What personas are good for
Personas exist to make abstract customer segments concrete enough that a room full of non-researchers can argue productively about them. A well-drawn persona collapses dozens of interviews, support tickets, and sales conversations into a character that feels real. It gives teams a shorthand. “Would Sarah use this?” is faster and more memorable than “Would a mid-market procurement manager in manufacturing with a compliance background use this?”
Used well, personas improve decision-making in three places. Product teams use them to prioritise features. Marketing teams use them to shape messaging and channel choices. Service teams use them to anticipate friction and design better support experiences. In all three cases, the value is alignment — a shared reference point that reduces the number of subjective debates.
Where personas go wrong
The most common failure mode is that personas become laminated. They are created in a research sprint, printed on A3 posters, pinned to walls, and then never updated. The market moves. Customer behaviour changes. New segments emerge. But the persona — friendly, well-designed, and slightly outdated — stays exactly the same.
Another problem is over-confidence in the method. A persona built from three interviews and a gut feeling is not a persona. It is a stereotype with a name. Teams then make real decisions based on fictional detail, like “Margaret prefers email on Tuesdays” or “Dave is motivated by status.” Unless those claims are grounded in evidence, they can lead product and marketing teams astray.
Personas can also flatten complexity. Real customers often belong to multiple segments, behave differently in different contexts, and change their minds. Reducing them to four neat profiles can hide the very nuance that matters most.
Keeping personas useful
The organisations that get value from personas treat them as living hypotheses, not portraits. Every persona should have a “this persona is wrong if…” statement. If the evidence starts contradicting the statement, the persona needs to be revised or retired.
Review cadence matters. For fast-moving markets, quarterly is not too often. For stable ones, twice a year may be enough. The review should include fresh data: interviews, analytics, support logs, win/loss notes, and frontline team feedback. Personas built only on what customers say are weaker than personas built on what customers say and what they do.
Personas or jobs-to-be-done?
Personas work best when the audience varies by identity, context, or motivation. They work less well when the real insight is about the situation someone is in. In those cases, jobs-to-be-done framing can be more useful: what is the customer trying to achieve, what is pulling them forward, and what is blocking them? Many strong customer experience strategies use both. Personas describe who. Jobs describe why. Together, they give teams a fuller picture.
The final test is practical. If your personas are not referenced in product, marketing, or service conversations, they are decoration. If they are argued with, updated, and used to settle debates, they are doing their job.