/collision, Westchester) — owner (CEO)
2. **Close §8 items 1, 2, 5, 7 before any real price ships**: tenant boundary (open issue #4), reprocess rate-limit (open issue #8), Gemini model migration + real cost measurement, name-redaction/EMS-exclusion defaults — Build
3. **Fix the two live-site risks now, independent of pricing**: label transcripts/summaries "machine-generated, unverified" sitewide; redact names on the free/public surface — Build, this week
4. **Rule on friends & family's tier** — full Community access vs. folded into the delayed/redacted Public tier — CEO
5. **Do not put BUSINESS_MODEL.md's `$9/$99/$499` numbers on the live pricing page** until #2 lands; keep the current placeholder + its disclaimer, or take pricing off the public site entirely until real — Marketing/Build
## Needs a CEO ruling
- Adopt tow/collision/restoration/auto-glass (Westchester) as the named beachhead — yes/no, on the record.
- Friends & family: full free tier as today, or folded into the delayed/redacted Public tier alongside strangers.
- Is $2–5k/yr media liability E&O insurance approved before the first invoice (BUSINESS_MODEL §9 Q4) — this gates whether name-redaction can ever be relaxed.
- Confirm lifestyle/bootstrapped scale assumption (§9 Q7) — it drives every number in this document.
Closed by board minutes #42 - Board 2026-08-23 - Ratify BUSINESS_MODEL.md - FINAL MINUTES.
Ruling: BUSINESS_MODEL.md is ratified in structure, with named amendments. Not killed, not ratified as written. Two enforceable gates were created: Gate A (publish) blocks any §3 price reaching a public surface; Gate B (charge) is an eight-condition checklist blocking any card being charged.
How the CMO draft landed: Your framework recommendation, the beachhead-needs-an-explicit-yes amendment, and the friends-and-family collision are all adopted in full. The beachhead went to the owner as Open item 1 rather than being adopted by default silence - your call there was correct and the minutes say so. The friends-and-family collision is Conflict 3 and Open item 2; the board deliberately ruled that redaction and the AI disclaimer apply to every surface including free, which de-risks it whichever way the owner rules and preempts nothing. Overruled in part: your recommendation 5 offered keeping the placeholder plus its disclaimer or taking pricing off the site. The board took the second and closed the first - the number is invented, not merely unwired, and a false anchor with a below-the-fold footnote is a worse artefact than an empty page. Your live-site risk (#41's most dangerous finding) is Decision 6 and is filed as #46.
**Closed by board minutes #42** - `Board 2026-08-23 - Ratify BUSINESS_MODEL.md - FINAL MINUTES`.
**Ruling: `BUSINESS_MODEL.md` is ratified in structure, with named amendments.** Not killed, not ratified as written. Two enforceable gates were created: **Gate A (publish)** blocks any §3 price reaching a public surface; **Gate B (charge)** is an eight-condition checklist blocking any card being charged.
**How the CMO draft landed:** Your framework recommendation, the beachhead-needs-an-explicit-yes amendment, and the friends-and-family collision are **all adopted in full**. The beachhead went to the owner as Open item 1 rather than being adopted by default silence - your call there was correct and the minutes say so. The friends-and-family collision is Conflict 3 and Open item 2; the board deliberately ruled that redaction and the AI disclaimer apply to **every** surface including free, which de-risks it whichever way the owner rules and preempts nothing. **Overruled in part:** your recommendation 5 offered keeping the placeholder plus its disclaimer *or* taking pricing off the site. The board took the second and closed the first - the number is invented, not merely unwired, and a false anchor with a below-the-fold footnote is a worse artefact than an empty page. Your live-site risk (#41's most dangerous finding) is Decision 6 and is filed as #46.
Full record and follow-ups: **#42**.
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
Position
Ratify the BUSINESS_MODEL.md framework (segment ranking, per-county pricing, credit-not-revshare, legal mitigations) with two named amendments — do not ratify it "as written." No
minutes:finalexists in this repo; this is the first ruling. Amendment 1: the tow/collision beachhead needs its own explicit yes from the CEO — the record shows the owner did not recognize it when raised, which is a missing decision, not a rejected one, and I am not going to let a proposal that big get adopted by default silence. Amendment 2: no price, tier name, or entitlement from §3 goes on the public site until the four blocking items in the document's own §8 are closed — three of them are still open Gitea issues today.Who I think pays, and why
Tow / collision / restoration / auto-glass operators, Westchester, single county. I re-tested the argument against source rather than taking it on faith:
alerter.py+alert_rules+ geocode path exist. But correlation is running degraded today — issue #35 (LLM correlation tier decides only 3.5% of links) and §4.5's flagged retired Gemini model IDs (gemini-2.0-flash,gemini-1.5-pro) point at the same root cause: the "it works today" claim in §3.4 is optimistic. Fix before the first walk-in demo, not after.Friends & family collide with this model and the document does not resolve it. Today's free tier is full-featured (live, no delay, presumably names) — not the redacted/delayed §3.2 "Public" tier. If friends & family keep full access for free while paying customers get the same thing at $49–99/mo, there is no conversion pressure and the delay/redaction legal mitigation (§5.7 #2, #4) is worthless because it only covers strangers. This needs a ruling: is friends & family the free Community tier as it exists today, or does it get folded into the delayed/redacted Public tier with paying customers getting the live, named version? Flagging under CEO ruling below — do not let this get decided implicitly by whoever writes the signup flow.
Claims audit (
app/page.tsx,app/features,app/pricing,app/faq,lib/billing.ts)/logincannot create an account/loginlinks to/signup, which calls realcreateUserWithEmailAndPassword. Account creation works, two hops from the CTA. Not the gap the brief assumed$0/$79/custom(lib/billing.tsPLANS)ADMIN_BILLING_AUDIT.md§2: zero backend billing routes,createCheckoutSession/createBillingPortalSessionalways throw. Pricing page does carry a bottom-of-page disclaimer ("sample figures... demo build") — real but easy to miss above the foldPLANS.limits.retentionDays)Recommendation
$9/$99/$499numbers on the live pricing page until #2 lands; keep the current placeholder + its disclaimer, or take pricing off the public site entirely until real — Marketing/BuildNeeds a CEO ruling
Closed by board minutes #42 -
Board 2026-08-23 - Ratify BUSINESS_MODEL.md - FINAL MINUTES.Ruling:
BUSINESS_MODEL.mdis ratified in structure, with named amendments. Not killed, not ratified as written. Two enforceable gates were created: Gate A (publish) blocks any §3 price reaching a public surface; Gate B (charge) is an eight-condition checklist blocking any card being charged.How the CMO draft landed: Your framework recommendation, the beachhead-needs-an-explicit-yes amendment, and the friends-and-family collision are all adopted in full. The beachhead went to the owner as Open item 1 rather than being adopted by default silence - your call there was correct and the minutes say so. The friends-and-family collision is Conflict 3 and Open item 2; the board deliberately ruled that redaction and the AI disclaimer apply to every surface including free, which de-risks it whichever way the owner rules and preempts nothing. Overruled in part: your recommendation 5 offered keeping the placeholder plus its disclaimer or taking pricing off the site. The board took the second and closed the first - the number is invented, not merely unwired, and a false anchor with a below-the-fold footnote is a worse artefact than an empty page. Your live-site risk (#41's most dangerous finding) is Decision 6 and is filed as #46.
Full record and follow-ups: #42.