When I share with José Andrés the mailbag question that I’m struggling to answer, the first sentence to tumble from his mouth is the same as his second one.
“It’s complex,” says the chef who helped orchestrate the small-plates revolution in America. “It’s complex.”
The question? Why do small plates come out when they are ready and not in the order a diner would like?
In the past three decades, small plates have multiplied like bacteria on room-temperature chicken. They go by different names: tapas (Spanish), antojitos (Mexican), mezze (Mediterranean/Middle Eastern) or torizara (Japanese), just to name a few. But they’re all essentially small plates designed to share with others. They also tend to include a caveat from your server: “The dishes will come out as they are ready.”
Because of their approach, small-plates restaurants often feel like random dish generators. You don’t know what food will hit your table first, or how many dishes will arrive at once. Diners can find this frustrating and chaotic, particularly when compared with the kind of pacing practiced at countless restaurants in the 20th century: you know, the old avuncular, ever-predictable procession of appetizer, entree and dessert.
Given their experiences at these quotidian, three-course restaurants, diners often wonder why small-plate emporiums can’t follow a similar practice and send out dishes in the order they prefer. The problem, I’ve been told by people in the industry, boils down to numbers.
If a kitchen in a busy restaurant will prepare hundreds of dishes a night, a kitchen in a similarly congested small-plates establishment will push out thousands. Ivan Iricanin, the founder and chief executive of Street Guys Hospitality, says diners at Ambar, his chainlet of Balkan restaurants, order an average of 5½ small plates per person. Do the math: A four-top at a standard, three-course restaurant will generally order about 12 dishes. At Ambar, the same four-top will demand 22.
But it’s not just the sheer number of orders that can stress a kitchen — and the servers who must punch in those requests. It’s the randomness of the orders. Since there are no formal courses with small plates, customers are free to build their dinner any way they desire. Let’s go back to our imaginary four-top: One diner may request four dishes just for himself, spread out over a two-hour meal. His tablemates may want to split another 20 dishes during the same period. The time to cook and plate each of those dishes can vary significantly, from a pre-made hummus that takes just a couple of minutes to a chicken plate that requires 12 minutes.
Now multiply this kind of order many times over, one for each table in a tapas-style restaurant. As Werner Herzog might say, madness reigns. Restaurants do what they can to manage the madness. They might group like-minded plates on a menu, suggesting a number of dishes from each category that the kitchen can then cook and plate in a similar amount of time. They might manipulate portion sizes to cut down on the range of cook times among dishes. They might train servers to place orders according to the cook times of each dish: First punch in the chicken, then, a few minutes later, the asparagus, which cooks faster than the bird.
Or they can just have the dishes — all together now — come out as they are ready.
With the latter option, “you’re getting the product in the best way possible,” Iricanin says. “You can enjoy it the way the chef wants you to enjoy it.”
In fact, some diners, as Andrés points out, enjoy the randomness and spontaneity of the small-plates experience. Yet, because this is America — specifically Washington, where the Type A set gathers to throw its collective weight around — Andrés and his team at ThinkFoodGroup wanted to develop a system in which diners have more control. About a year and a half ago, they rolled out a new software system at Zaytinya, the Mediterranean small-plates powerhouse in Penn Quarter.
Zaytinya chef Michael Costa, Andrés and others spent months working on the system. Among other things, they had to program the software with the exact cooking times for each of the dozens of dishes on the menu (information that is updated regularly with new data collected from the kitchen). A server will punch in a table’s entire order but send only a few dishes at a time to the kitchen to prepare — and hold the rest until the diners are ready for the next round. The dish with the longest prep time will take precedence, and once a line cook starts on it, the computer will stagger the start times of the other dishes, so that the food in that round reaches the table simultaneously — and in peak condition.
The software has “overcome what I think are a lot of people’s principal objections to this style of dining,” Costa says. “It doesn’t have to be random anymore.”
The system isn’t foolproof, Andrés cautions, as he shows me how it operates in Zaytinya’s expansive, stainless-steel kitchen, where the heat can feel as heavy as the workload on a Saturday night. All by itself, the computer system “doesn’t mean we’ll always be successful,” says the celebrity chef, “but we aim for that.”
Human beings — stressed ones, distracted ones, fallible ones — still have a role in how your small plates reach the table at Zaytinya. Servers still have to monitor the tables, punching in new orders when diners are ready for them. Food runners still have to pick up the correct dishes and deliver them to the dining room in a timely fashion. Cooks still have to withstand the sometimes crushing demands placed on their stations.
“We still have to manage it,” Costa says. “It doesn’t happen flawlessly every time, but it happens way more often than it used to with the old system.”
ThinkFoodGroup has introduced the proprietary system to other restaurants in its empire, which means that if you want to exercise control over your small plates, your best bet is to dine at Zaytinya, Oyamel, China Chilcano or the downtown Jaleo. Otherwise, be prepared to mouth along with your server as she repeats, for the umpteenth time, that your dishes will arrive . . . well, you know the rest.