The lunch customer isn’t the late-night customer, even when it’s the same man

1The scene: one man, two restaurants

Composite scene, not a specific customer

Monday, 1:10 pm. Fahad, in thobe and shemagh, with three colleagues. They have an hour for lunch, a quarter of it in the car. They want a table now, food within minutes, a bill that’s easy to split and parking close by. Nobody wants surprises.

Thursday, 11 pm. The same Fahad, in the family section, with his wife, his mother, his sister and three children. They want a table for eight, privacy, dishes the whole table can share, something the youngest will eat, and nobody rushing them. The bill isn’t his first worry. His worry is getting through the night without embarrassment in front of his mother.

In your loyalty data he’s one customer: male, 34, Riyadh. In the marketing plan he sits inside “men 25 to 40”. Both descriptions are true, and neither tells you what to cook, what to charge or how to lay out the room.

2The image: same thobe, different job

Put the two pictures of Fahad side by side: 1 pm and 11 pm. Same face, same thobe, same phone. The only thing that changed is why he came. At lunch he came to get a work meal done fast and get back to his meeting. At night he came to gather the family and honour his mother.

These are two different customers in one person. Sell to “Fahad” and you get it wrong twice. Sell to “the work lunch” and “the family night” and you get it right twice.

3The idea beneath

Clayton Christensen and his colleagues put it in a line that became famous: people don’t buy products, they “hire” them to get a job done in a particular circumstance.¹ The circumstance explains the purchase better than the buyer’s attributes. Fahad’s age doesn’t change between noon and midnight. His job does.

Ten years earlier, Daniel Yankelovich and David Meer wrote that market segmentation had lost its way.² Segments built on age, income and personality look elegant in a presentation, but they rarely predict what people will buy. They argued that every segmentation should be built around a specific decision you want to make. If it doesn’t help you decide something, it’s decoration.

The evidence is in our calendar. According to Arab News reporting of Saudi Central Bank (SAMA) data, food and beverage spending jumped 74.9% week on week in the week before Ramadan 2025.³ In one Ramadan week, the same source reported restaurant and café spending down 28.3% while groceries rose.⁴ Same people, same ages, same salaries, same neighbourhoods. What changed was the occasion: iftar and suhoor at home, the ghabga instead of dinner out. Segment by age and you’ll never see that shift. Segment by occasion and you’ll see it coming.

And our occasions keep their own hours. A Saudi study of meal-timing habits, based on what participants reported about themselves, pointed to late dinners and skipped breakfasts being common.⁵ In other words, the late night here isn’t an exception. It’s a full occasion with its own demand.

Saudi Arabia also has an old example of occasion-based segmentation we never noticed: the family section and the singles section.⁶ The restaurant split its room not by the customer’s age or income but by one question: who’s with you tonight? That is exactly what an occasion is.

4Where the money leaks

Serve every occasion with one menu, one price and one branch format, and you pay in four places.

The menu. The lunch customer flips through pages of sharing dishes and orders something that takes twenty minutes. The late-night family can’t find a big platter for the table, so they order ten individual plates and the kitchen falls behind.

The price. Lunch is a per-head budget; the family night is value per table. A 20% discount for everyone subsidises a family that was coming on Thursday night anyway, and doesn’t persuade an employee on a fixed budget. (And did the offer bring in sales at all? That was Episode 2’s question.)

The channel. The lunch job may be done better by pre-ordered pickup or delivery to the office, the ghabga job by trays booked two days ahead. One channel for everyone means half the jobs get done badly. And every channel is judged by its contribution after commission, not its revenue.⁷

The branch format. A branch in the office district has lunch as its main job, a branch in a residential area has the late night, a branch on a highway has the drive-up pickup. Copying the flagship everywhere means tables for four where you need eight, and a big dining room where people needed parking. That’s why we treat each branch as its own unit with its own P&L,⁸ and test a neighbourhood’s job through delivery before signing a lease.⁹

5The Sunday-morning question

The question: What job did people hire us for this week?

Three occasion measures

7Credit, and the season finale

As always, the idea isn’t mine. The “job” comes from Clayton Christensen, Taddy Hall, Karen Dillon and David Duncan (2016),¹ and the critique of age-and-income segments from Daniel Yankelovich and David Meer (2006).² My share is the thobe, the shemagh and the family section.

This episode closes the first season of Beneath Marketing. Six ideas, one sentence each:

• Episode 0: Your customer doesn’t fall down a funnel; they circle a roundabout, and their “no” usually means “not now”.
• Episode 1: Price is a comparison in the customer’s head, which is why 29 riyals on the app isn’t 29 riyals in the branch.
• Episode 2: Before you credit a promotion with the sales, ask whether they’d have come without it.
• Episode 3: A campaign your operations can’t absorb sells a closed door.
• Episode 4: A brand is an asset that lowers the cost of being chosen, if the customer recognises you in one second and in their moment.
• Episode 5: Segment by occasion, not age, because lunchtime Fahad isn’t late-night Fahad.

One thread runs through all six: marketing that works beneath the surface is a decision about menu, price, capacity and branch, not just an ad. Season two is coming.