Back by night ten

For years I read Ramadan off one number, and it was the wrong one.

Revenue for the month against last Ramadan. Every operator in the region can produce it, and it moves for reasons that have little to do with anything anybody did.

The number I look at now is smaller, and it is available today, on about the eleventh day. Of the customers who came on the first three nights, how many have come back by the tenth.

Nobody takes a risk on a new kitchen on the twentieth night of this month. A late order at sunset does not spoil a meal, it spoils an evening for nine people who were all waiting on the same minute, and a household that has been let down once does not offer the month a second chance. So the customers arrive early and then stop arriving. What is left after the first week has nothing to do with acquisition. It is thirty nights of the same families deciding again, every evening, whether to use you.

Which means a good first night is not a result. It is an option, and it expires around the tenth.

The uncomfortable part is that the revenue line cannot see any of this while it is happening. The families who stayed order more as the month goes on. They order for bigger tables, they add the dessert, they order on nights they had meant to cook, and they cover the gap left by everyone who tried you once and went elsewhere. A month can read as growth the whole way through and finish on a smaller base than it started with.

A Ramadan customer who has not come back by night ten is not coming back this month. That is knowable tonight. The revenue will not admit it until after Eid.