When failure is not an option

Over on The Little Innovation Book, Malcolm Lowe asks the question:

What are your thoughts on organisations were failure maybe is not an option. For example nuclear physics, NASA or a government organisation that pays benefits. In these situations failure could be disastrous. What strategy would you recommend in these types of organisations?

This is such an interesting question, that I thought it worth a whole post all on its own.

The point I was making in that chapter of the book was that 80% – on average – of new things innovator try will fail.

When you start new things, there are all these processes you have to go through before anything happens. You have to write business cases, win money, and eventually, deliver stuff.

Failure that occurs at any point before delivery starts is “good failure”, and in fact, is something that innovators should develop competencies in if they want to stay in business. The fact is, before delivery starts, you probably haven’t spent much money on whatever-it-is, so the downside of stopping isn’t that great.

On the other hand, the further you get into delivery, the more money you’ve spent. If you have to stop then, its very bad indeed. As innovators, you don’t want that situation occurring if you can help it. It leads to what academics call “innovation trauma” – the scenario where everyone is so burned by a failed innovation that no-one will ever sign up for anything new again.

For a great case study of innovation trauma, do a Google search on “SunRay”, the Sun Microsystems thin client which wrecked the careers of a number of smart people, and made the company’s salespeople unwilling to take any risks on innovative new products for years.

The point is, failure which is “good failure'” should be most of that 80% that I refer to in the book. It doesn’t damage anyone, and probably offers instructive lessons to the innovation team

Even one bad failure, though, can close down an innovation programme. And clearly, in the cases that Malcolm mentions, that kind of failure has very dire consequences indeed.

One last point on this: “failure is not an option” is a mentality that leads to – you guessed it – failure. Trying new things is a process that requires lots of stops and starts. There will, inevitably, be more stops than starts, actually.  In an organisation that doesn’t celebrate good failure, what you get is a scenario where nothing new starts at all.

That, clearly, is a very bad situation to be in, and is one of the main reasons people complain they “don't have enough innovation”

Comments

[Peter Evans-Greenwood]

The to close the connection between Malcolm Lowe & James, failure of something like a heart lung machine is not an option, but that doesn't me we can use an innovation programme (chock full of failure) to create a better heart lung machine. We need to distinguish the journey from the result. In this context they're called experiments or proofs of concept and, as always, its better to fail early than later (as James points out).

Really, this is just the distinction between waterfall and concurrent engineering.