Re: [agile-usability] Who is using multiple personae for one job role?
- On 27 Nov 2006, at 14:19, Desilets, Alain wrote:
> Recently, I have come to the conclusion that Roles and Personaes[snip]
> are essentially orthogonal to one another, and that you need both.
Erm... have some people been saying otherwise? I always thought
persona aim to capture archetypal users - not archetypal roles... or
have I been missing the point all these years :-)
> For examples, we had no difficulty agreeing on roles like[snip]
> PhotoEditor (someone who modifies photos to make them look better)
> or StoryTeller (someone who uses photos to tell a story about an
> event or series of events). We had no trouble either agreeing on
> UserTasks for those UserRoles either.
> But when we started trying to prioritize tasks done by a
> PhotoEditor against tasks done by a StoryTeller, we found we could
> not agree. The reason was that while everyone was talking about the
> same UserRoles and UserTasks, we had very different models of the
> actual people acting in those roles and doing those tasks. For
> example, some were thinking of a personae which we later called
> JimTheEagerHobbyPhotographer, while others were imagining more a
> personae which we eventually called MarthaTheSeventyYearOldGrandma.
> Obviously, the priorities for Jim and Martha will be quite
> different. Jim probably puts higher priority on tasks related to
> PhotoEditing, whereas Martha probably puts higher pirority on tasks
> related to StoryTelling.
> So the only way to agree on prioritization seemed to be to agree on
> which persona would be our primary target (we never did agree on
> that btw).
Yes! This is why I like using Persona names rather than Roles in
stories. Because the interesting breakdowns when thinning stories
happen around persona differences rather than role differences. The
core features for JimTheEagerHobbyPhotographer's CropPhoto will be
different from those of MarthaTheSeventyYearOldGrandma CropPhoto.
By using persona in conversation with the customer you get them to
start thinking about which kinds of user bring the most business
value to particular features. I often start with identical stories
that only differ by persona name, which then get broken down into
quite different features for implementation.
> > With corporate software, there is an existingI do find that I use personas less often in the
> > structure that groups people by function,
> > education, and those introduce a minimal amount
> > of uniformity (again, I'm not saying that
> > individual differences don't exist within
> > a same work unit... Just saying that there
> > is a bit more uniformity there).
> Look closely at anything and uniformity disappears.
> ( http://
> www.despair.com/individuality.html )
enterprise and more often in the consumer world, but I
would attribute this to the relative "horizontal-ness"
of the system under consideration, rather than the
"enterprise-ness" of system.
When you have a very vertical application in the
enterprise context, roles can go a long way and are
often a good enough model to use--especially when
combined with feedback from actual users.
But when the system is more horizontal--a phone system
perhaps--personas become more useful because the role
"phone user" doesn't get you very far.
I agree with Jared that the closer you look, the less
uniformity you have. Thus "horizontal-ness" is
something of an artificial distinction. If are
motivated to look at any vertically-defined role
closely enough, if you spend enough money and time,
you can make a vertical role as horizontal as you
So what determines "horizontal-ness?" I would argue
that (in terms of deciding whether you use personas as
a design tool) you consider both the intrinsic
differences of behavior and motivation, but also the
extrinsic factors: namely the relative value of
investigating those differences on any given project,
and the relative motivation of the client to pay for