Personas and Behavioural Axes Analysis

[John Norman]

Those who have spoken with me about personas in user centred design  
will know that I have been a bit of a sceptic. Nevertheless, we are  
about to start paying a commercial company to help us with UCD! They  
are called Flow Interactive (www.flowinteractive.com).

The reason for this post is that in the kick off meeting, they shared  
something that was far more persuasive than anything I have read on  
the Fluid wiki to date so I wanted to share this insight as I  
understand it.

So here is the big idea: Design personas are different to marketing  
personas. A design persona is not a description of a person, it is the  
result of analysis of *goals* and *behaviours* of users in a  
particular context. The analysis seems to involve identifying which  
goals and behaviours have high discriminating power in terms of design  
tradeoffs, describing them in terms of behavioural axes and plotting  
real user behaviours on those axes. Clusters of user plots one one  
axis become a single characteristic of a potential persona. Profiles  
of plots on different axes become design personas. They *express the  
user research* in terms of different behavioural profiles *that are  
relevant to design choices*.

So for most of you this probably still sounds like gobbledegook and  
you are unimpressed that I have now joined the ranks of the speakers  
of UCD. But Flow gave me an example that really helped me understand  
the concept:

They had a client who wanted some work done on a hotel booking site.  
The client told them the users of the site had 5 personas - the  
business traveller, the short-break weekender, the family with small  
children, the single holidaymaker, and - well lets just say the  
romantic short stay customer. These personas are valid marketing  
personas, with different communication channels, different price  
sensitivity and so on. Flow's user research phase discovered that from  
a site design perspective there were only 2 design personas - those  
who knew where they wanted to stay and those who didn't. That is, the  
marketing personas did not exhibit behavioural differences when it  
came to the factors that impact on the site design and the user  
behavioural differences that did impact on site design revealed a  
different clustering of users into 2 design personas - destination  
known and destination unknown.

Now, this makes sense to me. The next test is to see if it works as  
well for some projected theoretical future (Academic Social  
Networking) as it does for some real-world goal seeking behaviours.

Note that the slideshow linked at the bottom of the main Fluid  
personas page (http://www.slideshare.net/toddwarfel/data-driven-design-research-personas 
) includes a couple of pages (27/28) that appear to illustrate this  
concept, but the slides are not very explanatory.

I hope this helps someone else find their way into this UCD stuff.

John