9 Tips to keep Personas / User Types alive (and useful)

Why do Personas have such a short half-life?

Most “Personas” developed within and for companies reach their high point at the moment they’re presented. After that, they usually die off quickly. At best they might offer a quick but useful glimpse into the landscape of your users. But what to do with them next? How can they solve your or other teams’ problems? How representative are they? From whose perspective were they designed? And why do they have these (silly) names?

There are steps you can take to increase the value of your  Personas User Types* and extend their use beyond the first presentation. If done properly, you can make User Types an indispensable component in shaping strategy and developing a user-centered approach.

* We suggest using the term “User Types” instead of “Personas,” as the latter term often carries negative associations from past experiences. The term “Segments,” which is especially used in marketing departments, lacks the necessary human and evocative element.

1. User types need  a proper name and granularity

Avoid giving User Types real names and demographic data, like “Anna, 2 kids, 37 years old,” as this might be more damaging than helpful. You need to find the right “granularity,” or level of abstraction, useful for your company and project. And User Types should represent useful data, not uncontrollable interpretations. Everybody in your company likely knows an “Anna” and so may project their own (different) associations and prejudices on them.

Instead, consider giving User Types a descriptive name that reflects a pain, or an identifier that makes them clearly distinguishable. Furthermore, use memorable names that raise curiosity and are sticky, so that in three months’ time everybody still knows who you’re talking about. Examples: “Solo fire fighter,” “Picky bargain hunter,” “Informed fan boy.”

Also: consider using avatars instead of real images. As with real names, real images may invoke associations that you cannot control. And photos can cause problems around fulfilling diversity requirements.

2. Make user types accessible

Not everybody will have attended your User Types presentation. Not everybody will know where to find the User Types afterwards. And, at least since Covid, it’s no longer sufficient to post them on the office walls or in the “war room.”

We recommend giving User Types a dedicated place in your intranet, assigning an owner, and sharing the link with everybody.

The easier they are to find or to stumble upon, and the more self-explanatory they are, the more likely people will ask for further details, or contribute to them.

3. Use their names, mention them often

User Types allow you to stop referring to “the user.” Whenever possible, talk about users in a differentiated way and name specific User Types whenever possible, e.g. in presentations, weekly stand-ups, onboardings, and briefings.

Clearly communicate that there is something important to know, and let them know where to find more details. Be prepared to answer questions, or to help.

It will take a while for some to accept the added layer of information, but once they see that the gained precision helps decision making, they will likely even ask for them.

4. Demonstrate applicability

Many colleagues may not see the value of User Types because they don’t know how to derive actionable insights from them (“yet another UX presentation”). A User Types poster will never cover all the valuable aspects, and everybody has different questions about them. So it’s highly important to show how to draw interesting conclusions, and be ready to dig deeper on request.

Example: The “Nerdy decision prep” will never order directly, but will go to the CFO. However, he or she will likely prescribe what to buy – so, showing them a direct-order button will not help, while giving them a referrer link to send to the CFO might.

5. Transparency

User Types are political in nature, if only because you can prioritize actions through them. If somebody dislikes them, they will not use them, will question them, or – worst case – set up their own competing ones. The more open and transparent you are about the (data) base for the User Types, their benefits and limitations, the more credible they will be. This includes:

  • When and from whom were they developed?
  • What was the research angle and who paid for it?
  • How representative are they?
  • How much of the user base are they covering?
  • How can they be extended or contributed to?

6. Updating

To make User Types useful over time, they need to be updated, complemented, and extended. Some older User Types will need to be cycled out once they’ve become outdated. User Types should be continuously reinvigorated through ongoing research.

Use them to recruit study participants and update them through other data sources (see slide “Data & Analytics”). In this way, they’ll include other teams and perspectives and gain contour, detail, and credibility.

And of course, these updates (e.g. new survey, new UXLab) can and should be communicated in weekly stand-ups, etc.

7. Show Business Value

Different User Types have different needs, goals, behaviors, resources, etc. But they also have different value for and impact on your organization. By showing the potential value creation of each User Type, you can increase the interest in them from teams that otherwise don’t look much at the user level.

Typical parameters to communicate are: Share (%), acquisition costs, Customer Lifetime Value, Multiplier-Value.

These factorings typically also assist with prioritizing of actions and roadmapping.

8. Connect with Customer Journey Maps

Just as “the user” does not exist, the same is true with “the customer journey” or “the customer life cycle.” Sometimes you have no other choice, and a single CJM is better than none, but differentiating between types of users can be a significant improvement, e.g. when looking at emotional peaks, specific touch points, or exits.

And having multi-type CJMs helps keep User Types in the loop. The same applies to JTBD or mental models.

9. Connect with Data & Analytics

Likely the most powerful step, but also the most challenging one, is to connect user types with analytics and other quantitative data. We would even state that any qualitatively developed User Types will die early unless / until they are backed through analytics.

Basically this requires identifying and mapping User Types to individual sessions, in real time or ad-hoc. In real time, you might map them to behavior patterns or signifying activities (e.g. visit of a specific webpage, use of a specific contact channel). You can also use micro-surveys to find out which category they (likely) fall into (given their consent). All this is obviously easier with existing customers and diverse data “traces,” compared to new customers / unidentified users.

However, once you are there, you can create extremely powerful analyses and high-resolution dashboards, that are able to reflect your user types, instead of falling back to “the user“ or another competing perspective.

Quick update: If you like to have the ChatGPT summary:

  1. Give User Types appropriate names and granularity, avoiding real names and demographic data. Focus on descriptive, memorable, and distinguishable identifiers.
  2. Make User Types easily accessible by providing a dedicated space on your company intranet and assigning an owner to maintain them.
  3. Frequently mention User Types in conversations, meetings, and presentations to familiarize your team with these concepts and encourage their use.
  4. Demonstrate the practical application of User Types by illustrating how they can generate actionable insights and support decision-making processes.
  5. Maintain transparency regarding the development, limitations, and representation of User Types to establish credibility and prevent skepticism.
  6. Keep User Types up-to-date by incorporating ongoing research and input from various teams and perspectives, cycling out obsolete types as needed.
  7. Show the business value of User Types by highlighting the potential value creation, acquisition costs, customer lifetime value, and multiplier value for each type.
  8. Connect User Types with Customer Journey Maps to capture variations in user experiences and emphasize the importance of differentiating user needs, goals, and behaviors.
  9. Integrate User Types with data and analytics to create powerful, high-resolution dashboards that provide valuable insights into user behavior and preferences, thereby enhancing the utility of User Types within your organization.

When is an MVP Good Enough?

The Minimum Viable Product (MVP) is a key concept in the Lean Startup framework. In many MVP sprints and workshops the question came up: How elaborated, worked-out, designed, complete must the MVP be to be released, to generate valid feedback? Do we have to user proper grafics, or a stockphotos enough? Do we need good copy, or can we use Lore Ipsum? Does it have to be a proper “brand” or URL, or can we use any?

I think it is a crucial question, because time for both a MVP and in a sprint is the (artificially) scarcest thing, and everything you can leave out helps you. At the same time, if necessary parts are missing, the MVP experiment may fail, not because the valua proposition is irrelevant, but because the design of the MVP is insufficient, thus creating wrong results (and in the worst case dropping the idea).

My standard answer has been: It needs to be good enough. It needs to convey the value proposition to the user, everything else can be in a minimal fashion. However, this was not always clear or convincing or satisfying. And all those famous MVP examples (inclusive the wonderful Zappos concierge MVP) are sometimes difficult to apply to the own project.

So when I was reading his amazing book “Shape”, Jordan Ellenberg is discussing the concept of topology, where a circle does not have to be perfect to be a circle. And when he showed and described one proof for the famous Pythagorean Theorem, I found a wonderful “illustration” of “good enough” in an MVP.

The big square is the same in both pictures. But it’s cut up in two different ways. In the first picture, you have four copies of your right triangle, and a square whose side has length c. In the second picture you also have four copies of the triangle, but they’re arranged differently; what’s left of the square is now tow smaller squares, one whose side has length a and one whose side has length b. The area that remains when you take four copies of the triangle out of the big square has to be the same in both pictures, which means that c2 (the area left over in the first picture) has to be the same as a2 + b2 ( the area left over in the second).” (p.23)

Just enjoy for a moment the “proof” of the four rectangles being the same for both boxes, the proof of the famous Pythagorean Theorem (I enjoyed it, and would have hoped to have that at old school times).

For me, a good enough MVP is exactly this: To show and to reveal the value of an idea and to surprise the user with a concept. To induce the “klick”, as we say in German (both mentally and physically).