Is analysing success more important than analysing failure?

At some point between the birth of Silicon Valley and the release of the iPhone, failure became cool. Rather than keeping their failures quiet and focusing on their successes, entrepreneurs began parading their botched ventures and strategic missteps proudly.

It was a mark of credibility, showing not only resilience (by definition; you tried again) but also that you had something to learn from.

success and failure

There’s a lot of wisdom behind the TED Talk glamour. Failures can be a remarkable learning opportunity. Some lessons about what not to do can only be learned the hard way, particularly where it concerns recognising your own limitations. And the very act of reframing failure as a learning opportunity will help to develop that all-important growth mindset.

Yet it can also go too far. Why should we assume that a person would actually learn the right lessons from failure, for example? I might conclude from an unsuccessful product launch that it was the right idea at the wrong time; history might eventually suggest it was just a bad product.

The same applies at a prosaic level. Most businesses spend considerably more time debriefing after an initiative with poor or mediocre performance than after the ones that did brilliantly. If done well, this will result in improvements the next time. But what if there shouldn’t be a next time?

Focusing on the lessons of failure in granular detail can inadvertently lead to incrementalism, without addressing whether the starting proposition was fundamentally weak. At the same time, it can be bad for morale to spend too long dwelling on what we all did wrong.

Even if you take a purely no-blame approach to analysing failures, there’s a risk that instead of trying to pass the buck and bury our mistakes, we end up on the other end of the spectrum and become blasé to them.

Embracing failure is clearly not as straightforward as it sounds. Perhaps we should spend more time analysing our successes. It’s too easy to take them for granted, using them to reinforce existing beliefs about our value proposition or strategy or personal genius, without considering what else they could tell us.

At the same time, perhaps we should also not spend too much time dwelling on the past at all, whether successful or otherwise. After all, even incremental improvements on the past are only possible if conditions are stable.

It’s hard to find any examples of truly game-changing innovation that emerged from obsessing over lag data. Amazon’s data scientists may have used it to work wonders in its supply chain and logistics functions, but all of its big ideas – from one-click pay, to ‘frequently bought together’ suggestions, to the Prime subscription – came from looking forwards, not backwards.

So yes we should analyse failure, and success, but we should approach them with great care as well as great rigour.  Indeed, the real lesson from learning from failures is just this: that whatever happens, we should keep learning.

Five tips for analysing failure (and success)

  1. Think like a scientist. Don’t just do something and then look for a reason it went wrong. Where possible treat innovations as an experiment, with a hypothesis and a control.
  2. Choose your data carefully. Historical data can be very useful but you need to recognise its limitations very early on. If conditions are changing or you need to know a question that data can’t easily answer (for example, why a customer did something) then don’t dwell on it.
  3. Beware of too much blame – and praise. You are looking for lessons to learn, not to browbeat anyone or to inflate their egos. In the debrief at least, try to keep it neutral.
  4. Encourage dissent. It’s so easy for groupthink to set in. One person sees a reason why something happened, and then it becomes the narrative, with the whole team looking for evidence to back it up. Instead, get someone to play devil’s advocate, challenging assumptions before they become accepted reality.
  5. Evaluate your process. You can also learn lessons from past debriefs themselves, and whether or not they resulted in improved outcomes. Keep transparent records, examine the process and don’t just look at the last year’s examples.