These days (not that I remember those days) we speak a lot about a willingness to fail. ‘Fail fast. Fail often’ we say. This is great, because it can reduce the fear of failure, and encourage an experimental culture, which drives innovation. What’s not so great is when this attitude is abused to excuse poorly thought out work.

When scientists run experiments, they’re aiming to learn. Often they’ll have hypotheses they’re testing, and they want to learn if their current assumptions are correct (or not) in order to move towards an understanding. To them, a failure is not about things turning out how they expected. Failure is not the same as being wrooong. Failure is the lack of understanding or learning. The same should apply to business.

I’d like to see businesses iterate through a series of experiments that increase their understanding and move them towards their goal, rather than trying things and seeing if they work…

Before we move on, I’d like to be clear about what I mean by understanding. Understanding isn’t just about knowing something, it’s deeper than that. Understanding means you can apply and manipulate your knowledge to different situations. We want to try and avoid any post hoc ergo propter hoc assumptions.

In (made up) Practice

I’ve got 12 shops and we have a central warehouse. Sales aren’t looking so good at the moment. Not bad, but, you know…. I think it might be because people can’t carry their purchases home.

My silver bullet? Home delivery.

So, I’ve set up this super sweet new system that allows employees in store to place orders for home delivery. This all goes through to the warehouse, the delivery guys pick it up and then deliver it to the customer. I sit and wait for the money to roll in… But it doesn’t. I’ve failed at increasing sales, but don’t worry… I’ve learned.

Learned what?!

I never set out to learn anything, I set out to make more money! After two months, my business is making no more money than before! All I’ve got now is this apparently not so super sweet delivery system, a big red number in the books, and the knowledge that not a lot of people want the delivery service I thought would elevate my business. What can I do about it, carry on, wait for improvements? What am I going to do differently? How can home delivery improve my business? Can it?

While I know that what I did didn’t work, I don’t understand why, or what to do about it.

Doing it differently

Success can be learning, but it’s planned and leads to understanding that can inform future actions/decisions. Failure is the lack thereof.

In The Lean Startup, Eric Ries talks about the Build-Measure-Learn feedback loop (for those of you that are more familiar with dog fighting tactics, this is very similar to the OODA loop). This is an incredibly powerful tool that, if used properly, can efficiently help you truly understand your business and your customers.

Build: Build/develop/implement something

Measure: Measure the impact of what you’ve built/developed/implemented

Learn: Learn from your measures and use them to inform future decisions

Of course, planning needs to happen.

Remember our goal, or measure of success, is understanding. Despite Build-Measure-Learn being the chronological order of the stages as they’re carried out, we plan in reverse. First we need to look at what we want to learn, how we’re going to measure it, then what we’ll build to enable it all.

Learn: Is the absence of home delivery affecting sales?

Measure: Percentage of visitors that leave the store without a purchase because there’s no home delivery option

Build: A colleague bribing customers on the door to answer some quick questions

This is a simple example but effective. By thinking about what we want to learn and carefully choosing our measure, we can drive some very valuable and cheap understanding.

Use of this tool is key. We don’t want to just know if people want home delivery (If you ask anybody in a shop if they want home delivery, they’ll probably say yes). Similarly, a good measure is the key to understanding, and the trickiest hurdle.

We set out to learn if the lack of home delivery was affecting sales. Once we’ve carried out our experiment, we will have learned much more and at a much lower cost than actually building a fancy home delivery system!

Note that the Build phase doesn’t always mean you have to build any software. We should always try to achieve our  by ‘building’ the simplest thing possible. In our example, we’ve achieved our goal without any development time/resources wasted at all. Yeah, Lean!

It’s also worth noting that this is just the first experiment we run. We should run further experiments and complete more iterations of the BML loop, making sure that understandings fuel future ideas.

Redefining failure as a lack of understanding can encourage experimentation by shifting the definition of value. It can also instil a culture of questioning. Do we really understand this new marketplace? Is that really something that people want? Do we understand this enough to move forward? If the answer to any of those questions is no, then you need to think about how you’re going to better your understanding before implementing a solution. With better understanding, you can move to towards a much better, much more suitable solution. Maybe even the right solution.

And if that ain’t good for business, I don’t know what is.

Speak soon,

Ollie