Imagining the machine

Photo by zigazou76

You’ve probably been there: a meeting where there’s a great idea to make a new feature. “Wouldn’t be great if…” is often the trigger to a neat, as yet unproven idea. So how can you get more information to help your team decide whether the idea might actually work?

Start by imagining how the process might work. If you were a machine that followed all of the steps in the process, could you end up with enough data points to make a persuasive case one way or the other for a feature?

Consider a hypothetical idea: “wouldn’t it be great to be able to search information about a person so that when you got an email from them you could be better informed.” It sounds neat to get instant information without doing much work. But what data points are available with just an email address? Make a (wish) list, and then google an email address to see what you can find.

So great – you’ve got a prototype. And does it apply at all to the real-world problem you’re trying to solve? Take the next 50 or 100 emails you receive and try a similar web search on them. Capture the results in a spreadsheet and see what kind of accuracy your prototype had in relation to the real world condition.

In this example you’ll probably find that installing a browser plug-in like Rapportive makes more sense than reinventing the wheel. But by making yourself the machine, you created a quick real world prototype and found a direction much faster than coding or making a complicated design. When your quick prototype gives you a stronger signal, you might be on to something that you should test more broadly.