Thread by @rickdzekman

Some thoughts on using Personas in UX:

1/ I’ve found that doing the research to make a persona was incredibly valuable. But that if I wasn’t the person also designing the product then my personas would never get used. That’s on me.

2/ If I was the person designing the product then the persona itself wasn’t even as useful as the mountain of research needed to get there.

3/ I’ve found that, for a specific user scenario, there are *usually* 1 or 2 attributes about a person that affect how they interact with the company, service, or app. e.g. whether they used a physical shopping list for their online grocery shopping + how they search products

4/ Often when a persona is presented as a real person with a name and a photo, I found that the most important details for a given scenario were lost among a sea of distractions

5/ But the exercise in working out what separates users from each other was often invaluable in helping to understand how different people might interact with an experience

6/ I’m a big proponent of Hierarchical Task Analysis. And that it’s not possible to look at motivations for a task without thinking about the kinds of motivations and behaviours that matter most to different kinds of people

7/ From time to time I have found them useful as a communication tool. E.g. A project team full of analysts & developers realised that not everyone plans ahead. That some uni students will reach the subject selection screen having never thought about what to do this semester

8/ I’ve also seen a lot of personas used where User Groups would have been far simpler. Sometimes two groups of users just have different needs. Business vs Consumer, Undergrad vs Postgrad, Admin vs Consumer, Organiser vs Attendee, Pickup vs Delivery

9/ Another pitfall I’ve seen is a lack of coverage of types of users. This is often because personas are done witnout quant. The talk to too many of the same kinds of people because they have no data about diversity.

10/ And sometimes they go for the wrong kind of diversity. Yes it’s often important to speak to a mix of genders and ages. But if you’re making a banking app you need to look at a range of financial situations

11/ And then there’s ignoring marginalised users. Customers in financial trouble, blind customers who can finally shop on their own online, students failing their course, single mums operating their phone with one hand. Often the UI we male ignores the marginalised

12/ Then there are times where personas are impractical. e.g. personas for a national bank, a newspaper with millions of readers, a government portal used by every citizen. IF personas are used they need to be broken down for smaller tasks than the whole experience

13/ I’ve never once come into a new team and been taken through their personas and felt like I understood the research they’d done to date

14/ On the other hand when I’ve done a task analysis I’ve found it much easier to onboard new designers and researchers to my work. Especially as a consultant presenting back to a client

15/ “Here are tasks our users complete and the range of motivations associated with that task, and the motivations for those motivations” seems to be a much simpler way of showcasing your research

16/ But there are situations where task analysis falls flat without personas. Consider my example where students come to an enrolment screen not having planned which units to study that semester. You need to know those users exist to design for them. Persona research finds them

17/ The pretty posters that get stuck on a wall but never looked at are the reason that personas get such a bad wrap. Design research needs to be done so as to make design decisions.

18/ When I’ve interviewed UX’ers for a job I always ask “can you give me an example where you made a design decision because of research you’ve done”. This is a hard question for some because they spent time doing things like personas and then never using.

19/ Not once has a candidate ever answered that question by telling me about personas

20/ Ultimately research needs to be planned to help guide a design. Personas shouldn’t be done because that’s what you always do. They could be created with a plan for how they affect design decisions.

Would be interested to hear your thoughts: @MrAlanCooper @jmspool @GeorgeCockerill @SamuelHulick @stevenfabre @vfowler