Making Sense of it All: Personas and Archetypes

First ResearchOps Berlin Meetup kicked off at smava demonstrating their collaboration within the team and across methods and finished with inspiring discussions.

On 07.02., we held our first meetup at the office of the loan comparison portal smava. They were not only the host, but also allowed us to take a peek behind the scenes of their research/product department. In an insightful presentation, they laid out smava’s own approach and challenges related to this edition’s topic: “Qualitative and quantitative approaches to Archetypes and Personas”.

Archetypes? Personas? Segments?

There are many different names and purposes to describe users of your product or service. More recently, personas seem to be the go-to method in user research with different levels of fidelity — from rather superficial, hypothetical “proto-personas” to rich descriptions of specific contexts and needs of a persona based on data (e.g. these “jobs to be done personas”).

However, the exact purpose and added benefit of personas is often unclear. Especially for large corporates that have traditionally referred back to market segments in marketing and product development, the distinction between segments and personas is often blurry.

Personas were originally meant to help designers and developers empathize with the people they design for. In this context of service and product development, they are usually quite specific to an already existing or conceptualized service and product. They are also usually more qualitative — in their nature as well as regarding the data they are based on.

On the other side, marketing and market research have been using approaches to find out about user groups on a more generalizable level, using quantitative methods. This often includes making statements about broad market segments and identifying potential future customers. Classic approaches in this regard describe groups based on socio-demographic categories.

And then, there are many different approaches that fall in between and use a mix of qualitative and quantitative methods. One of these approaches is archetypes, as shown by smava at our meetup.

Mixed methods approaches

It seems that, by the growing popularity of personas, their downsides were challenged by different disciplines. This has opened an opportunity for the integration of other concepts under the umbrella term of personas. An example of this would be the critique about personas for not being generalizable or scientific. This is something, Jeff Sauro picks up in his article about how to make personas more scientific. He describes a mixed method process towards creating personas that is close to how smava came up with their archetypes:

1. Conduct qualitative interviews and observations.
2. Survey a large sample of users and/or prospects.
3. Identify segments using a statistical clustering technique.
4. Determine key variables that differentiate segments.
5. Predict segment membership using a typing tool.
6. Personify or qualify your segments

smava’s archetype development

The process smava’s research team (Frederik Niedernolte, Mehmet Aydin, Katrin Pereslavskaya) used was quite similar: An integrated mix of qualitative and quantitative methods to identify, validate and evaluate generalizable segments of their customer base. If you want to take a deeper look into it, you can follow smava’s presentation on slideshare. smava started their archetype development with qualitative insights from interviews and a dinner with customers and employees. This led to the creation of four first archetypes of loantakers based on a 2x2 matrix.

Based on the qualitative data (quotes from the research), as well as referring to known scales like the Big 5 personality traits, smava’s quantitative user researcher created a survey. The purposes of this survey was to validate the previously created archetypes and to evaluate the distribution of them in the German market. Running different statistical models, they found out that the best number of archetypes was actually three, not four. Furthermore, the survey was used to identify a smaller subset of questions to shorten the identification of archetypes in future studies. After a factor analysis, the smava researcher had found 14 items distributed among 5 factors, such as spending or planning.

The next challenge for the UX team was to communicate the three archetypes properly. Here Cansu Kerestecioglu, the graphic designer of smava, took us through her process of making sense of the data and playing with the different categories and their visualization. For example, a typical quote from the archetypes was one crucial part to display as well as using a spider chart to show the characteristics of the five different factors among an archetype. It also became obvious, that the design is not just about “beautifying” the content, but a deciding factor by itself.

Discussing and sharing experiences about personas

After a Q & A session, in the last part of our ResearchOps Berlin meetup, people split up in three groups and discussed their experiences and opinions on these questions:

1. What are your worst and best experiences when using archetypes and personas?
2. What additional value do personas provide compared to archetypes? Do we need different formats for marketing and UX?
3. How do you continuously evolve archetypes/personas, and continue to make them actionable within your teams?

Each group had a lively discussion and shared their findings in the final plenum of our meetup. You can read the notes and follow the thoughts on the last question via this article by Nikki.

If you want to be part of the next discussion, follow our meetup page. Our next meetup will be on 09.04. and RSVP’s open about 2 weeks ahead.