Harnessing unpredictability

It’s spring, when a restaurant’s fancy turns to thoughts of money. This is inevitable after any January slump, when hope has been briefly revived by Saints Valentine and Patrick (who says religion and commerce don’t mix?), and a brief thaw reminds us that it’s only a couple of months to beer on the patio.

Financial angst becomes particularly strong when, as now, the economy su—excuse me; is underperforming. Credit is hard to find, commodity prices are insane and customers are worried—and you, blithe optimist, think that a couple changes to your menu will make it all better. We could save a lot of time here if you would take my advice: buy a tent and some fishhooks and head north. Instead, everyone seems to want to talk about strategy. Fine. Suddenly, fishing’s not a strategy? Have you tried it?

Last month we discussed the virtues of numerical self-analysis; dissecting your menu item by item to get a clear picture of what each one contributes to the business. You did this in your abundant free time and wound up knowing the total production cost of each item on your menu (food plus labor); the weekly sales of each in both units and dollars, the percentage of total sales that each item represents, the percentage of total margin and of total labor for each item, and whether a butterfly waving its wings in China will cause the price of Wisconsin cheese to rise (yes). With all this knowledge, you are prepared to forge confidently into the jungle of your menu and hack out an enclave of fiduciary sanity.

This would work well if your customers were sane. To put it a bit more graciously, there are limits to statistical modeling when unpredictability is built into the system. At Au Bon Pain a few years back, a steak-and-Gorgonzola salad was taken off the menu because numbers didn’t support it. It was put back on when customers screamed for it.

So it’s best to harness that unpredictability by going right to the source. You presumably have an idea of who your customers are (and therefore who they aren’t), and your sales numbers will give you an idea of what stuff on your menu they like. It will not, however, tell you what they like that is not on your menu. So ask them.

If you take a look at the Web sites of a few restaurant chains, you’ll notice that the comment card has undergone some notable evolution. Though many are still mired in the primordial “Whadja like/Whadja dint,” others now ask about the frequency of your visits (overall and by daypart), what new items you’ve tried and how you liked them, your value perception of the restaurant, and—crucially—what other restaurants you eat at, and what you have there. I can’t think of a more direct way of identifying your competition and stealing their thunder.

Of course, having heard all those lectures (ahem) about not diluting your menu identity, you aren’t going to import foie gras for your sports bar’s sandwich-and-burgers section, even though your customers like it when they dine elsewhere. To avoid such awkward marriages, the better-designed questionnaires will phrase their questions to narrow the categories: if you couldn’t eat at our place tonight, where would you go that’s similar, and what would you order? This will occasionally identify something that they can’t get from you right now that could fit your menu, inventory and price point. Then you need to decide what to do with it: every new item needs a business plan; they involve too much work to just throw them into menu and hope that they can swim.

We’ll talk about this in more detail next month, but it hearkens back to the original question about menu change: Why bother? That question and its companion—So what?—need to be asked throughout the process. You like fettuccine Alfredo? So what? You want three new burgers? Why bother? Without answers expressed in clear, convertible currency, it’s best just to go fish.