Intentional Innovation: Giving ideas direction
Having ideas can be easy. Brilliant ideas, harder. Generating an idea which you will build to meet a need is even harder. All too often teams pursue an idea because they believe it will lead to something good without considering the customer need they’re trying to meet. Sometimes there is a place for the “build it and they will come” approach, but more often this creates waste and excess cost which can be avoided.
Ideas need to have direction and that comes through having a culture that not only embraces experimentation but does it with structure. There are lots of methods you can follow to do this, but I’d like to suggest there are some key things that you should ensure you include if you want to benefit from experimentation.
Have a hypothesis
Most science starts here and remember science is more art than it is a fact. Science provides an understanding of what happens but accepts you cannot prove everything. There should always be room for doubt.
The hypothesis is the statement of belief that you want to test. This is your scope and focus. It is key to get this right for the team who will work on this to understand what they should look at, and what isn’t in consideration now.
If you’re unable to arrive at and agree upon a solid hypothesis, then there is the risk that when the experiment happens and you get feedback, then it won’t bring clarity that supports decision making. If you’re not clear at this stage, then spend more time developing that hypothesis until it has a laser-sharp focus.
Identify the riskiest assumptions and plan to test them
Life is full of risk and assumptions. We need to tolerate and accept them to do anything. The problem comes when we act in ignorance of their existence. As you look at the hypothesis and the plans, think about any assumptions that sit behind this viewpoint. If any of those assumptions prove false, then what would it mean for your idea? Tease those out, ideally as a cognitively diverse group, and then assess the risk attached to them. These are the things that need to receive attention in the experimentation and testing stages.
Plan for fast feedback
Getting feedback is only half of the picture here. Being able to act because of what you learn and change your experiment accordingly is just as important. It’s all too common for the evidence to point towards a different answer, but the sunk-cost fallacy, or other bias and strongly held opinions prevent the pivot taking place. If you have planned to change at the beginning and built-in pivots to the methodology then it will be easier to embrace it when the evidence points towards needing to do so.
Know why you’re doing this
There has been lots of discussion about organisations needing to be data-driven and there is good justification for this conversation. However, you need to get the data that will support decision making – and that can only happen when you’re prepared to be experimental. Having data differs from having the data to decide. If you have run a marketing campaign but haven’t run an A/B experiment, then you’ve only got half the picture and the data you’re using to decide will not tell you everything.
The culture of the organisation also plays a part in experimentation as you need to continually learn, and to do that means accepting that you will sometimes (perhaps often) fail. Again, be aware of chasing the sunk-cost fallacy. If the hypothesis doesn’t prove to be correct, this is useful – perhaps more useful than if they had proven it. But the culture of the organisation needs to let that learning emerge to be shared and built upon. If everything always goes to plan, you don’t know if you’re just getting lucky or if there is a reason for it. Failure can teach us what works well.
The value of the experiment
Market research and focus groups can tell you something, and they have their place. The difference with experiments is that the data you get is more useful to your decision making. You can test specific assumptions, learn from feedback and make changes before testing again. The data is specific to the hypothesis you’re working towards and so has focus.
Finally. and perhaps most importantly, when you run experiments then you interact with people as they behave and not as they say they will.
Well-designed experiments are better than market research – you get more data, test more assumptions, can interact with people as they behave, not as they say they will. We have pursued many good ideas through to development only to fail because humans don’t do what they say they will.
Building through experimentation, failing and learning, knowing your assumptions and testing their validity, will not only make for more interesting work but lead to better customer outcomes being achieved.
