{
 "id": "1.10",
 "title": "Risks",
 "department": "Company and strategy",
 "department_slug": "company-and-strategy",
 "direction": null,
 "branded": false,
 "status": "draft",
 "owner": "Param",
 "last_verified": "Mon Oct 26",
 "holds": "Product, platform, AI, legal, money and people risks, with a mitigation each",
 "urls": {
  "html": "https://handbook.joinsquareup.com/company-and-strategy/risks",
  "markdown": "https://handbook.joinsquareup.com/company-and-strategy/risks.md",
  "json": "https://handbook.joinsquareup.com/company-and-strategy/risks.json"
 },
 "other_directions": [],
 "tldr": {
  "headline": "Six risk areas, each with a sign to watch",
  "points": [
   {
    "icon": "warn",
    "text": "Meta business verification is stalled, and Shopify app review is not approved yet."
   },
   {
    "icon": "money",
    "text": "Cost per brand above a third of $79 would break the plan."
   },
   {
    "icon": "spark",
    "text": "An AI key ran dry on 20 Oct; one env switch swaps us to Gemini."
   },
   {
    "icon": "lock",
    "text": "Only preview terms and privacy exist; a partner agreement comes before graduation."
   },
   {
    "icon": "people",
    "text": "A team of three: the handbook and Next carry every plan."
   }
  ],
  "stat": {
   "value": "3",
   "label": "platform reviews pending"
  },
  "decide": "Do we write the partner agreement now? Proposal: yes, before any partner graduates to paid.",
  "mood": "urgent"
 },
 "sources": [
  "brand-directions/book/content/b.js b-11 risks",
  "brew origin/main a559de6: docs/connectors-approvals.md:26, legal-pages.js:16-40, tools/prod-env.sh:9, SCOPE.md:13",
  "brew working tree docs/pricing.md:13, 23",
  "handbook content/ai-and-models/cost-per-brand.md, content/product/design-partner-program.md",
  "memory: brew-meta-app-review, brew-shopify-app-review, brew-google-oauth-scopes, brew-pro-switch-and-flush, brew-gemini-provider, brew-connect-refresh, brew-move-freshness, brew-higgsfield-wiring, brew-real-connectors, brew-pricing-parked, brew-live-stores, never-disconnect-arata-meta, brew-sloosh-finals"
 ],
 "open_questions": [
  "**Who owns the monthly review?** Proposal: Param, with Sai on money and partners and Arvind on platforms and integrations, as in [team and roles](/people-and-ways-of-working/team-and-roles).",
  "**Do we write the partner agreement now?** Proposal: yes, before any partner graduates to a paid plan."
 ],
 "parts": {
  "Answer": "The risks that could hurt Brew, in product, platforms, AI, legal, money and people, each with the sign we watch and what we do about it.",
  "Today": "### Product\n\n| Risk | Sign | Mitigation | Source |\n|---|---|---|---|\n| Ideas feel generic | Many Not for me, few hearts | Lean on the store's own evidence; proof on every card | [src: brand-directions/book/content/b.js b-11 risks] [src: brew/SCOPE.md:13] |\n| A matcher flags the wrong Plays | A short product name matched a variant and flagged 54 of Arata's Plays on 16 Oct | Check matchers on real brand data with variants before deploying | [src: memory brew-move-freshness] |\n| Background work runs away | Cozy's Plays were rewritten about 116 times in 3 hours on 16 Oct | Settle waits, once per set of accounts, caps per run | [src: memory brew-connect-refresh] |\n| People expect posting | Support asks where the post went | Say plainly that v1 never posts | [src: brand-directions/book/content/b.js b-11 risks] |\n\n### Platforms\n\n| Risk | Sign | Mitigation | Source |\n|---|---|---|---|\n| Meta business verification stalls | In review since 16 Oct, still on 24 Oct | Keep the read only Ad Library path; App Review asks only for ads_read, business_management, public_profile and the Marketing API tier | [src: memory brew-meta-app-review] [src: memory brew-real-connectors] |\n| Shopify app review | Submitted 21 Oct, not approved | Own app OAuth for partners until approval | [src: memory brew-shopify-app-review] |\n| New Google scopes restart review | Any scope beyond adwords and GA4 | Never add scopes without a plan for review | [src: memory brew-google-oauth-scopes] |\n| Arata's connections break | A disconnect or reconnect | `PROTECTED_CONNECTIONS` blocks it in code | [src: memory never-disconnect-arata-meta] [src: brew/tools/prod-env.sh:9] |\n\n### AI and models\n\n| Risk | Sign | Mitigation | Source |\n|---|---|---|---|\n| A provider runs dry | The Anthropic key ran out of credit on 20 Oct at 07:50 IST | One switch to Gemini through env vars, no code change | [src: memory brew-pro-switch-and-flush] |\n| A preview model changes or goes | Plays run on gemini-3.1-pro-preview | Probe models before deploys; keep a fallback model set | [src: brew/tools/prod-env.sh:9] [src: memory brew-gemini-provider] |\n| Quality differs by model | Opus Plays read clearly better than Pro on Arata, confounded by inputs | A fair rerun before a client launch | [src: memory brew-model-compare-2026-10-20] |\n| A maker's terms do not fit | Higgsfield terms make a multi tenant app on one CLI login risky | Ask Higgsfield sales about Enterprise first | [src: memory brew-higgsfield-wiring] |\n\n### Legal\n\n| Risk | Sign | Mitigation | Source |\n|---|---|---|---|\n| Only preview terms and privacy exist | A partner asks for a contract | Write a partner agreement before graduation | [src: brew/legal-pages.js:16-40] [src: content/product/design-partner-program.md Tomorrow] |\n| Tax on wallet top ups unclear | GST on top ups unconfirmed | The CA confirms before billing | [src: brew working tree docs/pricing.md:13] |\n| The entity named on platforms | Yellow Monk Technologies LLP is on the Meta business | Keep one entity across platforms, invoices and contracts | [src: brew/docs/connectors-approvals.md:26] |\n\n### Money\n\n| Risk | Sign | Mitigation | Source |\n|---|---|---|---|\n| Cost per brand above the plan | Measured month over a third of $79 | Make each Play cheaper, never fewer; raise caps only with a reason | [src: brew working tree docs/pricing.md:23] [src: content/ai-and-models/cost-per-brand.md Numbers] |\n| Credits run out | $20,763 of $25,000 Google credit left on 16 Oct, ends 2028-05-19 | Track the balance monthly in Finance | [src: memory brew-gemini-provider] |\n| No revenue while parked | Pricing parked, free during the preview | Set prices after a measured month | [src: memory brew-pricing-parked] [src: brew/legal-pages.js:40] |\n| Paid makers bill per use | Sloosh finals cost credits per piece | Every paid make shows its price and waits for a yes | [src: memory brew-sloosh-finals] |\n\n### People\n\n| Risk | Sign | Mitigation | Source |\n|---|---|---|---|\n| A team of three | One person holds a whole front | The handbook and Next carry every plan | [src: handbook AGENTS.md] |\n| Parallel sessions overwrite work | A rebase dropped another session's lines | Stage only your lines; deploy from a clean worktree | [src: memory brew-move-freshness] |\n| Five partners carry the learning | A partner goes quiet | Never wipe them; check fixes on them first | [src: memory brew-live-stores] |",
  "Tomorrow": "| Item | Version | Tag |\n|---|---|---|\n| Posting needs write scopes and a fresh platform review | v2 | New risk |\n| Results can be wrong or late | v3 | New risk |\n| Money spent by Brew inside rules, with undo for spend already gone | v4 | New risk |\n| Every risk above | v1 | Stays, reviewed monthly |",
  "How we do it": "1. Review this page on the first Monday of each month.\n2. A risk that fires gets a line in the decisions log with what we did.\n3. A new risk needs a sign we can watch and one mitigation, or it is a worry, not a risk.",
  "Numbers": "| Metric | Definition | Where measured | Status |\n|---|---|---|---|\n| Risks with a fired sign this month | Count | This page | Proposed |\n| Platform reviews pending | Meta, Shopify, Google Ads Basic access and any new scope | Connectors approvals doc | 3 on 26 Oct: Shopify, Meta verification, Google Ads Basic [src: memory brew-meta-app-review] [src: memory brew-real-connectors] |",
  "Open": "1. **Who owns the monthly review?** Proposal: Param, with Sai on money and partners and Arvind on platforms and integrations, as in [team and roles](/people-and-ways-of-working/team-and-roles).\n2. **Do we write the partner agreement now?** Proposal: yes, before any partner graduates to a paid plan."
 }
}