Fail to succeed

I love failing. I fail a lot. I go seeking it!....Eh!

I want to fail early before it becomes a big deal. I want to fail quickly so that I haven't spent loads of effort on a dead end. I want to find the right path for ultimate success.

Many, seeking to avoid failure, work really hard to understand the problem. They do in-depth studies, recruit experts and compile detailed reports. In some situations, this is perfectly valid. But in this context, I am mainly considering complex systems. Ones that are hard to predict and for any innovative digital solution this is the case.


Why is digital complex?

Digital technologies involve the interaction between people and machines. Through apps, websites, screens and increasingly voice. People are tough to understand. They are unpredictable; they act on a compilation of logic (on occasion!) and emotion. With people in the loop, even simple processes become complex to get just right.

No expert can pull on their draw of experience and standard solutions. That will not address the need. To really understand what works, we need to test and see the response.


Natural Selection

By building from a platform of failure, we are mimicking the process of natural selection, but for ideas. We allow the ideas that work to continue while learning the lessons of those that fail. This way, we iterate and perfect our ideas across generations. We are improving upon generations before getting us quickly to great solutions.

We could spend hours researching, planning, and designing to develop the 'perfect' solution that completely tanks in the real world. With an interactive approach, we can prove our ideas before we have made an expensive commitment.


Washing Powder

Unilever had a challenge. The process of making their washing powder involved spraying chemicals through a nozzle. This nozzle was not very reliable. It created an inferior product quality (powder of varying size) and regularly clogged, leading to downtime and productivity loss.

Being Unilever, they got the best minds on the problem. Experts in the field of fluid dynamics that could research and re-design the nozzle. After digging down into the theory and designing the ideal replacement nozzle, they achieved nothing! The new nozzle was just as ineffective and just as likely to clog.

So who do you turn to now to solve your engineering challenge? Biologists obviously! With no knowledge of fluid dynamics or engineering design, they did it. They created a nozzle many times more effective. Simply by understanding the power of natural selection, they were able to create something the best experts could not. By creating several variations, testing them and failing many many times, they achieved something incredible. By learning from each failure and progressing what worked to the next phase, they were able to succeed.


What does this mean for me?

When we embark on an innovation project, we are trying to do something new. Something that has no go-to solution and has not been experienced before. Although I have poo-pooed the idea of research, of course, some are needed. We can learn from other success and failures and build a base of theoretical knowledge. This will accelerate our early progress.


What we must be careful of, though, is being too wedded to our ideas. We should be willing to test our theories, our solutions and our plans. By subjecting them to the possibility of failure early, we can adjust and rethink based on our new, precise information.


There are some key areas that this can be most applicable:

Strategy - When we set out our idea, it can be tempting to launch into realising our vision. We have found though there is enormous value to testing your plan before you begin committing your resources. For one of our clients, we found that their strategy didn't support the business model when it was put to the test. We did find that a different product could achieve their vision. Before they had committed to designing and building a white elephant, we had course-corrected. By accepting the possibility of failure, our client has been able to learn and adapt very quickly.

Prototype - It is important not to assume that what we think is an excellent solution that makes a good product in a customer's eyes. We need to subject our ideas to failure and, as always, the earlier, the better. Just as for Unilever's nozzle, a great way fo ding this is creating generations of prototypes that build on each other. Starting simply, you can use quick sketches to try out many ideas. Picking the successful one, you can take these up a level to wireframes, increasing detail and complexity. You can even begin introducing real customers to the testing to see how they react. Finally, you could even have a simulated solution that gives customers a real feel for how your product works and get precise feedback. All this and you have not yet committed to the time and cost of software development.

The testing never ends - Even once you have launched your product, it will not be the end; you must continue to search for failure. Digital products do not stand still because customers do not. Expectations increase, situations change. Keep searching for failure. Things that don't meet expectations and missing features that customers need. Keep learning and developing — staying ahead of your competition and giving your customers brilliant value.

Far from being something to avoid failure is something to search for. Suppose you don't find it, then you are on the right path. If you find it, you learn quickly and before you have spent too much energy on it.

Where are you going to look for failure next?

woman using loud hailer

do you agree?