The few bets Brew is making, what would prove each right or wrong, and the order we test them in; the sources give the wedge and the ladder but never name the bets, so the list below is a proposal for Param.
Today
What the sources already commit to
Commitment
Source
The wedge: read the store before drafting anything, with a source on every number
brand-directions/book/content/b.js b-11 edge
The shelf: most tools either read the store or ship the work; we read, decide, draft, then hand off
Five real design partner prospects on production since 25 Oct
memory brew-live-storesmemory brew-prod-backend
Keep every Play; make each cheaper
memory keep-all-moves
Global product, acting on the brand's own stack
memory brew-name-and-scope
The moat grows with use: what we know about a store, then what worked
brand-directions/book/content/b.js b-11 moat
The tension the bets must settle
The pricing proposal estimates v1 brain cost at about $15 a brand a month against a $79 plan brew working tree docs/pricing.md:20-23
The caps in the code allow far more: the cost per brand page proposes up to about $180 a month before drafts, not measured content/ai-and-models/cost-per-brand.md Numbers
No brand has a measured monthly cost yet content/ai-and-models/cost-per-brand.md Open
Tomorrow
The proposed bets, in the order we test them.
#
Bet
Right if
Wrong if
Version
Tag
1
Reading first beats a blank box
Partners heart and hand off Plays from their own evidence most weeks
Not for me outnumbers hearts across partners for a month
v1
New
2
A weekly hand off habit forms before any posting
Three of five partners hand off in three weeks running
No partner hands off two weeks running by the end of the partnership
v1
New
3
One brand costs well under its plan
A measured month per brand sits under a third of the plan price
Measured cost sits above half the plan price after the cheapest models that keep quality
v1
New
4
Trust earned in v1 carries into posting
Partners say yes to the first scheduled posts and keep the dial up
Partners turn posting off after trying it
v2
New
5
Results make next week smarter
Graded results change the ranking and hand offs rise
Results arrive too late or too noisy to change a pick
v3
New
How we do it
Test one bet at a time, in order; bet 3 runs beside bets 1 and 2 because it needs only logs.
Each bet has one owner and one review date.
When a bet is proven or broken, write it in the decisions log with the evidence and update this page.
A broken bet changes the plan, never the truth rules.
Numbers
Metric
Bet
Where measured
Status
Heart to Not for me ratio per partner a week
1
Card reactions
Proposed
Weeks in a row with a hand off, per partner
2
Brief records
Proposed
Measured cost per brand a month
3
Spend files plus AI tags per brand
Proposed, not measured
Share of scheduled posts approved
4
Need your yes log, v2
Proposed
Open
Are these the bets? Proposal: accept bets 1 to 3 for the design partnership and keep 4 and 5 as the later order.
Who owns each bet? Proposal: Param owns 1 and 2, Sai owns 3 with the cost per brand work, Sai reviews 2 with partner calls, since team and roles proposes Sai for design partner relationships.
Review date for bets 1 to 3. Proposal: the end of the six week partner programme the book proposes, or 7 Dec 2026, six weeks from today, if no programme date is set.
