Dogfooding your product

“Eating your own dogfood” or “dogfooding” is a term that always felt a bit odd to me, but the principles underlying it are incredibly important to product teams small and large. In short, Dogfooding means using your own product in order to better empathize with your users. When you build more empathy for your users, you build a better product. I’ll be sharing some thoughts on why dogfooding is important and some pointers on how to dogfood well.

Why dogfood?
Nothing is more humbling than watching a user struggle with your product.

Devs watching QA test the product pic.twitter.com/uuLTButB3x

— sanja zakovska 🌱 (@sanjazakovska) January 22, 2021

The tweet refers to QA, but really the experience is akin to anytime a product team watches a user using their product in the wild. Dogfooding is important because it forces the team to put themselves in the shoes of their customer or user, which then makes the team more likely to prioritize the work to make the UX better.

One of the key benefits of dogfooding is that it helps you understand how the UX could use improvement or is straight-up broken. Often individual features in a product can be bug-free and work in isolation, but when you jump from feature to feature, the experience just feels… off somehow. Dogfooding, when done properly, puts you in the mindset of a user and forces you to go thought potentially multiple parts of your product.

Dogfooding also helps teams identify bugs. While automated testing is incredibly important in the developer workflow, some bugs are more easily surfaced and identified when a human is using the product. While dogfooding isn’t supposed a replacement for a test suite or a QA pass, you’ll almost certainly find bugs when your team dogfoods.

Make dogfooding easy
Dogfooding is easy if you’re building a product for yourself. Often consumer products (Instagram) or productivity tools (Slack, email, etc.) are quite easy to dogfood since everyone on the team can easily use the product either in their personal life or day-to-day at work.

Dogfooding can be trickier when you’re building a product for an audience that doesn’t target you. For example, if you’re building software for flower shop owners, but you yourself don’t own a flower shop, you likely won’t be using your product on a daily basis, so it’ll be harder for you to understand what your users are experiencing.

Because of this challenge, here are some ways to make dogfooding easier for the team:

• Subsidize it. At TaskRabbit, the leadership team gave everyone a monthly stipend so everyone at the company could use the product and hire people to get work done. While you’re spending a bit more money per month, that money more than paid for itself in terms of generating feedback and ideas on how to make the product better from everyone in the company.
• Make it a team ritual. Schedule time for everyone on the team to sit in a room together and use the product together. It’s a ton of fun to get everyone together to use the product, talk aloud as they run into issues and brainstorm new ideas for how things could be fixed. Sometimes the best ideas come out of these sessions, so have fun some with it! And bring the donuts, of course.
• Have clear use cases. Dogfooding is not QA testing. The core of dogfooding is about trying to think like a user. Users don’t often use a product because they think to themselves, “I want to use Product X right now”. Users have a different goal in mind like getting a specific task or job to be done. As such, it’s important to identify what jobs those users are trying to do, and dogfood the product from that perspective.

Even though we all know dogfooding is important, it can be easy to put off or to make excuses that you know the product or the user better. No excuses-- just dogfood. When we started dogfooding in earnest at Grouparoo, unsurprisingly we identified areas of improvement small and large almost immediately, and as we continue to dogfood, we’re continuing to learn new things every day.