Proof
See it work
Before selling any of this, we ran it on a real business we know well — a neighborhood food business in Wynnewood, PA — and scored it with the same scorecard we'd give you.
The scorecard arc
Before
~4.0 / 10
A good business with a working website — and almost no way for an assistant to find, describe, or act for it.
After signposting
~6.3 / 10
The basics assistants look for, added and verified live. Findable, partly readable — not yet bookable.
After connecting
~8.5 / 10
An assistant could check live availability, hold a claim for the customer, and complete it only after the customer confirmed.
The moment it became real
On a Saturday night test, the owner opened his assistant app and asked about that night's offering. The assistant found the business, read the live availability, and placed a hold. Nothing was final — no message sent, nothing claimed — until he confirmed personally. Then the claim completed with a pickup code, and a follow-up status check answered correctly. Every step asked permission. Every step was recorded.
What we'd tell any owner about this pilot
✓ The improvement was verified by measurement, five separate times, with the same rubric.
✓ The business itself never changed how it operates — same kitchen, same customers, same systems.
✓ Privacy tightened as we went: public surfaces show initials only, and private details never pass through the assistant channel.
✓ Discovery keeps compounding: assistants' indexes take days to weeks to catch up, which is why we re-measure monthly rather than declare victory once.
This scorecard uses our Find · Read · Book rubric and is re-measured over time; it's a measurement of readiness, not a guarantee of rankings or revenue — any studio that guarantees those is selling something else.