A short email, every week
Stars vs. quiet losers
Know who just walked in
vs. venues like yours
A read on your week, before the first coffee.
No dashboard to dig through. Copilot watches the same numbers your POS, payments and stock already capture, then writes you a short email — what drove revenue, what's trending, and the three things genuinely worth your attention. In plain language, the way a sharp GM would say it.
- Revenue, decoded. Not just the number — what moved it, and which day, shift or channel did the work.
- Three things, not thirty. Ranked by what they're worth, with a clear next step — and a button to act on it.
- Ask it anything. Reply to the email — "why was Saturday down?" — and get an answer grounded in your data.
Friday bar drove the lift — covers up 22% after the late-night menu went live.
Margin watch: Truffle Mac & Cheese down to 41% — prawn cost up 12%.
Tue–Wed lunch soft again. Suggested: 15% 2–5pm credit.
The dishes that make you money — and the ones that quietly don't.
Copilot crosses every sale with its recipe cost, so you see what each plate actually earns. Some dishes are stars: popular and profitable. Others sell all night and lose on every cover. It tells you which is which — and what to reprice, re-cost or rework.
High margin, high volume — protect and promote.
Sell well, earn nothing — reprice or re-cost.
Great margin, low orders — push them harder.
Low both ways — candidates to cut.
Are good numbers actually good? Now you can tell.
A 28% food cost means nothing on its own. Copilot anonymously compares you with similar venues nearby — same cuisine, same size, same city — so you know whether your basket, your labour and your margins are ahead of the pack or quietly leaking.
- Like-for-like peers. Matched on cuisine, format, price band and neighbourhood — never named.
- Where you lead, where you lag. Basket size, food cost, labour, turn time — at a glance.
- Fully anonymised. You see the benchmark; nobody sees your books. Aggregated, opt-in, private.
Most loss hides in the ordinary. Copilot catches what doesn't add up.
A comp here, a voided item there, a bill reopened after the guest has left — most loss looks like normal trade. Copilot reads the full event log behind every order, recognises the sequences that don't occur in honest billing, and grades each one by risk — so a quiet pattern surfaces before it becomes a habit.
- Graded by risk. Each flag is scored high, worth-a-check or clear — so you spend your time on the one that matters, not the hundred that don't.
- The whole sequence, not just the total. It follows every void, comp, item move and reopen in order, and shows you exactly why a bill was flagged.
- You're told, not buried. A short alert lands the moment something high-risk appears — the order, the pattern it matches, and the one question to ask. The clean shifts stay quiet.
Bill reopened after payment, cash removed, then the items re-rung on a fresh tab — a known skim signature.
An item was voided seconds before a large cash bill printed — cheap to confirm it was actually served.
Single payment, no comps, no reopens — nothing irregular.
It already knows everything, because it reads everything.
Copilot isn't a separate tool you have to feed. It reads the POS sales, the payment ledger and the stock counts you're already capturing in Papaya — so the day you switch it on, it's working with the full picture. Nothing to import, nothing to wire up.
- One source of truth. POS, payments and stock in a single model — sales, margin and cash always reconcile.
- Better the more you run. Every shift, recipe and receipt makes the read sharper — no setup on your side.
- Open API & MCP access. Builders can query the same insights and pipe them into their own tools and agents. new
Pipe the same insights into your own agents and dashboards.
A GM's instincts, applied to every number you log.
Revenue & trends
- Sales by day, shift & channel
- Average basket & covers
- Week-on-week movement & anomalies
- Soft-shift detection
Margin & menu
- Dish-level contribution & margin
- Stars, gems & quiet losers
- Supplier price-rise alerts
- Reprice & re-cost suggestions
Guests & growth
- Regular arrival & win-back alerts
- Item pairings & upsell ideas
- Neighbourhood benchmarks
- Ask-anything, grounded in your data
Fair things to ask.
It's your data — you should know exactly what it does with it.