Direct-order share is the health metric
Two restaurants can do the same revenue and be in completely different businesses. One takes 90% of delivery through aggregators at 25–30% commission; the other takes 40% direct. The second one keeps materially more of every dinar and owns a list of customers it can reach for free.
Measure it monthly as a percentage and watch the trend. If it is rising, your marketing is building an asset. If revenue is rising while direct share falls, you are growing someone else's platform and renting the growth.
Cost per cover, measured without software
Run one campaign with a code or a phrase said at the table, log redemptions for two weeks, divide spend by count. That gives you cost per new customer — and then compare it to the gross profit on an average ticket, not the ticket itself.
If a new customer costs 3 KD and your average ticket leaves 6 KD of gross profit, you are profitable on visit one and everything after is upside. If it costs 9 KD, the campaign only works if they come back, which makes retention the thing to fix rather than the ad.
Watch the review trend, not the review average
A 4.3 average built over four years hides what happened last month. What matters operationally is the last twenty reviews — because that is what a potential customer reads, and because it is the earliest warning that something in the kitchen or the service has slipped.
Track the rolling average of recent reviews alongside your marketing numbers. A campaign that increases covers while the recent-review average falls is not a success; it is a problem being amplified. Restaurants that watch both together catch this months before it shows up in revenue.