Persona Development What/How Thoughts

by on May 2, 2007

From time to time it’s fun to think things through using the “what/how analysis.” This can be summarised by the statement “One man’s ‘what?’ is another man’s ‘how?’” and it can be applied to lots of things in order to work out where you are in a set of processes and how, or whether, some things have a natural relationship or hierarchy to describe.

I’ve been trying to apply this technique to the process of persona development, because in particular this seems to me to cut the designer off at the point where they actually need to design the end product (the UI of the system in most cases). In short, I wanted to know whether performing a thought experiment like this would reveal whether modelling users necessarily supports the design of a better system for them or not.

The what/how analysis would paint this picture of the development process of, say, a website as follows:

  1. CEO: What do we need to do? We need to increase profits.
  2. Marketing Manager: How do we need to do that? We need to develop our brand.
  3. Programme Manager: How do we do that? We need to launch a website that supports the brand.

and so on until it gets to somebody saying “We need to create an effective web site.

  • Designer: How do we do that? We need to understand our audience.

As this point, the designer is committing to a UCD process. The next step has to be something like “We need to model our users” or “We need to research the users’ needs” and other similar activities in the persona development process.

The trouble is that I can’t think of an intellectually satisfying what/how link between “We need to model our users in huge detail” and “We need to design the UI.” That is to say I can’t do so if the persona model contains data on what shoes they buy, what papers they read and what what they think about DIY. Why is the huge detail necessary? Only a rather simple understanding of the audience (their age, sex, typical activities relating to the target activity, etc.) seems to make a natural link with UI development itself.

I’m probably being far too intellectual about this, or more likely being stupid and missing some obvious point and will be quietly deleting this post out of a hideous realisation of sophomoric angst.

Comments

Hm. Well, I work as CSS designer for… something like 4 – 5 years or more…? and… neevr asked myself the questions outline by you in this post :-D

Maybe I’m just too simple… dunno:)

When a website needs to be done, I try to do it in the best possible way know to me, for the widest range of possible auditory and… that’s all, no philosophical questions at all:)

Good point. I’m just too bound up in the methods of large consultancies selling “solutions” to multi-national corporations with money to burn.

Well, I guess when you are not so large, then you simply want the job done, in a good way – the website created, the client happy (the graphics nice and the CSS/XHTML clean and well-structured) and… seriously… I sometimes wonder, why we need to make things happen in much more complicated ways? :-)

Is it just excess of money or simply we want to make them perfect?…

Not sure I completely understand the point of your analysis method Jonathan but that is not unusual for me. However my experience of personas is that they tend to be developed at length and then put in the cupboard and left to die. A shame really as they would be good people to ‘invite’ to QA sessions on the efficacy of the designed interactions although I have never seen this done. I have also never come across a UI designed directly from persona work, rather it seems to be used to get a ‘feel’ for the audiences and what kind of interface might suit them.

Michel – given the number of websites which exist with opaque purpose and particularly poor interaction we can only assume that not enough people are making the effort to learn a bit more about their audience before building don’t you think?

I’m not sure I understand the point of my analysis method either, but it’s something to do with validating whether a large complex activity like persona development actually has a connection with another large complex activity like designing a UI. So I break it down into action/consequence pairs and see if I can follow a thread from one end to the other. If at some point I find a non sequitur (like “What I want is to sing – how do I do that? I hit myself with a hammer.”) I know something is wrong.

At least in part, this is about my current angst that persona development (in the way that I currently understand it) may in fact be unnecessary. Either that, or it may just be misguided in the context in which I have applied it. Either way, it makes money so maybe I should shut up.

Well… I am even not sure about what is the discussion >:-o

;-)

As to the websites with opaque purpose and bad interaction – I am sorry, but I am afraid that in most of the cases there was a CLIENT who didn’t know WHAT HE WANTS TO SAY to his audience… and the result is this :)

Yes, I admit it, I did this at least once. I designed and put online a website on some project… It was nice CSS+XHTML, nice, clean graphics, easy navigation, etc. Problem was, the project guys & girls couldn’t explain to me WHAT WAS THE PROJECT ABOUT! Nope. Nothing. Zero. It was about some “consistency development sustainability related to abstract forming of…” — err, I forgot. Anyway, you get the picture.

I even redesigned the site TWO times!

I was the designer.

I did my job.

They provided me with the texts, images, etc.

Is it my fault that the website said nothing to the user?

I guess, no. I got no real content, I then did some nice design, that’s all…

Not always the poor webiste is the result of a bad designer’s work.

It’s often the inability of the client of the website to say anything to the audience…

Cheers :)

Leave a Reply

Your email address will not be published.

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>