My understanding of UX Personas is that they're descriptions of fictitious people designed to represent the common traits and attributes of a broader audience demographic.
If that's the case, in what way is an invented profile more relevant/helpful than picking an actual member of the group being represented?
The answer is in the question.
You said: They're descriptions of fictitious people designed to represent the common traits and attributes of a broader audience demographic.
Traits and attributes. Plural. They are made up combinations of several traits and attributes. Most people might have one or two, maybe even none. You might well spend a while finding the perfect real person who could represent a persona that you might otherwise come up with. Besides you still have to come up with personas to get to this stage.
Then you'll find they don't know what you think they know - or they know too much.
Don't get me wrong - nothing wrong with real users, but no need to get obsessive about finding users that could be personas. Just get a selection of real users - you can learn a lot from most users - they don't have to be the dream team.
Hopefully, you'll find personas are available 24/7, don't eat, don't sleep, don't get bored and don't need paying.
Move onto testing with real people when you think you've got enough content to make the feedback useful, but design/test/iterate with a representative selection of personas in mind, not for Dave in accounts.
EDIT: Added the following as examples of personas where the real person is going to be very hard to find:
Depending on how big the project, how long it's lasting, how much stakeholders need to be persuaded etc etc, then you might want to create quite detailed personas:
In A Project Guide to UX Design by Ross Unger and Carolyn Chandler (website: http://projectuxd.com/) they give some examples of advanced personas. You can find these on the website ( http://projectuxd.com/?page_id=5 ) under Chapter 7.
You can also Look inside the book at http://www.amazon.co.uk/Project-Guide-Design-Experience-Designers/dp/0321607376 and then Search inside this book for Advanced Personas and going to the second result (using the tab at right of 'page')
Now these personas are intended to be detailed and for good reason. As a developer or designer you should get to know these people; to understand them; get their needs, desires, influences, goals, doubts, etc. The more detail, the more real they seem. They should live on your wall; in your head, affecting your design. You should feel empathy with them.
BUT - These are not real people - they represent a type of person in your target audience but they are totally fictitious, maybe made up from hundreds of collected details, all accumulated through researching your audience carefully. To research your audience on a broad scale and then pick a real person as your persona, is like arranging a whole bunch of focus groups and then listening only to the one person that shouted the loudest.
Where are you going to find a single person that encapsulates all the information from one of these advanced personas.
Wikipedia has a pretty good section on the benefits of personas. Here, they summarize what Alan Cooper stated in The Inmates are Running the Asylum (1999):
- Help team members share a specific, consistent understanding of various audience groups. Data about the groups can be put in a proper context and can be understood and remembered in coherent stories.
- Proposed solutions can be guided by how well they meet the needs of individual user personas. Features can be prioritized based on how well they address the needs of one or more personas.
- Provide a human "face" so as to focus empathy on the persons represented by the demographics.
In essence, I believe the benefit is that one consistent set of characteristics are communicated and agreed upon within the design / development team. Had it instead been real persons, you run the risk that "stakeholders may define the 'user' according to their convenience."
AFAIR, Alan Cooper and Robert Reimann treats the subject of personas extensively in About Face 2.0 (and 3.0 too, I assume).
I prefer to go in the other direction personally. http://commadot.com/sticky-personas/
Years ago, Alan Cooper pioneered the concept of personas, which are basically fictitious people who represent the customer. They are stand-ins that are used to create empathy for the real people. They are created by amalgamating different characteristics of potential customers into one pretend person. Personas are very useful for decision making, especially for the question, “Who do we optimize this interface for?”
The problem with this approach is that the names are clearly fake and the people have no built in frame of reference. Generic makes them inherently non-sticky. A little over a year ago, I had worked on a new approach to personas that was meant to be more sticky. I used Muppet characters.
For instance:
Gonzo, marketing director, was out there and would try anything. Miss Piggy, field sales, was sweet until you messed up her sale, then KARATE CHOP!
Your idea isn't awful though. I often will mention actual customers when saying something like, "Do you really think [customer name] will want that feature?"
For an entirely different take, return to one of the original sources of personas in user experience: the book by Alan Cooper, The Inmates are Running the Asylum: Why High Tech Products Drive Us Crazy and How to Restore the Sanity.
In it, Cooper specifically talks about personas not necessarily being representative of common traits. He has some excellent wisdom both about designing for an 'individual' (persona) who is not the most common of users, and the best descriptions of the reasons for them not being 'real.' That section is only a few pages long, so is an easy and informative read.
The whole book is worth the attention: a lot of gold there about every aspect of conception, design, and development.
Sometimes personas need to be unflattering. You might found out about motivations for using a product that no single end user would admit to having. In the absence of honesty or openness from a user, the discovery could be made from anonymous feedback, or from a knowledgeable stakeholder's insight. Tying this uncomplimentary or sensitive information to a real person would be problematic, but a fictitious representative gives deniability to your interviewees while maintaining realism.
