Re: Desktop personas (draft)

On Wed, 2005-12-07 at 10:35 -0300, Santiago Roza wrote:
> yeah, and i agree with this of course.  what i don't like is the idea
> of creating "living" characters, at the cost of making them A LOT less
> representative of your target audience.
> for example, alex replaced marcus' typical teenager with someone who
> uses openoffice, firefox and gaim... that must be like 5% of
> 18-year-olders (who have computers), with the other 95% not even
> knowing that alternatives to ms office, internet explorer and msn
> messenger actually EXIST.

Right, but you're actually conflating two problems here (and it's quite

 i. the persona is not representative
ii. personas are not representative

The first one is easy: a bad persona is useless. So, in my example, if
I've invented a persona that doesn't really reflect our user base, then
it's a bad persona. Care does need to be taken to get them close to

Number two is harder. The intuitive side of you probably doesn't like
the idea of detailed persona: this is true of everyone, I think, even
Cooper admits that they are counter-intuitive. 

The "footprint" of a persona - how much of your userbase it covers - is
clearly going to be very small. But, forget trying to cover the
userbase: what you're after is a number of stakes in the ground,
hopefully evenly dispersed, which represent important points.

It's a bit like averages. It's a common saying over here that the
average family has 2.4 children (whether or not this is still right, I
have no idea). Of course, *no* family has 2.4 children. But, the
important point is that (assumption: population follows a bell
distribution) families are likely to be near that figure.

If you can think of it that way, you're likely to see the use of
personas. Of course, like any tool, they can be over-used, but it's a
different idea to market segmentation (which I think is another valid
tool to be used). 

Does that make sense?

[It's also valid to argue that personas are not a useful tool; many
people hold that viewpoint. I don't, personally, but there are
significant limitations to how you can use it IMHO]



[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]