Complete guide: How to create personas based on data

Table of contents:

1. What is a persona?
2. What do you need personas for?
3. How do you create a persona?
4. Collecting data for perfect personas
5. Formulating and designing personas
6. How can you use personas in your daily work?
7. What are personas not able to do?
8. Conclusion: Personas are wonderful when they are made scientifically

Definition of persona

A persona is a fictional description of a person which is nonetheless realistic. Fictional and realistic – that isn’t necessarily a contradiction: Fictional means that you can’t simply choose a real member of your target group, one who you may even know. Instead, you invent a person that doesn’t exist. They combine important qualities of multiple representatives of your target group. However, you need to keep it realistic. The person you’re describing should be true to life. So the persona gets a name, an age, a gender, they get a family, a job, preferences and habits and much more.

Example of a persona

Why you should create personas

Personas are a method of making ‘the customer’ or ‘the user’ tangible and concrete. They are given a name and a face. Personas have been used for almost 20 years, mostly in the field of user experience. Gradually, more and more colleagues from marketing departments are using them too. Personas have the advantage that they put you face-to-face with the person you are conceiving, designing and developing for. This makes the end product more usable and more attractive for people to buy.

The idea of personas originates from Alan Cooper. He developed them in the 1990s and presented them in his book ‘The Inmates Are Running the Asylum’, which is still worth reading today.

How do you create a persona?

As mentioned already, personas are fictional. Hence the widespread misconception: Many people think personas are invented. But that’s not true. You determine what the persona looks like, what qualities they have, what name they have and what is important to them. But you don’t simply dream up these things off the top of your head. Nor do you make them up according to how you would like them to be. The basis for all good personas is always data – data and insights about your users. About the people you would like to develop products for. To ensure you create realistic personas, here’s a tip on how to proceed. The ideal process for creating personas is divided into two parts:

1. Data collection
2. Formulating and designing the personas themselves

Collecting data for perfect personas

Let’s start with step one, data collection. The best way to do this is as follows:

1. Market research, statistical analysis and conducting surveys
2. User interviews, focus groups, contextual inquiries
3. Collect information and generate hypotheses
4. Validate hypotheses by means of a survey

This is the model procedure and I don’t know many people who really always implement it that way. Later I’ll give you some tips on how to do this with less effort and when you can possibly do without certain things. Let’s take a look at why these five steps make sense:

1) Market research, statistical analysis, conducting surveys

This is a step you always need, and there is no excuse to skip it. In doing so, you become familiar with the background context. You learn which applications/products/systems your (potential) users are already using. You see what industry they are in and what topics are discussed there. And also what others have already learned about the users. There is a lot of information about end customers/consumers on the internet. Statista.com is a helpful resource. So are the statistical offices of Germany, Austria and Switzerland.

If you have corporate customers (business-to-business, B2B) as your target group, then you’ll have to search a little harder for information. But even then, you can find a lot of information online, such as on XING or LinkedIn forums. If you have the feeling that you still don’t know enough, then conducting your own surveys makes sense. Good surveys require experts – or expert knowledge. Simply compiling a few questions and then unleashing them on your target group is of little use. Designing questionnaires and evaluating surveys is a science in itself. So stay humble and start with very few questions if you don&’t have any experience.

2) User interviews, focus groups, contextual inquiries

Step two, on the other hand, is something that anyone can do. I even recommend involving as many people from the team as possible. After all, talking to users is always very rewarding. And, often, you learn a lot more from personal contact than just the facts, which is why conversations are worthwhile: You can see how users work or live, how they present themselves, how they react, how they work. This gives you a sense of who you’re dealing with. In interviews, you simply talk to the users. You ask them about their lives and about their work. About the area that interests you regarding your product. However, it’s best not to ask users direct questions about the product. Don’t show them anything, whether ideas, screens, presentations or prototypes. Just find out what they’re currently doing and what problems they’re having.

In focus groups, you don’t just talk to one person, you talk to five to 10. They get into conversation with each other and interesting discussions arise – and sometimes even useful ideas. Contextual inquiry essentially means visiting someone in their home. You visit the users at home or at their workplace and let them show you how they currently work. For example, if you’re planning an app for recording employees’ time, you’ll go to the office and ask them to show you how they’re currently recording their working time. You can see if they’re using another app or web form, and you can see if they’re using notes, post-its or other things. Contextual inquiries of this kind have many names or variants, and are often referred to as contextual analysis, participatory observation, or ethnographic study.

3) Collect information and generate hypotheses

In the next step, you will see your insights from steps one and two. You are looking for similarities and differences between the users. Does a clear picture emerge? Can you explain all the differences you found? Using this as a starting point, you formulate hypotheses. In our time recording example, you could put forward the following hypotheses, for example:

• Employees find it annoying to write down their hours.
• Currently, they only record their hours once a week.
• During the work day, they are too busy to record their hours.
• They are good with Excel and Word.
• They are open to technical innovations.
• Employees spend 90% of their time working in the office.

You can also make hypotheses that contradict each other. After all, it is possible that you may have observed several different target groups. For example, one target group that welcomes technical innovations and one that is very sceptical. Your hypotheses should be based on observations or insights. If you have just used your intuition to formulate hypotheses, then the next step is particularly important for you:

4) Test hypotheses

Ideally, you should put your hypotheses to the test at the end. To do so, you would carry out a quantitative study – such as a survey. This means that you test the findings from your qualitative studies using quantitative methods. For this you need a large number of people – anything under 100 participants will hardly produce statistically meaningful results. The easiest way to recruit these participants is through your website. In this case, you have to keep in mind that you are only attracting customers and people who are already interested in you – not people who don’t know your website.

If time or cost constraints prevent you from conducting a study, at least test your hypotheses by talking to potential users again. Now you can ask much more specific questions than with your first interviews, when you were still in collector mode. As far as the process is concerned, it would also be conceivable to start with your own surveys in step two, prepare the results and identify the most important groups with a cluster analysis. Then you conduct qualitative interviews to check the content from the survey findings.

Formulating & designing personas – including a downloadable persona template

The process of creating personas holds a big opportunity: Your personas will be much, much better if you don’t create them on your own. You can achieve better results when working in a team. Especially if you manage to bring everyone together who is knowledgeable about your existing customers and about your desired customers. So not only the bosses, product managers, graphic designers and programmers, for example, but also your colleagues from sales and support departments. Because it is usually them who speak directly to customers and who often know much more about their problems and wishes than anyone else.

Most of the time, you will realise: Your colleagues’ experiences are all different, and they only start to form a complete picture when considered all together. This is why you should try to gather colleagues from all relevant departments for a persona workshop. Ideally, you would work with five to 15 participants.

Creating personas in a workshop

Your task is to brief all participants so that everybody is on the same page. To do this, you present the results of your interviews and surveys. It is best to decide together as a group which personas you are going to create. For example, one persona for an ambitious aspiring young manager, one for a long-standing employee and one for someone at the start of their career, for whom starting a family is currently more important than their work. Following that, you can divide the work up and, as long as you have enough workshop participants, delegate it to smaller groups. There should be at least three colleagues in each group.

After an hour or two, everybody comes together again and each group introduces their personas. Then you have a group discussion about whether the characteristics are appropriate as they stand, making any necessary changes. At the end of the day, everybody will be exhausted, but the bulk of the work will have been done. After that, your task simply consists of presenting the results in an appealing way. A few tips on that later.

What characteristics should you include?

The question often arises as to which characteristics should be included for one’s persona. This goes hand in hand with the question of what criteria to use to differentiate between different characteristics. Unfortunately, there is no correct answer, since it differs significantly from project to project. What’s important is that you aren’t working on too many characteristics. Remember: Personas are a tool that can help give you a quick overview. The more information you include, the more confusing it becomes. The minimum set is made up as follows:

• Name
• Age
• Profession
• Marital status/living situation
• Place of residence
• Interests/hobbies
• Values/beliefs

Depending on the product, you might add things such as:

• Technical ability/software skills
• Media consumption
• Purchasing behaviour/habits
• Preferred brands
• Position in social circles/within the company (advisor on certain subjects, influencer)

If you are targeting corporate customers: Even in the field of B2B, personas need to have a personal life. After all, in the end it is always people who have the final say on a deal, even if they are doing it on behalf of a global corporation. Every persona also needs a photo. This helps enormously to aid your memory and develop real empathy. To collectively choose one is not always easy, but in my experience even photos that have been chosen in only a few minutes work well.

Our human brain works in such a way that we assign the characteristics to the appearance of a person, even if at first we might think that a photo doesn’t fit. It’s best if the photo doesn’t give the impression of a supermodel. A snapshot usually gives a more realistic and accessible impression. Photos of celebrities are unsuitable – everyone in the team will associate them with something different. When organising a persona workshop, it’s important that everybody in the group knows what characteristics they are defining. You either agree on this together or you specify it from the beginning – depending on how much time you take for the workshop.

How do you represent personas graphically?

The better your personas look, the more likely they are to be used. That doesn’t mean you need to create a masterpiece of graphic design. You just need to make an effort to design the personas in a clear and appealing way. The more you work with images, symbols or graphics, the better. After all, the personas should be clear, memorable and pleasant. There is no universal template that suits all needs. However, you can download this template that can serve as a starting point for the design of your persona:

What are good personas called?

I always choose realistic names. They should be serious in any case. Names such as ‘Debbie Downer’ or ‘Silly Lily’ are off limits. Personas are the people you’re working for in the development process – and you don’t want to work for people you don’t like, do you? I personally don’t like ‘Ollie Online’, for example, because descriptive names like that always seem a bit silly, which makes acceptance within the company difficult. Some colleagues argue that this makes it easier to remember the names. But I’ve found that my colleagues can remember even non-descriptive names in no time at all. I would always use first and last names, simply since it makes the person even more realistic.

How many personas do I need?

Keep your persona workforce to a minimum. Three to four people is good; if you can manage with just two, all the better. It should never be more than five. Personas are there to make your various target groups clearer, and that becomes more difficult the more personas you have. Even numerous large online shops with an extremely wide product range and a correspondingly diverse customer base only work with three to four personas.

What makes a good persona?

For good personas, there are really only two criteria:

1. They must come to life in the reader’s mind when they read their description.
2. They need to be realistic.

You can easily verify the first point: Does the description give the impression that this person could exist? Do you feel like you would recognise them if you met them on the street? Give the persona to an impartial colleague and ask them these questions if you’re not sure yourself. It’s a little harder when it comes to the second point. In order to test this, you need to go over the findings from the user and market research again. Do they match the descriptions of your personas? Again, check carefully that you haven’t created a wishful persona.

I recently came across a persona in a project that appeared to be very thoroughly conceived and perfectly designed. Willi Winkler, 44, employee in the purchasing department of a large company, two children. As a tech enthusiast, friends frequently ask him for advice when a device isn’t working or when they are looking to buy a new one. Doesn’t use any wireless music speaker system such as Sonos, because he doesn’t know how it works. I was convinced up until the last point. But then I had to ask myself: Interested in tech, but not able, or rather willing, to understand how a wireless speaker system works? Even without user research, I would say: Such people may well exist. But most likely not very many.

That is why Willi Winkler is not a very good persona. Sorry, Willi. You should also avoid modelling personas on existing customers or real people. After all, personas are meant to combine the qualities of several people in one. This is also problematic because some of your colleagues may not know the people in question that well, and may not have the same opinion about them. In other words, they will not have the same persona in mind when talking about them. If the persona is truly fictional, everybody has the same information. Besides, it can be very embarrassing if the person in question gets to see ‘their’ persona – that feels a lot like stalking and you should avoid it at all costs.

Creating personas with minimal effort

What do you do if you’re the only one in your team responsible for the UX? And what if you don’t manage to get a big enough budget or time from your colleagues to put on a persona workshop? Don’t give up. Of the four steps of data collection that I explained above, focus on number one (market research, statistical analysis, conducting surveys) and two (user interviews, focus groups, contextual inquiries). You can perform these tasks alone in a few days, which should at least give you a solid basis. When the colleagues start working with your personas, they will learn to appreciate them. Then next time, you can approach the whole process more professionally and get better and better.

How can you use personas in your daily work?

Your personas are ready. Congratulations! But what now? Unfortunately, they don’t come up with how to be useful by themselves – they depend on your support. You have to introduce the personas to the team. If your colleagues have taken part in the persona workshop, this won’t be so difficult. In that case, it is enough to send an announcement by e-mail, intranet, slack or equivalent and then put up posters. If not everyone was able to make it, you should organise a round of introductions.

The simplest way is a short presentation. At the end of the presentation, you can then distribute attractive printouts and/or posters that everyone can hang up at their own workstation. There are even teams that print life-size posters. On the front you see a full body portrait of the persona, on the back there is a description of the persona. Or they are holding a blackboard with their own description in their hands. I even know a company that has set up a room in the office for each persona – designed the way the persona’s living room or kitchen would look. There you can sit down and literally immerse yourself in the persona’s everyday life.

In team discussions about design, prioritising new functions or fixing bugs, you should refer to the persona as often as possible in the future. ‘Would Willi Winkler understand this function?’ or ‘Does it make things easier for Andrea Aufmann?’ You will find: After a few weeks, even those colleagues who were previously critical of personas will start referring to your personas. Finally, it is important that you take care of your personas from time to time. At least every one or two years you should carefully check whether they are still up to date. User habits may have changed. Your product’s target groups or the market situation may also have changed. The faster your industry moves, the more often you will need to update your personas.

What are personas not able to do?

If personas are well prepared and based on real data, they can be very effective. However, there are also some things they are simply not suitable for. The most important principle is: Personas are no substitute for exact descriptions of your target groups. That’s a mistake some marketing people make: They think that you are simply representing each target group with a persona. However, this is not the case. Target groups are abstract and comprehensive – and you usually have significantly more target groups than personas. Target groups are especially important for the development of the product vision, for strategic planning. Personas, on the other hand, are particularly helpful in daily work, precisely because they are concrete.

Of course, no matter how good your personas are, they cannot replace usability tests. Personas can help you to improve conception, design and implementation. They help you better define your needs, previous experiences, problems and requirements. But you can only find out whether you have succeeded in implementing this if you conduct tests with real people from your target group. Personas are very helpful in the search for test persons. After all, personas describe exactly who you need for your user tests.

Conclusion: Personas are wonderful when they are made scientifically

To conceive, design or develop without personas is to stumble in the dark. The better you know your users, the better your product will be. That’s why personas help everyone in the team to focus their work on the users. To make it easy for yourself to get started, download the persona template for free:

User-centred development benefits enormously from personas. You just need to make sure your personas are genuinely realistic. To invent personas means to live in the desired country. Then you can develop very well for these personas – but if they are unrealistic, real people won’t understand/use/buy your product either. But if data and real knowledge of your users form the core, then there is every reason to develop personas for your current project – or to update the existing ones.