
In sales, avoiding the long, slow no is an effective strategy. We know that for every yes, there will be countless nos. It’s beyond frustrating to invest time and effort in a sale that won’t go our way. We want to get to the nos as quickly as possible so we can focus on the maybes and the yeses. Minimizing time wasted is more valuable than trying to nail every sale.
The same is true with technology. We can all kid ourselves that we’re immaculate thinkers, unsullied by failure. That every idea we have is perfect. But the brain fart is a universal experience. We just need to weed those ideas out as quickly and cheaply as we can.
I’ve personally reviewed projects where tens of millions have gone on developing a piece of technology, only to find that the whole approach and its components don’t deliver and never could have. By that point, you’re too invested to call it what it really is: a failure. And it’s not a rare as you think. High profile duds are everywhere you look: the Post Office, TSB, BA, the NHS, HS2.
Rather than build out an idea into a complete project plan right off the bat, we can develop a set of experiments. Each experiment helps us validate part of the idea or test an approach to solving the problem.
First, we plan all our experiments up front. We record what we want to test, why we want to test it and what it enables. We then group those experiments into phases and attach time, effort, money and value to each of them.
We design the experiments so that they give us an answer on the key risks as quickly as possible. If we get a no, we’ve minimized the cost and risk. If we get a yes, we move on to the next experiment. We do that until we get a no or until we’ve validated our ideas and we’re confident in the project.
For this to work, we need to be clear on our success criteria from the beginning. And we should be super cautious of moving the goalposts once we’ve started: even when a cherished idea face-plants against our success criteria! Instead, digest the learning points and move on.
Remember, the experiments themselves do not fail, they either support or reject our ideas. Knowing what doesn’t work is just as valuable as knowing what does. In fact, knowing what doesn’t work is often more valuable. It saves us from spending millions on sure-fire failures.
To make the experimentation model possible, you need three things. The first is meaningful, executive-level buy-in. The second is the honesty of a data-driven approach that forces us to confront what we see. And the third is to avoid letting personal brand or innovator aspirations trump risk management, data and reality.
This way, we steer clear of committing to a 2-year project only to find that in the end, we can’t deliver what we wanted. Instead, we find out 1 month in for a fraction of the budget. That’s what I call a win.