11 Oct 2007 0746H

On personas, part 2

Last time we talked about personas and why we use them. This time we’ll talk about what ought to go into a persona.

It’s commonly believed that the persona is little more than a made-up name, a picture of the person, and a little story that goes into the context surrounding usage of something, but I’ve seen how often this is given more focus than the data and research that surrounds a really good persona. Because it’s a story, which is a powerful way to convey information, it is cuddly in terms of being able to relate to it, more so than cold hard numbers, sometimes we fall in love with the story, and begin to lose the reasons why things are the way they are. And then we might begin to think that there is only one true story, when in fact, there are many many stories like this one.

Here’s an example of a persona from Weber. I don’t know if we could say Jim’s a persona, in the strict sense of the word, because he seems like a real person but they may have just found a fan/friend who was enthusiastic enough and who fit this particular persona. Here they focus on the narrative but it’s okay here because it’s a consumer facing document and may have originally derived from an exec summary document. As usual, the content is determined by which audience you intend it for:

Persona Jim from Weber Nation

Sometimes too, due to fears that personas won’t meet the scrutiny of quantitative rigor of more demanding (i.e., metrics-focused) clients, we focus too much on quantitative analysis and this has the effect of losing the forest for the trees.

It’s surprising how many people pick one of these approaches and then fall back in surprise as their personas fail to work for them. Probably they do so out of time or budget constraints, since this work generally takes about 3-6 weeks to do properly.

Either of these approaches alone misses the big picture. What people don’t realize is that these approaches are two sides of the same coin: the numbers never tell the whole story and the stories are anecdotal without the data. We have to use a mix of approaches and data sources in order to make sure that we’re looking at the users in a wholistic manner, and of course, by doing so we strive towards face validity so that we know our data is representative of the groups we’re seeking information upon.

It’s too much trouble at this writing to go into the ways to ensure that what we are studying is in fact what we ought to be studying, but it wouldn’t hurt to revisit say, social research methods or a basic sociology, stats, or psychology methods course to refresh. Still, it is likely that because of our constraints we won’t achieve the kind of rigor that would allow us to draw inferences from the data, but that isn’t really what we’re trying to achieve when we create personas for the web or for interaction design. Of course our observations must be scientifically valid and rigorous, but we need to seek the proper level of rigor. What we’re trying to really do is to bring ourselves and our team into alignment with our customers in order to design better products and thereby, more passionate customers.

Remember, the personas are always a means to an end: we use them to guide, align, and communicate our design thinking so we can make better decisions. And they don’t just get created and are there forever, but they grow, they have to be nurtured and fed, matured, and eventually, retired.

Next time we’ll talk about how we use personas in our work.

Permanent link to On personas, part 2

Filed under User Experience, Web

2 responses to "On personas, part 2"

RSS feed for comments on this post. TrackBack URL

11 Oct 2007 1115H

Tim writes:

Jim looks like the kind of guy who keeps his car forever, and has a box of plastic bags from safeway in his garage.

18 Oct 2007 1927H

Gino writes:

See? You’ve latched onto his persona.

And now it's your turn.

Fire your weapon, soldier. Just be careful of friendly fire. NAME & EMAIL required.

Your response