The question of “who?” can sometimes be a barrier. Unless your company or team is structured in such a way that a specific person or people must take responsibility for creating personas, I’d say that anyone can do it. If no one’s done or doing it already, go ahead and do it yourself. Better yet, recruit some people to help you, to try to conquer bias and improve diversity.
“Process” can also be a barrier that people use to avoid creating personas. Who is responsible? Where should we start? How many should we have? There are no perfect or one-size-fits all answers. Ideally, personas should be based on facts. Is there user research or statistics that we can draw from to create personas that are representative of our users? You might not have a UX team to do this user research for you, but surely you have some contact with real users. What about your sales or customer support team - what information can they give you? What type of product is it, and who is it marketed towards? In what geographical locations is the product available / advertised? What kind of people are likely to use the product, based on that? What information can you get from customer reviews? Do you collect anonymised user data already?
Use all these sources as ideas for personas and then just start writing them. Or do it the other way around and create personas as “hypotheses” and then set out to verify them. They will never be “perfect”, but they are a model to be wrong and something to use to get started with. Update and improve them over time as you discover that things are missing or not representative, or just not relevant / useful for the intended purpose. Do you know what the purpose is?
Starting is the hardest thing, but once you get going, you’ll be more fuelled to continue.
I’m not using BDD currently but I’m not sure if I understand your question fully. Are you talking about using user personas or user roles? I can see how you might want to implement a particular feature for a specific user role with set permissions and goals (e.g., IT administrator, team manager), but I’d caution against creating an entire feature just for one persona, as you might end up excluding the other personas who can, technically, also use the feature, but have a bad experience doing so. Personas can, of course, be a good source of ideas for thinking about use cases and scenarios that you might want to use for BDD. However, creating scenarios based on only one persona seems strange to me.
Thanks for your questions, everyone. It was great to see lots of good points and thought-provoking questions. Feel free to continue the discussion by answering some of the questions yourself.
Thanks, Cassandra, all of that is really helpful for me! I like the idea of offering to create the personas myself and then work with others to refine them. I’m starting to think the best way to overcome unconscious bias is to collaborate with other people. At the very least, we hopefully have different biases. I think creating a diverse team is also key, we are making a much bigger push to do that now where I work.
In general testing terms, of course, ‘edge cases’ are fruitful areas for testing because they are fertile scenarios where we find departures from expected or specified system performance. As with general testing methodologies, so with personas.
We are always being told that our software products have to have the broadest possible scope and appeal, because they can potentially be used by anyone, anywhere. So should it be with personas.
(Edit: And I wrote that before reading Cassandra’s response to the question…)
Hey folks, I forgot to follow up with the companion blog to my talk on personas at TestBash Germany a couple of weeks ago - (Mis)Using Personas with the Seven Dwarfs. Maybe this will help with some of your questions and help you on your journey to finding out more about personas
