The thing that is wrong with “fail forward”

I just encountered a presentation by Bruce Nussbaum, one that inspired a post on leadership. One of the almost throwaway remarks he makes is that the “fail fast, fail forward” approach is flawed.



You can see the distinction he makes, suggesting that “fail fast and fail often” — what I call “fail forward” — should be replaced by what he calls gaming, by which he means exploring a design through playful or speculative means.

In the world of social tools, I see little speculative design. It’s been years since I’ve seen a really out-there mock-up of some radically different way for people to co-work. Consider how many dozens of augmented reality mocks-up came out in the past five years leading up to Google Glass. But who is using a speculative design approach to elicit a discussion about the fundamental premises of social tools and co-work?

We are stuck in a world of small advances and small retreats. I was poking at Sandglaz recently, and I mentioned that I liked some of its design choices, like its grid display. I received an email from the company informing me that it is dropping the various sorts of grids and going to a single one. I guess it decided it was a bug, not a feature. But if I were actually using it as my team task manager, I might be pissed off.

Instead of this “fail forward” approach, it would have been potentially better to provide the simple grid and make the various others beta features for those who wanted to try them, and then learn from that experience. Instead, it introduced a point of failure into the mainstream use of the product instead of a play area to gain insights.