Learning Fast vs. Failing Fast
By SVPG Partner Chris Jones
At this point, nearly everyone in product at some point has heard phrases like “fail fast” or “embrace failure”. Though they are a bit tired, the intent of these phrases is still good: many product organizations are stuck in conservative patterns where they do little more than optimize on their tried and true ideas, while sensational phrases like “embrace failure” direct attention to breaking out of that pattern.
The problem with this phrasing is that it robs us of the opportunity to call out the behavior and mistakes that truly need to be corrected. In the worst case, lumping together the concepts of risk taking and preventable mistakes leads to sloppiness.
We need to reclaim the concept of failure and redeploy it to cover our inappropriately cut corners, bad decisions, and preventable mistakes. A prototype that falls flat in user testing is not failure, but releasing a product with a foreseeable security flaw is. An A/B test that shows no improvement in customer retention is not a failure, but shipping a product that failed to account for a critical facet of the customer’s’ workflow most definitely is.
As for breaking out of the very real problem of overly conservative product development, I prefer the framing suggested by Tom Chi. Tom talks about breaking out of a culture of “right and wrong” or “success and failure” and replacing it with a culture of “learning”. Instead of praising the ideas that work and condemning the ideas that don’t, we continually draw attention to the insights we’re gaining along the way. Regardless of whether an idea was successful or not, we celebrate what we learned in getting to that point.
Product discovery is where most of this learning happens. Its techniques are all optimized to get to insights as quickly and inexpensively as possible, much more so than the techniques used for creating delivered product. Where product delivery is about creating the product we are confident in shipping, product discovery generates the insights that inform what that delivered product should be.
Adding to the list of genuine failures: shipping a product without doing sufficient product discovery is almost always a failure. So let’s focus our attention on learning and save “failure” for the really bad stuff, like not learning.