I'm Writing a Cookbook. Now What?

Note: Mexican food expert and journalist Lesley Tellez blogs about Mexican cuisine at The Mija Chronicles and runs Eat Mexico, a food tour company in Mexico City. Now she's writing a cookbook, her very first. Here's how—and what the publishing people don't tell you goes into the job.

When I signed my cookbook contract late last year, to write about the foods of Mexico City, I was excited and overwhelmingly grateful that someone was actually paying me to write a book.

Then the excitement petered out and was replaced by terror. I had to write this book. I was contractually obligated to create 115 recipes in nine months—that meant about three fully developed recipes a week, assuming I didn't get sick or go on vacation. And I didn't know exactly what was involved, because I had never written a cookbook before. I'd written lots of recipes for my blog, but that was different—these recipes would live forever in a real, tangible document. (Or at least I hoped they'd live forever.)

I swallowed my terror and talked to other cookbook authors, and attacked the project like a new job. I assembled a team: a recipe tester to cook the recipes I developed, and a kitchen assistant—really a friend—who'd visit a few times a week and help me work faster. (It helped immensely, I realized later, that both were positive, cheery people who could bring me out of the isolation of working alone and dealing with recipe failure after recipe failure.) I paid my recipe tester out of my advance, which is also how I'd pay for groceries, supplies, and my research trips. My friend I couldn't afford to pay, but she accepted my promise to ply her with food and an acknowledgement in the book.

I created a workflow. I wrote up all my recipes in a large spreadsheet with different tabs for each chapter, and tracked their progress with colored highlights. Green meant the recipe was ready to be sent to the tester. Yellow meant it still needed work. Another spreadsheet contained my weekly recipe schedule, and in another, I tracked my expenses.

The actual heart of the dishes lived in my head, though, and in the memories of living and eating in Mexico City for four years. What did the purslane in salsa verde taste like all those times, in the various fondas where I'd tried it? What was that spice mixed with the pickled white onions I'd eaten at all those taco de canasta stands? What was it really like to bite into a taco al pastor, sliced fresh off the roasting spit?

I would cobble together a recipe based on what I thought should happen, what the vendors themselves did, and what other respected Mexican cooks recommended. (Although that latter part was difficult—lots of Mexican cooks swear that there's only one right way to cook something, and if you cannot cook it that way, you should stay out of the kitchen.) I'd print out the recipe draft, write the date I was cooking, and execute it.

Recipe testing for a cookbook isn't like making lazy Saturday breakfast. The ingredients need to be weighed and occasionally measured with a ruler. The process and cooking times, noted exactly, for adjusting the recipe afterward. And when the recipe doesn't come out right—is it missing some salt or acid or...something else?—it has to get made again and again. With all of that, plus regular food (the small number of dishes I eat that have nothing to do with the book, which includes Thai takeout), eventually you run out of space in your fridge and freezer. Today, any dish that doesn't make the first cut ends up in the trash. The stuff that's book-worthy, I send with my husband to work.

Generally I'm actively testing about three recipes at a time. Once those are done, I move on to the next three. My average right now is about five recipes per week, but it's no doubt going to get busier as the manuscript deadline approaches. All the grocery shopping, I do two or three times a week—I'm lucky enough to live in Queens, not too far from the Mexican bodegas in Corona.

After updating the recipe and settling on a final version, I wrote a headnote that explained what the dish tasted like and why it was important. Then I placed the completed recipe in a new file for my tester. She'd cook the dish at her house and make comments. I'd tweak the recipe. Repeat, repeat, repeat.

I've been writing Eat Mexico for six months now, and I've realized that while writing recipes is a science, it's based on creative variables—smells and memories and flavors, and a gut feeling about whether you'd gotten it right. Like writing, you've got to trust that you've chosen a compelling way to tell the story, and that you're not being careless with readers' time.

I really won't know until the book has been published what people will think of it. For now I'm keeping my head down, and washing my kitchen towels as fast as I can.