If I have to quote one valuable learning from our innovation journey, it has to be demystifying the true meaning of failure. Failure is one of those casually used words carrying one too many meanings; everyone defines it differently. In this post, we will try to deconstruct that meaning and see how we can fail fast in the rapidly changing landscape.

So, let’s begin by imagining these 3 experiment scenarios:

• An experiment is about to be halted after the first iteration because we failed to reject the hypothesis.
• An experiment based on launching a new business model was halted mid-way due to COVID as it posed a risk towards on-ground personnel.
• An experiment around a new business model successfully proved the hypothesis but was halted because the direction of the company changed.

Without sufficient context, these scenarios might appear like failures but once you start exploring the after-outcomes in each of the cases, I am sure your perspective will change. To me, none of these scenarios were failures as they helped us gather new information which further aided in creating better chances of success.

• In the first scenario, we iterated and created a different variation which ultimately ended up being in production.
• In the second scenario, we extracted learnings from whatever was done and applied it to our existing business model.
• In the third scenario, the new model saved the day for us during COVID disruption in areas where our existing model wasn’t able to help.

Well, then, this brings us to the question — what really is a failure then?

Failure to me is anything where we have inconclusive results, where we make no attempts or quit too early and lastly where we haven’t learned anything.

Let’s look at each one of these in detail.

Inconclusive results

The most commonly observed factor for inconclusive results is the lack of goal clarity. So, we always make it a point to repeat hypotheses every time an experiment is being discussed. I believe the same idea can be pursued for two very different goals and can have very different success outcomes. So, understanding the goal clearly is critical because your entire experiment depends on it.

Additionally, every good experiment is designed to validate the most critical risk or the most critical assumption. If you aren’t doing this, then use it as a red flag to see if you are off-track.

Another piece of advice that always works is bringing a good balance between details vs abstract picture and short term vs long term. Often, inconclusive results arise because the team isn’t engaging in thinking on both levels.

No attempt/Quitting too early

Generally, failure to attempt happens because of the fear of the unknown. Relating this to the evolution perspective, humans are designed for taking the path of least resistance. To avoid such traps, as leaders we need to create environments that provide incentives to try new, unobvious and unknown.

Furthermore, the temptation to quit too early happens whenever we see inconclusive results. This leads to a wastage of time because someone will have to pick up the same project again to understand if the idea works in reality or not. To conclude, one must never quit without asking why.

Not learning anything

The primary reason for not learning anything has a lot to do with — not having permission to fail. We are conditioned to not make mistakes from childhood. We are told — one mistake and boom; you are done. To remove this existing conditioning, we need to enable people by reducing the cost of failure for them. We can do this by:

• Serving big risks in small chunks
• Rewarding people for mistakes; not all mistakes, only the new ones

Now that we have a new definition of failure, how can we fail fast to produce better results and drive innovation?

• Keep asking “why” until you have a conclusive answer; having a conclusion is important.
• Experiments with structures and frameworks
• Self-impose constraints by
  Keeping high benchmarks for goals for revolutionary ideas
  Keeping high benchmark of scale for evolutionary ideas
  Adding stricter constraints for speed of execution

These three things not only ensure we don’t get lost in the details and nitty-gritty but also enable the team to ask retrospective questions like Are our existing models of working impeding our learning and speed? If yes, what needs to change? Is it structure or the benchmark or the goal?

Lastly, let me leave you guys with this amazing quote by Kenneth E Boulding

“Nothing fails like success because we don’t learn from it. We only learn from failure.”