There’s going to be a lot of designers, marketers, and product managers reading this and thinking to themselves, ‘what the hell is he talking about?’, and given the widespread use of personas in business today, such a reaction isn’t surprising.

However, just because something is done by a lot of businesses and the people within them, that doesn’t make it an effective practice, or the right thing to do. We see this throughout our lives on a regular basis. It’s a recurring theme.

Usually, what I’ve found to be effective in all aspects of my life, especially in product where we’re expected and often given the license to go against the status quo, is to do the complete opposite of what is the accepted norm and conventional wisdom.

Does this mean that it always works? No, but my own experience and the evidence elsewhere in the market has shown it tends to lead to better outcomes. Therefore, it is something I will continue to do, and I encourage you to do the same.

So what’s the issue with personas?

Personas are outdated snapshots before they’re even created

Firstly, personas are often used in enterprise level organisations trying to introduce user-focused behaviour in an environment whereby users were not the real focus (i.e. an agile and/or digital transformation).

The issue here is enterprise organisations should understand the market, just like wider society, is a fluid organism that constantly ebbs and flows like water. Change is the only constant. Nothing is static. Especially the people.

Users themselves are people that operate within the parameters set by society whether they be regulations, cultural norms and practices, common and/or criminal law, or any other constraint imposed upon a populace within a given environment.

Lewin’s Equation is a heuristic formula proposed by psychologist Kurt Lewin, 1933:

B = f(P, E) — Behaviour is a function of the Person and his/her Environment.)

The market is not standing still. It’s a living breathing organism of interactions that’s continuously evolving. A multi-variate equation that I’m not even going to pretend I can work out or make sense of in this post.

It would be disingenuous to do so. Anyone who could work this shit out would be worthy of a Nobel Prize in the social sciences. I can assure you that isn’t me.

In reality, a user persona has as much value as a Kodak moment. A nice photo may provide us with some degree of insight into what someone once did, or their personality at that specific moment or period in time.

However, it doesn’t give us any real valuable insight into their present situation, their character or personality, whether or not they’ve changed, or how their future will pan out.

They’re a gross over-simplification and absolutely worthless at scale

In the same way CV’s and resumes are a gross over-simplification of people, so are user personas. All of the aforementioned belong in the rubbish bin of the business world as it evolves and accelerates into a new digital dimension.

Even the best crafted personas only begin to scratch the surface on the complex nature of an individual or group of individuals interacting with one another, or engaging with the environment around them.

Some include emotional drivers, others go deeper into psychological analysis, but the reality is even if we put together 50 personas and found correlation between some of those developed, correlation does not necessarily equal or imply causation.

Those 50 personas simply do not go anywhere near providing enough authentic insight into each individual, the market, and especially the thousands, if not millions of monthly active users (MAU), that any serious business wants to adopt it’s product(s).

In summary, they do not really get to the root cause of why people made those decisions within their environment at that specific moment in time. They are an inauthentic snapshot put together in a feeble attempt to make sense of an unknown.

They’re often bound together with another time-wasting exercise in a digital globalised world — user interviews

Interviews or workshops are often the first step in speaking with users before user personas are put together based on the information obtained during those sessions. Sometimes insights obtained elsewhere are also included. Often it is not.

The problem with this approach is that it’s built on the premise that users are, in a gamed and set up environment as opposed to in their normal day-to-day life, going to tell you the truth on what they do, and behave in the same way they usually do.

The reality is, all the research indicates the exact opposite and it’s something that has even been picked up on by Eric Ries, author of The Lean Startup.

“We must learn what customers really want, not what they say they want or what we think they should want.”

And:

“If you are asking, you’re not there yet.”

And finally:

“Most of the time customers don’t know what they want in advance.”

Does this mean that user interviews are a waste of time? Absolutely not! It all depends on the desired outcome of the workshop.

If the goal is to run a Jobs To Be Done (JTBD) workshop to understand what potential capabilities and use cases could be required, then that’s a perfect reason to run a workshop with a subset of users before committing to build.

However, the focus should be on quantity of ideas obtained before carrying out experiments that adhere to the scientific method to validate those ideas at scale. This will help identify which are viable and add real value, and which do not.

It’s important to remember, features and use cases that are not being adopted and do not add value are effectively technical debt. They are still part of our code base and need to go through the same testing cycles as those that do add value.

What should we do instead?

This will depend on the context of our particular situation but the approach that has worked for myself is to define our target audiences, and then seek to understand the jobs to be done by each of those audiences.

From there the goal is to build a working MVP of the product with as frictionless a user experience as possible (we’ll probably fail — a learning opportunity) so that we can start testing something real and tangible at scale using digital tools, and start making data-driven decisions.

We do not want to fall into the predictive trap. Once we realise that the world is organised by jobs to be done, the task at hand becomes a lot clearer, less scary and the idea of initiatives and projects seem absolutely ridiculous.

The focus should be on continuous improvement via a build, measure, learn feedback loop. Would we plan a three month project to go to the gym? Or would we go to the gym, measure the progress, learn from it, and get back in the gym? Think about it.

We go to Skyscanner to book flights to go from A to B. We use Uber to solve essentially the same problem just within a smaller geographical range. We go to the supermarket to buy food so that we can ensure we don’t go hungry or thirsty.

The world is organised by jobs to be done to help solve problems that exist. Content, user journeys, and everything else should focus on complimenting and supporting those jobs that solve problems. Everything that doesn’t is a waste of money and time.

What about the risk?

You may be thinking, ‘isn’t this a huge risk? Shouldn’t we build and test a prototype first?’ — If this is applicable to your situation then go ahead but it isn’t always, and I would question anyone who wants to do this without a strong rationale.

How do we add Google Analytics and Hotjar to a prototype and obtain real qualitative and quantitative insight at scale from actual users if we are not getting it direct from early adopters of the product?

Early adopters are our best source of insight and the majority tend to be more forgiving permitting we remain committed to improving the product and remain in close communication with the community.

Anything that’s going to grow and be successful, there’s always an element of risk involved. That’s just the nature of the game. The real risk is in the cost of not doing and simply letting fear lead us into the delusion that we can predict the future.

We can further derisk by building a real MVP with low investment of both cost and time, and putting a focus on delivering small batches of valuable increments.

There is a lot more risk involved in investing $1m+ in a huge enterprise level project with personas, extensive up-front testing and user interviews only to see if fall flat on it’s face because we find out the hard way you cannot predict the future.

It’s time to get comfortable with being uncomfortable.