The persona that describes everyone describes nobody

Here is a persona. See whether it sounds like one of your users:

They're busy and juggling more than they would like to. They value their time and get frustrated when a tool makes simple things complicated. They're comfortable enough with technology but wouldn't call themselves technical. They've been burned before by something that overpromised, so they're a bit sceptical of new software.

If you recognised your customer in that description, it's worth sitting with the fact that I've never met them. I wrote it in a minute, knowing nothing about your product. It works the way a horoscope works. However, this is the case for most product teams!

Why it feels accurate

In psychology this is known as the "Barnum effect". Essentially, when you give someone a profile made of statements general enough to apply to anyone and tell them it was written for them, they will rate it as impressively accurate! This is because each reader supplies their own specifics without noticing.

Persona docs can be prone to it. Claims like: "Values simplicity", "time-poor", "wants to feel in control": probably true of your customers, and equally true of mine. A trait shared by everyone can't distinguish between different individuals, so if your persona would be just as accurate for a different product in a different market, it probably isn't describing your users rather than someone else's.

The awkward part here comes from the fact that this is what makes personas feel good in a meeting or in front of your team. Everyone can recognise this vague description as they can apply it to their own perception of your users or customers. Ultimately that agreement gets translated as validation when it's usually evidence of general vagueness. One good tell is that real specifics tend to produce disagreement within teams. A persona nobody argues with probably isn't compelling enough yet.

When the summary becomes a person

The second problem perhaps matters more. Philosophers know it as reification, which is basically treating an abstraction as a concrete thing. A persona usually starts as a summary of research, then gets a name and a photo applied, and somewhere in there your team will likely stop treating it as a summary.

You can almost hear when it's happened, because the question will most likely change from "what did the research show?" to "what would X want?". The first can be answered with evidence, and sometimes the answer is "we don't know" *shrugs*. The second question can always be answered, because Sarah is imaginary and imaginary people could theoretically have a view on everything. Ask her about a pricing change that never came up in an interview and you'll still get an answer, and it will feel like it came from a customer when it came from your perception.

So what happens is that the two compound. The persona is vague enough that everyone agrees with it, and reification turns that agreement into an authority you can consult about anything, while it still carries the credibility of the research it started from.

What I'd do about it

Two habits, and they are both straight forward:

First, a thirty-second test. Ask whether your persona would still be accurate for a completely different product in any completely different market. If it would, you've essentially written a horoscope. The traits that survive are the ones doing the work, and there are usually fewer than you'd hope.

Second, notice which question your team is asking. When someone says what would Sarah want, redirect to what did we actually learn, and be willing to answer that we don't know. That's unsatisfying in a meeting, which is exactly why the imaginary person is tempting, since she never leaves you without a view.

A persona is a summary of evidence, and it stays useful for as long as everyone remembers that. Once it becomes a person with opinions, you've stopped consulting your customers and started consulting yourself, in a costume!

This is where Corha comes in; any evidence absorbed from any source will build and evolve your personas, even if you just leave the platform to do its thing with your data - and the best part? any leading areas of ambiguity will be highlighted to your team by Corha, and will help you build research projects (and will run for you if you want) to uncover that ambiguity. So lets say that one of your personas only had 1 call's worth of info about how a representative user views your pricing, the model will detect the need for more real-world data on this, and write you up a research plan and/or automatically send out a short questionnaire to users from this representative sample. No matter how strong your team is on research, you always have a way to gather this information!

Thanks for reading and drop me a message on Linkedin if you want to chat!