You Can’t Test An Idea

One of the things I hear really often is something like “feature X doesn’t work, we tried it.” If you built an idea and it didn’t work, there is a good chance the problem isn’t the idea.


Sometimes the history of a company’s product haunts you. Everything that has been tried becomes a fact, whether it worked or not.

Maybe you added social sharing buttons and nobody used them. Maybe you did a couple A/B tests and didn’t get great results. Or maybe your boss was a UX designer that one time in college and now they just “know”.

Apparently nobody in the world should ever use those ideas, because one person tried it and it didn’t work out.

I have bad news. You didn’t test an idea. You just failed.



When you build an idea, and it doesn’t work, you have failed. Not the idea.

Your implementation of the idea didn’t work. The idea might be great if you do a better job of designing it.

Certain things, like virality, or on-boarding, or social features are notoriously hard. They almost always take a lot of work or experience to get it right. There are whole categories of UX strategies, like social, that need to be grown. Not just launched.

But then somebody — the lord of user experience, apparently — gets impatient, decides that “social doesn’t work”, and shuts it down.


There are a million reasons why an idea might not work for you.

Maybe you don’t have the kind of problem that this idea is good for. Maybe you just copied something you saw without understanding the mechanics behind it. Maybe your users have different needs. Maybe your Facebook clone isn’t working because there is already a Facebook! 

*cough, Google+, cough*

Or maybe the on-boarding is fine and your product just isn’t interesting enough to keep people around.

Blaming the idea might just be the excuse that feels best, because it doesn’t blame the designer.


So how do you fix it?

User research and data.

Blame-based thinking is natural to people. Everybody does it. When you blame your toaster for turning your toast into charcoal it’s the same.

UX is not a recipe. It’s a process.

Invite some users to try your solution and watch. Ask questions. Make them tea, curl up with a blanket, and share your feelings. i.e. — Get intimate with the problem.

*lowers the lights, puts on Ginuwine*

Why did you create social features in the first place? Because your competitor had them? Or because your team loves social shit? Or because your investors think social messaging is the future? Did you even have a problem that social could solve?

Or did you realize that your users could do something really valuable if they work together? Do you have a bunch of information that a community could rank by voting? Is your product something that has more meaning if it is shared among close friends?

You can’t solve a problem if you don’t understand it.

An “idea” is too vague to argue about. User research, data, and concrete problems aren’t. 


One last thought about this: some ideas are actually shit. 

There are types of things that everybody seems to try, even though they have never worked, but that’s a topic for another day.

One quick solution is to ask yourself why you should build this idea. Not why the idea is good or bad, but why it is the right solution here and now.

If nobody has ever made this idea work, or if you’re just doing it because that’s the way you have seen it done, you’re probably not solving anything.

Then again, it only takes one person to figure out something like Tinder or Uber or the iPhone when nobody thinks the idea will work. They all did it by designing a new version of an idea that failed for everyone else.

Not by copying.