« fact or fiction? | main | proof IA's are born not made »
from Rashmi Sinha's Weblog: Creating personas for information-rich websites
"Cooper's suggested for persona creation suggests detailed interviews, identifying patterns, picking up nuggets. However, there is no tight coupling between user research and persona creation. Furthermore, personas are supposed to be representative of large group. However, interviews are not an appropriate method to find out who representative users are. I am developing a persona creation method that is similar in spirit to card-sorting in conjunction with cluster analysis. As with card-sorting, the persona creation method gets some user input and subjects it to exploratory statistical techniques to find patterns. "
Cooper gets anxious, Jared cheers from the sideines, Goodwin makes an official statement-- gosh, comments just make everything more fun, don't they!
I like the card-sorting approach. The thing about personas that has always bothered me is that they seem a little too specific. By "specific" I mean that they don't lead to a general understanding (an overview) of the needs of groups of users. Perhaps I don't have a good understanding of how personas (nae?) are used.
His approach seems like a good compromise - find the extremes in user needs/wants, and group them.
By the way, have you ever seen personas described as simply a marketing tool to sell clients on a particular user group analysis? I've seen this criticism before. I don't necessarily agree with it, but I think that there's an element of truth there.
Personas aren't meant to lead to an understanding of the needs of groups of users; that's the sort of approach that personas specifically avoid.
It can be very hard to anyone used to thinking in general terms (like "our market for this is 25-40 year old car owners") to make this switch to thinking in terms of one person. Marketing departments, in general, can provide an overview of the needs of groups of users. But general overviews leave lots of gaps that are often filled in by guesswork.
The idea with personas (Cooper's and Rashimi's) is to say "right, we take care of this guy--this specific guy--and a lot of our general needs will take care of themselves."
Also, *I've* used personas as a marketing tool in one very useful way: helping clients understand a site design. Use a combination of personas and scenarios to "tell a story" of a user using the new site. "Jim, the 34 year old car owner, goes to the site to schedule his maintenence check. His schedule's tight this week, so first thing he wants is to see when he can bring the car in..."
(Rashmi's a she BTW)
Actually, Personas do lead to an understanding of groups. But, not groups in the traditional sense. It doesn't necessarilly lead to understand "Yuppies in the late 20's with over 100k in income" or "African American mothers with three children." Personas group individuals by goals, motivations and needs. There may be someone who falls into both those groups I mentioned above who has the same goals and needs. One Persona can be created and this person reached, as well as anybody else who has the same goals and needs. I would venture to say that looking at grouping human beings, the best way to make someone happy as well as to reach the largest group of people with a right-on-target solution is the focus on goals and needs.
Cooper's thinly veiled attempt to boost his class attendance on Rashmi's board makes me sick...