When are personas actually worth it?
As part of my studies, I looked closely at personas. My starting point was the question of why personas get created in many projects but don't always make it into everyday use. In theory, they're a central tool for making users tangible and supporting decisions. In practice, it often looks different.
Classic personas and their biggest problem
Personas are often used to make users tangible and to support design decisions.
In practice, however, it turns out that they're often not used long-term or are applied incorrectly.
This is mainly because many personas are based on one-off research and then never maintained further. Over time, they lose relevance or stop being taken seriously by the team (cf. Alex & Robra‑Bissantz, 2024).
So the method itself isn't the problem – how it's handled is.
One possible approach to solving this is data-based personas. They combine different data sources and can be adapted more flexibly.
In theory, that sounds very sensible. In practice, though, it quickly becomes clear that this approach is labor-intensive. Data has to be collected, analyzed, and continuously maintained.
Especially in smaller teams, that's not always realistic. For me, this led to the realization that it's often not about the "best" method, but about a solution that works in the given context.
Synthetic users in the context of AI
A current approach that's gaining importance through the use of AI is synthetic users. AI-generated personas can be especially helpful in early phases, for exploring hypotheses and developing a first understanding of target groups (Nielsen Norman Group, n.d.).
At first glance, this looks like an evolution of classic personas. The profiles are more detailed, faster to create, and often feel more "alive."
But that's exactly where the challenge lies for me.
Synthetic users can foster empathy and help make target groups more tangible. At the same time, they're not based on real behavior, but on generated assumptions.
If they're not regularly checked against real data, or if they get treated as equivalent to real users, a distorted picture quickly emerges. Decisions can then miss the actual problem.
Challenges in practical application
Beyond the conceptual level, there are also practical challenges.
If synthetic users are meant to be used across a team, they need to be accessible and up to date for all stakeholders. Especially in large or interdisciplinary teams, this quickly becomes complex.
Possible approaches like central persona documents or using LLMs with predefined contexts sound sensible, but bring their own problems: coordination overhead, inconsistencies, and questions of how current the data is.
Cost also plays a role, especially when several people regularly access the relevant systems.
For smaller teams or exploratory projects, this can work well. As team size grows, though, the demands on structure and maintenance increase noticeably.
How I work with personas today
Today I see personas as a tool that should be used deliberately. Before I work on personas with a team, I first clarify the goal: who's going to work with them, which decisions are they meant to support, and what data is actually available? Only after that do I advise on which approach fits the available resources and challenges. What remains decisive for me:
- Is the knowledge actually being used?
- Does the content stay current?
- Is there a connection to real users?
If these points aren't in place, even the best persona won't add value. Personas can be a helpful tool for better understanding users and supporting decisions. But every approach comes with its own challenges, and my job is to guide the team to the right question before we commit to a method.
References
- Alex, S., & Robra-Bissantz, S. (2024). Implementierung von 360-Grad-Personas auf Basis von Action Design Research und Design Science Research – Ein Forschungsprojekt mit der Praxis. HMD Praxis der Wirtschaftsinformatik, 61, 802–819. https://doi.org/10.1365/s40702-023-01024-5
- Freepik. (n.d.-b). Side view of woman leaning on a wooden fence (Photo). Magnific. https://www.magnific.com/de/fotos-kostenlos/seitenansicht-der-frau-die-beim-anlehnen-auf-holzzaun-aufwirft_9468548.htm
- Kauer-Franz, M., & Franz, B. (2022). Usability und User Experience Design – Das umfassende Handbuch. Rheinwerk
- Nielsen Norman Group. (n.d.). Synthetic users. https://www.nngroup.com/articles/synthetic-users/