With one outlet, an owner can feel when something's off — the vibe on the floor, the state of the walk-in, the look on the cashier's face. With three or ten, feel doesn't scale. What replaces it is a short list of numbers looked at the same way every week. Here's ours.
The weekly eight
- Sales per outlet vs same week last month/year. Trend, not absolute. A 12% drop at one branch while others hold is a management conversation, not a market one.
- Average order value. Falling AOV usually means upsell has stopped or the menu mix has drifted to cheaper items. Compare outlets; the gap is your training opportunity.
- Food cost % from recipe-level inventory. Theoretical (what the recipes say you used) versus actual (what left the store). A widening gap is leakage — over-portioning, wastage, or worse. This number pays for the POS.
- Top and bottom ten items per outlet. Menu engineering starts here: promote the high-margin winners, fix or drop the losers, and notice when an outlet's list looks unlike the others.
- Discounts and voids by staff member. Legitimate service recovery is normal; a pattern on one login is not.
- Sales by hour and day. Staffing against actual demand rather than habit is the fastest labour saving there is.
- Online vs dine-in vs aggregator mix — and the effective commission you paid. Watch the aggregator share; regulars should be migrating to your own channel.
- Repeat-customer rate on your own channel. The health of the customer base you actually own.
Making it a habit
Same day each week, same order, fifteen minutes. Consolidated view first for the trend, then per outlet for the outliers, then one action per outlier. If your POS can't give you these eight without exporting to a spreadsheet, that's the ninth number: how much of your time the system costs you.
Built for owners with more than one till
Central menus with local overrides, outlet-wise and consolidated dashboards, recipe-level inventory and staff-level audit trails are the multi-outlet core of our Restaurant POS. Ask for the multi-outlet demo and we'll walk through these eight on real data.

