Secrets of the Perfect Salsa

Among a certain circle of hipster connoisseurs, Papalote, a little Mexican restaurant on 24th Street in San Francisco’s Mission district, is famous for its salsa—rich, brick-colored, smoky with charred tomatoes, and packing a wallop of heat. The burritos are good, but they’re beside the point. People go to Papalote for the salsa. “Your senses are awakened and stimulated and your soul is transported up and down the majestic mountains and mysterious valleys of our beloved homeland,” says the website. Or, as one reviewer put it, “That salsa is like crack.”

I want the recipe. I call Victor Escobedo, one of the owners, and he tells me that it’s up to his brother Miguel. Miguel says no. The recipe, based on that of their late father, can never leave the bosom of their family. I wheedle, I plead. I claim salsa dependence. I ask him if he will sell it, and he scoffs: “The last time someone asked me that, the price was $300,000—and it’s gone up since then.”

Fine. I will do without his recipe. Instead, I will resort to the time-honored tradition of stealing trade secrets. Not exactly stealing, really; I will reverse-engineer the salsa.

Victor has let slip a clue: The tomatoes are romas. From the black pinpoints of charred tomato skin in the salsa, I deduce they are roasted or grilled. Papalote has craftily disguised the salsa’s remaining ingredients by blending everything to a puree.

Max La Rivière-Hedrick, then-CHOW food editor and tireless salsa Sherlock, suggests we sieve it. When we inspect the resulting particles, we can see fragments of leaf that must be cilantro, as well as other white and green specks. Max decides that, even though Mexicans tend to use white onions in their salsas, these look like scallions. So far, the salsa seems straightforward. But the chiles pose a challenge. Jalapeño or chipotle? Ancho or serrano? Then, as I stare at the tiny specks, I have a revelation: We have to go deeper. We will break down the salsa into the very building blocks of life itself. In short, we will get its DNA analyzed. Today, science has developed to the point where we can identify a murderer from a single hair; surely it will be child’s play to figure out what’s in this salsa.

I persuade a friend to put me in touch with Richard Hamilton, a biotech expert. Hamilton is the CEO of Ceres, a Southern California firm that makes genetically engineered plants. I ask him whether DNA analysis can indeed reveal the salsa’s secret. Yes, Hamilton responds via email, there is a laboratory in the Netherlands called Keygene that might be able to do it—if I have a million dollars. My heart sinks. For that amount, I could probably buy the recipe from Victor and Miguel.

We will break down the salsa into the very building blocks of life itself.

Shattered by the news, I can barely finish reading Hamilton’s email, which explains that there is not enough plant data in GenBank, the database of publicly available genetic sequences, to analyze my salsa. Essentially, if I want its DNA, I first have to pay someone to map the chili genome. I am sorely disappointed by modern technology. They can crossbreed goats with spiders and weave their milk into bulletproof vests, but it costs a million dollars to figure out a simple salsa recipe?

Researching the subject on the Internet, I discover that DNA analysis of food is chiefly used to detect things that shouldn’t be there: specks of nut that might trigger allergies, meat in vegetarian products, pork in kosher products. Strangely, most of the labs that do this sort of thing seem to be on the other side of the Atlantic. Europe, it turns out, is light-years ahead of the States in this field. Partly it’s because European manufacturers are legally obliged to mention on the label if a food contains genetically modified organisms (GMOs), and DNA analysis is used to disclose these. Meat scares have also fueled the food-analysis industry. After mad cow, foot-and-mouth, dioxin in feed, and other panics, European meat producers are using DNA tracking to restore consumer confidence. A DNA test will confirm what cow a steak came from, so that the seller can guarantee that the cow was, for example, organically reared and grass fed. (American meat producers call this “gate-to-plate tracking” and are now beginning to use it, too.)

I track down Genome Express, a food-analysis lab in France, where they are experts in exposing fraudulent foie gras. (Apparently some manufacturers are bulking out the pâté with inexpensive pork or chicken.) But can they figure out the sphinxlike salsa?

After several emails, I manage to explain my mission to Genome Express’s Franck Robert. (Because my French is barely enough to order a café au lait, Monsieur Robert is kind enough to correspond in English.) If I wish to identify a “species” in my salsa, he tells me, they can do it for a song, but identifying vegetables is much harder. You’d think it would be easier to identify a jalapeño than a horse, but the DNA analysis of food is chiefly used to detect the presence of animal substances, and the science of vegetable detection is not very advanced. “In plant,” the “marker” we want to use is “in developpement” at this time, writes Monsieur Robert in his heroic English. “So I make you a proposal: Without any garanty, we test our technology on your product. The cost for you: Zero.” Comme c’est gentil! I promise myself that if he solves the riddle of the relish, I will send him a crate of it.

I have a Pepsi cup of salsa in my refrigerator (last time I visited, Papalote was out of jars). I open it and see that, in his eagerness to assist me in my research, my husband has polished off half. But there is just enough—Monsieur Robert said 3 or 4 ounces will do the trick. I decant the pesky puree into a sterilized jam pot and screw the lid on tight. At the post office, I roll it in bubble wrap, pack it in a box, and dispatch it across the Atlantic. I trust that the scientists at Genome Express will have the self-control not to eat it—at least, not before they’ve analyzed it. I imagine them poring over it as if it held the secret to a crime.

If Monsieur Robert succeeds, the implications are revolutionary. In the future, no recipe will remain mysterious for long. There will be no more secret sauces, closely guarded through generations, and the folks at KFC will be forced to reveal their blend of 11 herbs and spices.

But the days pass, and there is no word from Monsieur Robert. Is it possible the French will fail me? Determined not to give up, I turn to Barb Stuckey of Mattson & Company, which develops new products for the food and beverage industry. Does Mattson ever have to reverse-engineer recipes? And if so, do they have any special techniques? Yes, Stuckey says. In fact, she recently worked on a “matching” project. A company that makes “frozen beef-based entrees” wanted to start a new factory on the West Coast, but it discovered that its East Coast manufacturer owned the rights to the recipes. The company hired Mattson to figure them out.

Stuckey is kind enough to explain, step by step, just how she and her team go about replicating one of the client’s frozen meatloaf dinners. Step one is to “assemble a team of trained palates” (food technicians, chefs, and project managers) and “identify the flavor components.” Step two is to get the “product specifications,” measuring such things as viscosity, pH, moisture level, and sugar content. This is all information your tongue will tell you, but not with the same accuracy and precision.

Seduced by the scientific sound of words like “pH” and “viscosity,” I ask Stuckey how much it would cost to hire her team of trained palates. “Between fifteen and twenty thousand a month,” she says firmly. Considering that the last quote was a million dollars, I think hiring her is pretty cheap—until I remember I don’t have any money. And what is step three? “Trial and error,” Stuckey says. Trial and error? That just means “messing around in the kitchen.” I brighten up. I could do that on my own.

I check in with Monsieur Robert. He asks for more time. But while I have been exploring the frontiers of science, Max has been busy in the kitchen making variations of the salsa and tracking his results in a spreadsheet. He explains how his salsa sleuthing has progressed. First, he knew that there were two chiles because he saw two different specks in his sieve. Looking for a chili with a kick, he tried jalapeños, then habaneros, but finally settled on dried chiles de arbol. He also discerned a fruity, chocolatey base note in the salsa, so he used dried pasilla peppers (often used in mole). Now he is pretty confident that he has the ingredients down.

But should he cook them? Papalote’s salsa has a uniform flavor, which seems to indicate that the ingredients are all cooked together. After charring the tomatoes, Max throws them in a pot with some water and the dried chiles, and then blends everything together. So close. He throws in a pinch of sugar. Closer. But he still isn’t satisfied. His salsa lacks the rich mouthfeel of Papalote’s—a salsa so thick and creamy that some customers have asked Victor whether it contains mayonnaise or peanut butter (Victor denies this).

Max tries thickening his salsa with cornstarch and pectin, but these give it a “glassy quality.” After hours of labor, Max’s eyes are taking on a glassy quality, too. He perseveres. He tries thickening the salsa with cassava and with masa harina (corn flour), sometimes used in salsa, but the cassava makes the sauce watery, and the masa’s flavor is too intrusive. He tries potato (sometimes sneaked into European sauces to make them creamier), but it doesn’t make much difference. He tries adding vegetable oil, but when he leaves the salsa to sit, the oil floats to the top, something that never happens to the vexing Papalote salsa.

Meanwhile, Monsieur Robert is missing in action. But Max is ready. In the end, he finds that toasted pumpkin seeds, one of the ingredients of mole, produce the closest simulacrum of the salsa, and he presents his creation. We dip in a chip and taste. It is absolutely sublime. The tomato flavor is more prominent than that of Papalote’s, and the chili does not overwhelm. While the Papalote flavor hits you all at once, the flavor of Max’s salsa unfolds in the mouth like that of a fine wine. In fact, I cannot stop eating it. Incredible to say, I think it’s actually better than Papalote’s. I imagine the Escobedo brothers begging me for the recipe. Perhaps we will sell it to them for $300,000.