Is it always OK to fail? Considering net failure value

‘It’s OK to fail’ is a valuable business mantra because our failures are useful - we learn from them and do things better next time.

Like any business mantra, though, it has its limitations, and needs to be applied intelligently.  We should assess the net failure value in pretty much everything we undertake, not just after the event but also in our upfront planning and decision making.  

I define net failure value here as the value derived from a failure in terms of learning, minus the negative financial, brand and customer impacts of the failure.

Even in our relentlessly positive thinking business culture, I think we need to recognise that some failures have a net failure value of less than zero.

 “This is your Captain speaking, welcome aboard. You’ll be delighted to hear I’ve been learning about business transformation and how it’s ok to fail. I’ve also been thinking about our fuel costs and reading up on some advanced gliding techniques which I’m looking forward to trying out today.”  

What makes it hard for us to accept that it’s ok to fail in this (guaranteed fictional!) case, or in the well worn ‘epic bridge fail’ example in the header photo?

Clearly it’s the impact of the failure. If we compromise safety or lose millions on a venture that doesn’t deliver, we’re unlikely to derive a net positive failure value.

Yet we can still point to big failures which have led on to big successes – without the Newton we wouldn’t have the iPad – so surely we have to be prepared to fail big to succeed big?

That’s true in some cases, and particularly for innovators and entrepreneurs, but it’s much riskier for an established business making large scale changes to an existing model, or for complex but ‘more of the same’ projects like building a bridge. We don’t learn much from the epic bridge fail because we already do bridges, so the net failure value is low.

The greater the innovation element in what we’re doing, the higher the likely net failure value. Nonetheless, we still need to consider potential collateral damage, particularly where we are innovating within an established brand, and especially if we are innovating to catch up with a competitor or a disruptor.

When you’re striving to overtake, every innovation failure on your part can be seen by competitors, customers and observers as another nail in your coffin, and that needs to be considered in assessing the net failure value.   

Digital is a pretty good area to fail too. We have Agile, web site optimisation tools and a host of other stuff that means we can try things out at relatively low cost and evolve them if they don’t initially deliver the benefits we intended.

Again, digital initiatives which add new capability tend to have a higher net failure value than ones which change an existing proposition.

If it’s new then there is both more to learn from failure, and less to damage when you fail.  If you can fail in new areas with high strategic value, even better.

A small agile development to deliver the first upselling options on your digital platform has a very high net failure value, particularly if you have a strategy to offer more services at point of sale. A development to streamline your existing order fulfilment process has a much lower net failure value as there is less to learn, and more risk of screwing up something key to your business.         

The move to Cloud IT is an interesting use case where we need to consider net failure value carefully. There’s a lot of innovation going on, and much potential to learn from failures, but for established businesses, the drive to move core business IT to the cloud is still changing the engine in the fast lane, with high stakes for the business.

Where possible, an incremental approach to Cloud, starting with new capabilities and ensuring learnings are fed back into migrating core functions will help keep the net failure value positive at every step.

So it’s not always ok to fail, but that’s okay, as long as we know the difference between a good failure and a plain old ‘fail’.