[IxDA Discuss] A question about Personas

[Petteri Hiisilä]

> I am wondering if it would be valuable to create a repository of  
> personas as
> templates or starting points for interaction design. Does such a  
> thing exist
> already? Would a persona repository be of value to design  
> professionals? If
> a repository would have value I envisage an interchangeable format  
> (?XML) to
> give some structure to the data. Even if a persona definition is  
> mostly
> narrative I believe a good persona would have some important  
> attributes.

Personas do have attributes, and they are the source of the  
narrative. We've also thought about making some kind of repository of  
the research data, but it isn't really needed.

It has been far more important to spend the limited time we have with  
real users. You personally, if you're going to be the interaction  
designer. For a simple consumer product we tend to interview 10-20  
people. For complex enterprise product set we tend to interview 40-80  
people, including end user (candidates), subject matter experts and  
stakeholders.

There's a well defined and reproducable way to create personas, and  
tons of ways they have been misunderstood. Personas were first  
mentioned (popularized) in Alan Cooper's "The Inmates Are Running the  
Asylum : Why High Tech Products Drive Us Crazy and How To Restore The  
Sanity" in 1999, but that book didn't describe how to make them ... ;)

 From the same source came "About Face 3: The Essentials of  
Interaction Design", in 2007 which finally (!) explains in detail how  
Cooper - the source - actually creates personas. I took their  
Interaction Design Practicum in 2004, which was the best career move  
I've ever made. Highly recommended. Take yourself, your colleague and  
your boss there if you can. Read those two books before you go to get  
everything out of it.

Some things to help you in the beginning, I hope I'm not too vague.

Do:

  + Perform the user interviews personally, with a partner.
  + Interview and observe until you you have a firm feeling that
    you know thy users. There comes a point where you can guess
    their answers pretty well in advance. You'll know it.
  + Research all the literature about the subject matter and interview
    experts until you've become an expert too. It's good to be a
    subject matter expert and a designer, but you need to keep those
    roles separate.
  + Base your personas in behaviors and goals, not demographics,
    eating habits or which car they drive. (check the books and/or
    take the Practicum for essentials on this.)
  + Behavioral variables are the source of the narrative. Well
    written personas will encapsulate behavioral information
    "between the lines", and further repositories won't add much
    value. This may be hard to understand/accept at first.
  + User personas as actors in your scenarios. Start with high-level
    context scenarios describing the future behavior. No interface
    details yet.
  + Create high-level requirements based on personas and high-level
    scenarios.
  + Use these requirements to come up with the actual design. There
    are (almost) repeatable methods for design, and I prefer Cooper's
    Goal-Directed Design. They know everything about personas and are
    helpful even after the practicum when you need tips or ideas.
  + Define the detail level interactions and scenarios using
    your personas and elaborating on the high-level scenarios.
  + Use a lot of paper and whiteboard, or even napkins during
    the design process!

Don't:

  - Don't design with a persona set that isn't based on thorough
    user observation and interviews. You'll lose credibility and
    make guesses. Users will surprise you often in the interviews.
  - Don't design with a set that wasn't researched for this product
    or service. It's an exception that the same set works with a
    different product.
  - Don't bake the scenarios inside personas. They are meant to
    describe how 1) people behave 2) right now. They are NOT
    meant to describe how the design will behave in the future!
    (Use separate scenarios for that and add details step by step)

Personas are tricky business. I suggest that you do some reading  
first and pick up an easy pilot project to become familiar with the  
tool and to make a positive impact inside your organization.

About personas:

http://www.cooper.com/insights/journal_of_design/articles/personas/

The Origin of Personas
Taking Personas Too Far
Getting from Research to Personas: Harnessing the Power of Data
Perfecting Your Personas

About organization, process and pilot project issues:

http://www.cooper.com/insights/journal_of_design/articles/index.html

Design Research: Why You Need it
Ten Ways to Kill Design
5 Ways to Get the Most from In-House Designers
The Iteration Trap

About behaviors and features:

http://www.cooper.com/insights/journal_of_design/articles/ 
features_talk_but_behaviors_cl_1.html

I hope this helps you forward. Good luck!

Petteri

--
  Petteri Hiisilä
  Senior Interaction Designer
  iXDesign / +358505050123 /
  petteri.hiisila at ixdesign.fi

  "Simple is better than complex.
   Complex is better than complicated."
   - Tim Peters