How to increase sales for a restaurant
Most restaurants that want more sales go looking for more covers, and the cheaper money is already sitting at the tables they have. The table that orders two mains and tap water costs you the same chair time, staff and gas as the table that takes the dear bottle, a starter each and dessert, and most of what decides which table you get was settled by the menu before a server opened their mouth. Below is what the research says is doing the deciding, how well each study held up, and one play you can run this week.
The principles at work
| Principle | Study | Held up? |
|---|---|---|
| Anchoring | Tversky & Kahneman (1974). A wheel rigged to stop on 10 or 65 moved guesses of the share of African nations in the UN from 25% to 45%. doi.org/10.1017/cbo9780511809477.002 | Replicated. One of 10 of 13 classic effects that held across 36 samples in Many Labs 1 (Klein et al. 2014). |
| Reciprocity | Strohmetz, Rind, Fisher & Lynn (2002). A sweet with the bill lifted tips; a second sweet with a personal word lifted them 21 to 23%. doi.org/10.1111/j.1559-1816.2002.tb00216.x | Real and published. Not independently re-tested, so treat it as one field study. |
| Social proof | Goldstein, Cialdini & Griskevicius (2008). A sign saying 75% of guests in this room reused their towels took reuse from 37.2% to about 50%. doi.org/10.1086/586910 | Real and published. No independent replication located yet. |
| Music tempo | Milliman (1986). With slower background music in a restaurant, diners stayed longer and bar revenue was 29% higher. doi.org/10.1086/209068 | One study at one restaurant. Not independently re-tested. |
| Choice overload | Iyengar & Lepper (2000). 30% of shoppers who stopped at a stand with 6 jams bought, against 3% at the stand with 24. doi.org/10.1037/0022-3514.79.6.995 | Contested. Scheibehenne, Greifeneder & Todd (2010) pooled 50 experiments and found the average effect close to zero (doi.org/10.1086/651235). |
Carluccio's, the scooter on the menu
In April 2012 the Carluccio's in Manchester's Spinningfields listed a Vespa scooter on its menu, between the food, from £2,400. The restaurant reviewer took it as a joke that raised a smile. Richard Shotton uses it as an anchor pushed as far as it will go. Once a guest's eye has passed £2,400 on the way to the pasta, £16 for a plate of it has stopped looking like a question. Nobody has to buy the scooter. Its job is to be read first.
Try this in a restaurant
You see a price and decide in about a second whether it is fair. Fair compared with what? Your brain cannot judge a number on its own, so it grabs the nearest reference point and measures from there, and it will use a number it knows is meaningless if that is the one that arrived first. Tversky and Kahneman proved it in 1974 with a wheel of fortune rigged to stop on 10 or 65, followed by a question the wheel had nothing to do with, the share of African nations in the UN. People who spun 65 guessed 45% on average and people who spun 10 guessed 25%. A carnival number nobody believed in moved the answers twenty points. Whatever lands in front of a buyer first becomes the ruler for everything after it, and when the Many Labs project re-ran anchoring across 36 samples in 2014, it held.
Your menu is a row of numbers, and the first one a guest reads sets the scale for all the rest. If the dearest thing on the page is a £19 main, your £16 main is the expensive choice. Put a £58 rib-eye for two at the top of the same section and that £16 plate becomes the sensible middle, without you touching its price. The rib-eye does not have to sell. Its job is to be seen before anything else is. Carluccio's did it with a scooter, and you can do it with a cut of meat, a bottle or a sharing board you would be happy to serve.
Steal this: At the table, before the order, say the dear thing first and then the thing you expect them to have. "The chef's cut tonight is the 900 gram rib-eye for two at £58, and if you want something lighter the lamb at £19 is what I'd order." On the menu itself, the dearest dish in each section goes where the eye lands first, top of the list, in the same typeface and size as everything else, so it reads as a real option and not a stunt. One anchor per section is enough.
The mistake: Taking the dear dish off because nobody orders it. It was never there to be ordered. Pull it and every other price on the page gets more expensive overnight, without a single number changing. The second mistake is pricing the anchor at a level that reads as a joke; it has to be a dish you would be glad to cook at a price you would defend.
Try this week: Reorder one section of the menu this week so the dearest dish sits first. Count how many of the second dearest dish you sold in the seven days before and the seven days after. If it moved, do the next section. If it did not, the anchor was not visible enough, or not credible enough, so fix one of those before you give up on it.
More plays for a restaurant
- Play #1, Reciprocity. The sweet with the bill, and the two other moments in a meal where a small unasked-for extra comes back as spend.
- Play #3, Social Proof. How to mark the dishes you want ordered so the menu does the recommending, with the exact wording that works and the wording that does nothing.
- Play #12, Charm Pricing. Which prices on your menu should end in 9 and which should stay round.
- Play #16, Bundle Pricing. Starter, main and a glass at one number, and when that lifts the ticket instead of giving margin away.
- Play #51, Music Tempo Effects. What the speed of the music does to the bar take and the turn time, and how to set it by service.
- Play #39, Choice Overload. How short the menu can go before it costs you, and how to test it for one week without upsetting the regulars.
- Play #33, Peak-End Rule. The last two minutes of the meal are the review. What goes there.
- Play #30, The Goal Gradient Effect. Why the loyalty card should never start empty, and how many stamps to put on it before you hand it over.