Current Issue
This Month's Print Issue

Follow Fast Company

We’ll come to you.

4 minute read

Mark Bittman

Yikes—Does My Startup Need To Become A Tech Company, After All?

Mark Bittman has always thought of Purple Carrot as a food startup—but now he wonders if he needs to think of it as technology company, too.

[Photo: Flickr user Didier Weemaels]

From late spring to early summer, during the first days of my involvement with meal kit startup Purple Carrot, we fielded many questions about our "tech side." These came primarily from a few influential Silicon Valley VCs who would directly ask questions like, "How are you using technology to make your startup more defensible?"

Being a newcomer to the company—and a startup world outsider to boot—I didn’t feel comfortable asking (perhaps as a challenge to their question) who in the world of food startups is really using technology innovatively or explaining that when I think of truly innovative technology, I think of 20th century revolutions like antibiotics or computers.

I suppose it depends how you define "technology," but my usual reply was that Purple Carrot is a food company. This doesn’t mean that we don’t use technology, obviously, but using technology doesn’t make you a tech company any more than publishing a hard copy of a magazine makes Fast Company a printing company. Nevertheless, the question confounded us until some equally savvy and influential people reassured us that we did not need to (nor were we likely to) invent anything technological.

That was a relief. But although I still spend most of my time at work thinking about food and other very down-to-earth things, our marketing strategy is making me wish I were a software developer—or at least could find the right ones.

Right now, our food is an extension of my taste and that of a couple key members of my team. We develop recipes based on our experience, and the meals Purple Carrot ships are essentially what we like, think is appealing, and guess our customers will appreciate and find not too challenging to make. I can see, though, that the day is fast approaching when we’ll be spending a good deal of time, energy, and money on the technological aspects of marketing, at least in the data-collection and identity-defining aspects of that ill-defined term.

Sales appears to be something we can manage with existing techniques and strategies; the main challenge of customer acquisition is lowering the cost of our product without sacrificing our quality or our values. But in almost every business that’s even remotely akin to ours—from supermarkets to magazines to online retailers—it’s customer retention that really matters, and for that we’re going to need to know our not only our current but our future customers.

This will take sophisticated data collection and analysis, and I’m not sure anyone has packaged what we actually need here (email me if I’m wrong please), because the notion of a mission-driven, for-profit meal kit company is a relatively new one. (Even if you allow that the mission can be defined as simply as "get people to cook more at home," the notion is only a few years old.) We’ll know if we’re connecting with our customers, of course, by analyzing user churn and our staying power. But we won’t know whether we’re maximizing those relationships, marketing to the right new customers, and generally being the best company we can be for them without making a series of data-driven judgments about how to best make those connections.

One can imagine a complex yet elegant tech solution to our prayers, a system that would allow us to ask individual customers exactly what they want to eat, and send them that. But aside from the ridiculously impractical nature of that "solution"—imagine a restaurant in which you could order pretty much anything, or even choose from 100 entrées—it’s not clear to me that people actually know what they want. In fact, a good deal of my job is to define that for them. Just as you go to a restaurant because the chef’s judgment appeals to you, you come to Purple Carrot because the product we’re making draws you to us.

No. The answer, it seems to me, lies in determining for people they people want, and giving it to them. I’ve written before that as a mission-based company, we’re looking to build a community; no one is going to come to us who doesn’t believe that eating more plant-based food is in their interests, and few will stay with us who don’t believe that the qualities that go into producing that food are as important as the ingredients themselves. This is what piqued my interest in Purple Carrot in the first place.

Of course, no one will stay with us if our recipes aren’t consistently appealing, interesting, revealing, satisfying, and worth the time (and money) that goes into preparing them. I can make sure our meals taste good; that part isn’t difficult. But those remaining qualities that will help build and retain our users are more challenging to determine—they probably include our agro-ecological values, but might also include the level of difficulty or diversity or constancy in our meals, or how we fit in with our customers's schedules and lifestyle rhythms, or other things we haven't realized yet.

I’m sure that if our data collection and reporting get stronger, it will become easier to see whether we’re hitting the mark with existing customers. On one level, they’re pretty easy to understand: If they stay, they’re happy; if they don’t, they weren’t. But to win them back or retain future customers, we need to understand what made them unhappy. This isn’t easy; the more complex data we gather about them—the most valuable data—is extremely noisy, at least to the naked eye: It may tell us what they think, but it doesn’t tell us what we should do, or how we should alter our recipes or our service.

What we need to do is figure out the right balance of variety, familiarity, challenge, ease, and novelty in our recipes that will allow us to keep as many existing customers as we can while also finding new ones. And that may take a kind of data analysis that doesn’t yet exist (we’ve been looking at the existing options out there). So stay tuned: We may become a tech company yet.

loading