Dogfooding your own SaaS

Dogfooding your own SaaS is definitely a great way of testing all the features of the product and producing something with fewer bugs, nicer UI, and a better user experience. Overall, it can surely help in making the final product better for the customer.

However, it can definitely be done poorly and divert you from the path. One of the downsides is that your product will never be good enough for you, and if the team has a hard time deciding which features should take priority and are more relevant, it can lead to unnecessary feature scoping and delays. In those cases, teams should definitely make clear boundaries on what's relevant for the business and prioritize. I'm sure we're all guilty of using our own products and realizing every single quirk; it's never as perfect as we want it to be.

Not only that, but engineers will never look at the product in the same way the customer does — this is especially true if engineers are not the target customers of the product. It's definitely not a replacement for QA, or even worse — unit testing. The team knows too much about the product and can never act as a true customer; they usually never explore all the flows and different scenarios that the customer might go through.

Can you even dogfood the SaaS if you're not its target customer? Dogfooding can help, but I don't think is a requirement for building a great SaaS. Just something that crossed my mind.