Answer
The story we tell about the store owner's problem and why it can be solved now: a small store's founder is its whole marketing team, the answers already sit in the store's own reviews, pages and ads, and AI can now read all of that in minutes and hand back a drafted week.
Today
The problem, in the founder's week
| Beat | What is true | Source |
|---|---|---|
| One person, every job | The founder sells, packs, answers customers and runs marketing, so marketing gets the tired hours | brand-directions/book/content/b.js b-1 monday |
| Five questions, no specialist | What should I post, what ad next, why visitors do not buy, does ChatGPT recommend me, how do I get customers back | brand-directions/BRIEF.md section 3.3 |
| The facts are already there | Reviews hold customers' words, long running rival ads hint at what sells, AI answers show if the store is named | brand-directions/book/content/b.js b-1 already-there |
| Every idea starts blank | Doing it yourself, a chat assistant and design apps all start from an empty box unless someone pastes the facts in | brand-directions/book/content/b.js b-5 alternatives |
| Hand offs are messy | A brief to a designer is a voice note and three screenshots | brand-directions/book/content/b.js b-1 monday |
The pains above come from marketing books written as Samples, never from interviews yet brand-directions/book/content/b.js b-1 monday.
What changed, as product facts we can show
| Shift | What it lets Brew do today | Source |
|---|---|---|
| AI reads a whole store fast | A read across the store, catalog, reviews, ads, SEO, AI answers, socials and rivals, live on screen | brew/SCOPE.md:11 |
| The evidence is wide | The source inventory lists 68 sources, 67 live today, 7 of them once an account connects | brand-directions/docs2/content/inventory.mjs SOURCES |
| AI writes real plans | 65 kinds of Plays, each with why, what we found, the plan and what to watch | brand-directions/docs2/data/plays.json |
| Drafting is real | Most cards carry a first draft: image, video, text or a mix | brew/docs/brew-loop.md:11 |
| AI search is a new shelf | The read checks whether ChatGPT, Gemini, Perplexity and Claude name the brand | brand-directions/BRIEF.md section 2 |
| Read only connections are normal | Shopify, Meta Ads, Google Analytics, Google Ads, Klaviyo and Judge.me read only on the live site since 16 Oct | brew/SCOPE.md:14 memory brew-real-connectors |
| Model cost fell | Plays run on Gemini 3.1 Pro and everything else on Gemini 3.5 Flash, paid by Google startup credits | brew/tools/prod-env.sh:9 memory brew-gemini-provider |
One true example to tell
- On 16 Oct the demo guide recorded Arata as named in 12 of 32 AI answers, and 0 of 20 when the shopper does not name the brand brew/docs/handover/product/08-demo-guide.md:14-16
- Shareable copy may use it only as an example finding from one brand's read brand-directions/BRIEF.md section 2
Tomorrow
| Item | Version | Tag |
|---|---|---|
| The story beats: get found, get chosen, get them back | v1 | Stays brand-directions/BRIEF.md section 3.3 |
| Replace every Sample pain with real founder words, with permission, from partner calls | v1 | Changes |
| The why now line grows: posting with a yes | v2 | New |
| The why now line grows: seeing what worked | v3 | New |
| The why now line grows: ads inside rules | v4 | New |
How we do it
- Tell it in three beats: the founder's Monday, what it does today, one concrete finding from the listener's own store.
- Never open with AI, agents or the stack. Open with the store.
- Any claim about how many shoppers use AI to find products needs a named source first; we have none yet brand-directions/book/content/b.js b-8 why-now.
- Invented examples carry a visible Sample label. The sample candle store is the only invented store.
- In shared copy, say Plays and leave the name out.
Numbers
| Metric | Definition | Where measured | Status |
|---|---|---|---|
| Real founder quotes on file | Quotes from partner calls with written permission | Partner feedback log | Proposed, 0 today |
| Read time | Median minutes from paste to a finished read on production | Read logs | Proposed; 3 to 4 minutes is the allowed figure, not yet re-timed |
Open
- Which finding leads the story? Proposal: the AI answers gap, since it is the newest shelf and the easiest to check live on a call.
- Whose words replace the Samples? Proposal: ask each of the five partners for one sentence about their Monday on the next call, with written permission to quote.