One of the things I spend a significant amount of time on is devising, designing and running experiments on various different ideas for new concepts. It is both fun and challenging.
The challenging part is mostly about not reverting to the same 2-3 types of experiments and use them again and again. But because it is wrong to do so, and you might develop bias. But also because there are actually a lot of different ways, you can design and run experiments based on what kind of hypothesis, you’re trying to (dis)prove.
For that reason I have built yet another Excel-model; a simple database of all the different experiments, we know and can run with titles, applicable stages, ‘how to’-recipies and our know-how and experience in running them with valid results. Using the filter option on that one quickly allows me to narrow down the list of useful experiment-types for any given idea, broaden our horizon – and generate better results.
It is really that straightforward.
There is no way around it: The greatest learning opportunity is when things don’t go according to plan. When you envisioned X and Y or Z happens, and you have very little idea about why that is.
When that happens – and it will happen – it is an open invitation to learn. It is an invitation to go back, investigate and walk through everything you have done in order to try and find the spot(s), where you missed something important.
It is by no means certain that you will be able to find it at the first time of asking. But if you adapt an experimental approach and try to adjust here and there in a controlled way, chances are that you will finally understand what happened, and what you could have done and should be doing forward to make it right. And voilá; you will actually have learned something.
When you try to do something completely new, the only way you will know whether customers like it or reject it is if you show it to them.
This is precisely the argument for why you should be running experiments again and again, as you try to move forward from idea to a product or a service; you need to take stock of your customers to see, if you are essentially on the same page as they are. It not, redo, retool, relaunch or just stop.
Show your ugliness. Give your idea a spin. A little time and money invested in the right experiments go a long way into guiding the big product decisions that are truly costly.
Some ideas intuitively makes sense. Other ideas seem just about the most stupid thing, you have ever heard of. Yet, while the former can go on and become a viable idea, the dumb idea more often than not end up making a killing.
Andrew Chen calls it “The Dumb Idea Paradox”. I would just suffice to say it is a pointed reminder of one of the core guiding principles I have; you just can’t sit at your desk and expect to figure out the next big thing. You need to go out there, be curious and – as part of that process – challenge your own assumptions.
For some people betting on dumb ideas is the thing that they do. Maybe they just do it once, but it will be all that matters, once the idea takes off. For the rest of us, we need to tell ourselves that “no idea is that dumb” again and again while giving it enough benefit of the doubt to at least experiment and play with. Because, more often than not, no idea really is that dumb.
*: Of course there are really dumb ideas out there, and most of us intuitively know what they are. So there are exceptions to the rule, but expect the rule to be the rule.