What Weekly Analytics Digests Should Actually Tell an Owner

A lot of weekly restaurant reports are, functionally, a restatement of things the owner already sensed from being on the floor: sales were up, Saturday was busy, the new dish is doing fine. That kind of digest isn't useless exactly, but it's not earning its place in your inbox either, since it isn't telling you anything you'd have missed by paying reasonable attention to your own restaurant.

What separates a useful digest from a restated one

A genuinely useful weekly digest tells you something you couldn't have known just from being present in the restaurant: which specific dish's satisfaction has been slipping for three weeks running, which item that used to be a reliable seller has quietly dropped off, or which new addition is already earning a like rate strong enough to justify a bigger push. Those are things that live in the aggregate data, invisible from any single night on the floor, however sharp your instincts are.

The specific things worth including

• Any dish that moved into or out of your top performers this week, not just the current leaderboard
• Dishes showing a meaningful shift in like rate, up or down, compared to the prior period
• New dishes and how they're trending in their first few weeks on the menu
• Anything flagged for an allergen conflict or a recipe issue that needs attention before it becomes a pattern

Why change over time matters more than a snapshot

A single week's numbers, viewed in isolation, rarely tell you much. Every restaurant has slower and busier weeks for reasons that have nothing to do with the menu, weather, a local event, a holiday. What actually matters is the trend: is a dish's satisfaction consistently drifting in one direction over several weeks, not just did it have one unusually good or bad week. A digest built around week-over-week change, rather than a flat snapshot, filters out most of that noise and leaves you with signals actually worth acting on.

It also helps if a digest tells you what to do next, not just what happened. A report that says like rate dropped for a dish is a fact. A report that says like rate dropped and the drop is concentrated among guests who flagged it as too spicy is a fact you can immediately act on.

What a digest looks like when it's done well

A well-built digest reads less like a report and more like a short memo from someone who already did the analysis for you. Instead of a chart of weekly revenue, it says a specific dish's like rate dropped six points over two weeks and the drop is concentrated among guests who also flagged the portion as small. Instead of a list of every dish sold, it says three new items are trending well in their first weeks and one is trending poorly enough to warrant a second look before the next batch of guests tries it. That specificity is what makes a digest worth opening every week instead of archiving unread.

How often is actually useful

Weekly tends to be the right cadence for most restaurants, frequent enough to catch a problem before it compounds, infrequent enough that there's always something meaningful to report rather than noise from a single day. A daily digest usually contains too little signal to be worth a daily read, and a monthly one brings back the exact delay problem a good digest is supposed to solve in the first place. Somewhere between those two extremes, weekly ends up being the cadence that actually gets read, week after week, rather than skimmed once and then quietly ignored.

TasteGraph's weekly digest is built around exactly this distinction, surfacing dishes that moved meaningfully, up or down, and sorting them straight into Promote, Review spice or recipe, or Flag as at risk, so what lands in your inbox is a short list of decisions to make, not just a restatement of numbers you already had a rough sense of.