What I Learned Running Marketing at a Startup That Shut Down

Before my current job, I spent nineteen months as head of marketing at a meal-kit startup for families with food allergies. We raised a seed round of $2.4 million, built a loyal small audience, and shut down on a Wednesday in November. The email from our CEO was two paragraphs long. I still have it.

I was twenty-nine, and I took it personally for a long time. Now, five years later, I think it was the most useful job I have had. Here is what I wish someone had told me while it was happening.

Love is not a market size

Our customers adored us. I mean that literally: we had parents sending handwritten cards about the first time their kid with a peanut allergy ate a takeout-style meal without a panic attack. Our net promoter score was in the high seventies. I used those cards in every pitch deck.

But there were simply not enough of those families within our delivery area who could pay $14 a serving every week. Our total addressable market, when we finally did the honest math, was a few thousand households across three states. Loving customers made it very easy to ignore how few of them there were.

Marketing cannot fix unit economics

For most of my time there, I believed that if I could just get our customer acquisition cost down a little more, everything would work. I optimized ads endlessly. I got cost per new customer from $140 to $95, which felt like a huge win.

What I did not fully understand was that each box cost us around $11 to pack and ship, refrigerated, to a small number of scattered addresses. We lost money on every order for the first four months of a customer's life. No amount of clever marketing changes that. I was polishing the front door of a house with no floor.

Watch the founders, not the dashboard

In hindsight the signs were there about five months before the end. Our CEO stopped coming to the Friday all-hands. Board meetings moved from the office to a coffee shop. New hires were quietly paused. Nobody announced anything, but everything changed shape.

I was too busy looking at campaign metrics to notice. If I had paid attention, I could have prepared my team better, and maybe helped two of them find jobs before the announcement rather than after.

What I took with me

Three things stayed with me and shape how I work now.

• I learn the unit economics in my first month. Before I plan a single campaign, I want to know what a customer costs to serve.
• I ask directly about runway. Founders usually respect the question.
• I treat customer love as a signal to investigate, not a conclusion. It tells you something is working. It does not tell you that enough people want it.

The handwritten cards are in a shoebox in my closet. I read them occasionally. They remind me that we did something real for a small group of people, and that real is not the same as viable. Both of those things can be true, and holding them at the same time is most of what I learned.