Dogfooding

Here’s a big mistake we made at my previous company: we never really dogfooded.

Dogfooding, from the phrase “eat your own dogfood,” simply means to “make sure you’re using your own products.”

It’s obviously “important”, a best practice even. After all, who has better feedback about a product than the people making it?

In real life, you get two problems: first, how do you rank your own priorities from dogfooding compared to the stuff actual customers want? And second, what if your product doesn’t quite fit your use case? Do you change it to fit better?

In other words: how important of a customer are you, for your own product?

Back in the day, we were under severe pressure to cater to customers. We were scrambling to close every deal we possibly could, so we could survive. Our product was still in its infancy, and every customer found gaps. Our roadmap filled up with commitments and critical issues to solve. Dogfooding came in as a very low priority. Yes, our product didn’t quite work for us, but it didn’t work for paying customers either. Our priorities, we thought, were clear.

Later on, when prospects came in, we were able to determine whether our product would be a good technical fit for them. We would green-light deals that seemed to mostly work, and pass on the ones that didn’t.

Ironically, our own company was not a good fit for our product. Not because of business or product reasons, but because of pesky technical challenges. We could definitely solve some of the issues and properly dogfood, but, alas, it was always a low priority. How do you justify working on the dogfooding pet project when so many customers are starving for R&D attention?

Over time the situation grew to be absurd. We made demo software, but we didn’t use our own software for doing demos. We would wax rhapsodic about demo tools, while screen sharing a demo in the most primitive way possible. I’m not sure how many deals were lost to this dissonance, but it did become a bit of an inside joke. Unfortunately, we were the butt of that joke.

I’m certain that if we decided to hunker down and commit to dogfooding, the benefits would have been immense. Yes, a deal (or a quarter) might have been sacrificing, but we would have come out much stronger on the other side.

Every challenge would have been a perfect practice run for real customers. And we would be the ideal customer: We’d be the earliest of early adopters; We’d be willing to try anything to make our product work. Sitting in the same office, we’d accomplish in days what would take weeks or months with anyone else.

Hindsight being 20:20, this could have been a pivotal moment for us, but sadly it was an opportunity we chose multiple times not to take.

My final words of advice: don’t let everyday roadmap pressures get in the way of proper dogfooding. The benefits of short-circuiting the R&D-customer feedback loop can be tremendous. You can literally save months and make far fewer mistakes. It’s worth it.