How we test multi-unit readiness

The first site usually works because the owner is there. They catch the under-portioned plate, they know the regular who needs the corner table, they re-route the dishwasher when the prep cook calls out. The second site cannot rely on any of that.

Before we model a second site, we test the first. It needs to run for two full weeks without the owner on the floor. If it cannot, the systems are not ready and the second site will pull the first one down with it.

We also cost the first site as if the owner were a paid employee. The first site's margin often includes unpaid owner hours that never appear in the P&L. When those hours have to be replaced by a salaried GM at site two, the unit economics look different. Standardize before scaling — recipes, prep sheets, ordering guides, opening and closing checklists, and the weekly scorecard should all exist as documents before site two is on the table.