I'm just running through an agency proposal and the misuse of personas has reared its ugly head again. I'm finding this is now a frequently reoccurring theme and so I thought I'd raise it as a question here on UX.

The problem:
There appears to be a tendency to assume that the creation of personas is inherently user centered design. I see people creating a variety of personas, ranging from very data driven complex models, to simple clearly defined segments. So what's so wrong with this approach? Well, the UX practitioner then appears to try to design the interaction from the perspective of a particular persona.

I fail to see how this is UCD when it is in fact PCD (practitioner centred design). Whilst the practitioner can try to be objective, or subjective from a 3rd party persona view, how much can the actually assume of that perspective? And isn't it a direct result of programmers making interaction design decisions that brought us to UCD in the first place...so how is this any different? The lunatics are still running the asylum - they're just hiding behind personas.

My reason for this mini rant is that I'm really failing to see the value of personas when they appear to revert UCD back to TCD, and bringing designer subjectivity to the fore. I do understand that they have some value (e.g. accessibility for older users) but it appears that the whole usage of personas has been extrapolated beyond original intent and actually in practice are anti-UCD.

Am I way off base here? It has really got under my skin and is one of those issues that is almost keeping me up at night.

Here's an Excellent article by Steve Portigal in Interactions, from the question Does anyone have guidence as to when to use data driven personas and when to make them up? Directly relevant to this thread, and I'm going to vote Portigal for President:

To make a crude comparison, guns don’t kill people, but they make it a lot easier. And personas aren’t solely responsible for bad design or solely to blame for the disconnect between designers and their customers, but they make bad design a hell of a lot easier. To compound the problem, personas enable all of this under a cloak of smug customer-centricity, while instilling bemused contempt.

  • 1
    Reading from the same paragraph as the quoted above, it is apparent that Portigal talking about pseudo-personas that are not the result of the study of real users. He is correct that pseudo-personas are worse than none at all because their apparent realism creates the illusion of designing for the user when in fact it’s no different than creators just building whatever they feel like building. – Michael Zuschlag Mar 5 '10 at 14:43
  • According to Alan Cooper - who thought the personas idea up in the first place (or at least published it) - Personas are not based on research data but on formalising 'what you have in your head' 'infodesign.com.au/uxpod/alancooper "So you’re externalising what’s going on in your head? - Yes" – PhillipW Aug 13 '12 at 10:14
  • There may be some references to academic research on which the Personas concept is based in "The Inmates Are Running the Asylum" - but I don't know as I don't own the book... – PhillipW Aug 13 '12 at 10:18

Lets get our definitions straight:

  • A Persona is a central tendency of a group of users as revealed by research. It’s a convenient shorthand for summarizing the research. If your personas are not based on research, they are not personas. See Spool and Cooper’s Steve Calde.

  • User-centered Design, as a method, is the process of iteratively designing a product through testing and observation of users (e.g., usability testing). It’s called “user-centered” because the goal is to create products optimized for user performance (as opposed to, say, minimizing development or manufacturing costs). UCD does not mean the user is doing the designing.

So, UCD necessarily includes user research (testing and observing users). UCD may include persona creation, since that’s one way to aggregate some user research data, but UCD does not have to include personas creation. There are other ways to summarize what you learn about your users. Historically, UCD existed before there were personas.

Are personas anti-UCD? No. Real personas based on research are entirely consistent with UCD since research is a key attribute of UCD.

Research controls subjectivity. Design inspiration is almost always personal whether it’s from a programmer or a trained UX designer or a user. However, that inspiration is more likely to be effective if it is derived from actual observations and other data from multiple users. Furthermore, you can keep an ineffective inspiration from making it into a final product by testing it on users first. That’s how user centered design works.


I agree with you to a certain extent but I think the same argument can be levelled against almost all the other UX Activities. Imagine creating wireframes (or lo-fi prototypes) without having done the research, analysis and positioning to back up what your building! This is as equally flawed as getting a developer to just get on with writing the code.

As you point out there is a big difference between UCD (User Centred Design) and doing activities which can make up part of a UCD process.

i.e UX Activities ≠ UCD

My answer

Personas aren't anti-UCD. Poor product development process is (but that's not new).

However, as I wrote in a blog post on user typology and the "perpetual intermediate" (a paradigm discussed in About Face and The inmates are running the asylum). If part of the persona data collection process included default attributes such as "beginner", "perpetual intermediates", "experts" rather then task centre they would be inherently more UCD.


Don't use the agency - they seem to have missed to point with personas. If the personas are not research based then they're meaningless. Personas are a method of contextualising user insight research into manageable reference and sense check tools during the initial design process. Subsequent designs iterations should always be stressed tested against real users anyway.

I heard that the misuse of personas described as "like trying to imagine your a Spanish mother looking to book a family holiday in Italy".


There are some excellent responses, here.

The best user personas I've written are based on decent research and are validated by others—specifically non-practitioners who work in the same domain as the user persona. I won't repeat the whole validation story, here. I blogged about validating your user personas on my site. Feel free to read it there.


What personas are great at is giving project teams, i.e. designers, developers, UXers, project managers, marketing teams, etc a clearer focus.

Is this truly a UCD approach? Perhaps not... But what is does do is minimise the risk of team members who shout loudest (who I'm sure we've all met) stealing the show, manipulating a team and ultimately creating a design for themselves.

If personas do nothing else - they can certainly maintain consistent thinking and design approaches across a diverse project team.


I agree with your point about personas reflecting 'PCD'. I can't say I've never devised personas based purely on personal insight and intuition.

While us UX-ers can provide value as experts with market insight, our real value is in knowing the right questions to ask. And when it comes to creating personas a lot of questions should be asked about statistics available to us, and marketing data from our own research and that of our clients.

Personas provide the most benefit when they're based on tangible insight. But I'd say they are still a step-up from developers just getting on with it when they're based on 'UX consideration'.


Any design by a practitioner is practitioner-centered. The practitioner-centered work can also be user-centered or not. It just depends on the quality of work of the practitioner.

If the practitioner creates personas well, that would be helpful in designing a product that is user-centered.