Fake it good

The value of realistic made-up content in usability testing.


After a quick glance at the screen, Tester #4 asked the moderator to turn the phone over. He couldn’t look at the mocked up design anymore: the data in the prototype made him anxious. This surprised me, as all the content in the app was fake.

We were in the middle of a formal usability test for a new quantify-yourself-type of medical mobile app. Leading up the the test, I thought the client was a little too concerned about getting the the fake graphs and charts just right. Do I really need to bring that curve down a touch? Does it matter that if certain values might be a bit high for that time of day? My faked content was accurate enough to not be a distraction, but maybe not 100% realistic. Who cares, I thought. It was close enough.

But Tester #4 wasn’t the only one of the participants who had a real reaction to the made-up content. All of them constructed stories to explain the charts they saw on the screen. For Tester #2, the spike in one line meant he must have been at a birthday party with his granddaughter. For #7, it meant he was eating his morning Cream of Wheat.

I probably should not have been as taken aback as I was. In his book The Storytelling Animal, Jonathan Gottschall discusses the uniquely human need to tell stories as a way to make sense of the world around us. This holds true for both big things, like the origins of the world, and for little things, like these faked jpg charts. For the testers, these lines were not arbitrary, like I thought, but a true visualization of their well being. So if the chart showed a decline in health, the testers reacted with real alarm. Heavy stuff, no?

Some takeaways I got from this experience:

1. If you can, don’t use fake content. Yup, this is a no brainer, but since getting access to real data is not always possible…

2. Make sure to use good fake content. If it’s bad in any way, it will be a distraction.

3. Good fake content makes the interface invisible. The users in our test didn’t focus on the UI. At first this was frustrating: I was there to test the UI and it largely took second fiddle to the content. But that is actually what you want. If they ignore the UI, that means it is working.

4. Users will create stories out of fake data. And if they don’t, try asking them to do so, to see if they can give it meaning. If they can’t, maybe your product isn’t visualizing the data as well as it can.

So I suggest: next time you do a little usability testing, encourage your participants to tell you a tale about the data they see. It will help you remember that you’re designing for people, not a demographic.