HCI VISTAS, VOLUME-III, 2007-2008
Human Computer Interaction
Ten Steps to Personas
Guest Author: Dr.Lene Nielsen
About HCI VISTAS
INDEX
Article INS-24./July 2007
<< >>

Having worked with personas before the method ever came to be known as personas there are, from my research and practical experience, three important areas that have to be considered: the data material, engagement in the personas descriptions, and buy-in from the organization which is part of the development process whether it is redesign or a development from scratch. This is the rationale behind my development of 10 steps to personas, an attempt to cover the entire process from initial data gathering to ongoing development.
In the following I will briefly outline the 10 steps. Any project that uses personas does not necessarily need to follow all 10 steps as long as the responsible party knows the consequences of skipping a step.
Step 1: Finding the Users
The initial step is to get hold of as much knowledge of the users as possible. The data can originate from several sources: interviews, observations, second hand information, questionnaires, reports, cultural probes etc. In my experience large companies have often a lot of information about the users, reports from marketing, call centers etc. these can in some extend substitute real life meetings with users, but they also create problems as they do not focus on the subject that the project is about. This might become visible in the next step.
Step 2: Building a Hypothesis
Working with personas is focusing on users in a certain context which originates from the project. Often companies have a certain way of talking about their users that does not take into consideration the different context the users might be in when using a website or a system. In a recent project for a national Danish authority concerning redesign of a web portal business reports to different governmental authorities, the national authority had a tradition for dividing Danish businesses into categories of size and trades. From interviews with staff in the call center and reading of several a hypothesis was formed.
The former division of businesses did not make sense in this project, as it does not matter which trade the one who has to do the reporting is in, what matters it seemed is how big the company is, and whether the persons who reports is employed within the company or a consultant of some sort. There had been a number of surveys performed, but none of these had this division in mind and had to be reread from the new perspective.
Step 3: Verification
In my experience the most difficult task in persona project are 'how to cut the cake' - coming from data to a decision of how may personas descriptions to include. This takes several of the 10 steps and involves more than a group of consultants or project members to just hand over some descriptions.
In 'Verification' the focus is on finding data that supports the initial patterns and at the same time supports the personas descriptions and the scenario writing. The persona method requires a certain kind of information that can help generate engagement in the descriptions and support scenario writing e.g. what does the users like ad dislike, what are their values, what are their attitudes towards the system/site, in what conditions will they use the system/site? When these data are collected do they then support or go against the initial data.
Step 4: Finding Patterns
My inspiration in this and the previous step originates from making sense of data in qualitative inquiries. The way you know that you are on the right track is when others can follow your argumentation and others can come to the same result. Therefore it is of importance to show the categorization to other team members, project partners etc.

Step 5: Constructing Personas
A crucial step is what to include in a personas description and how to avoid creating stereotypes. I have quite often seen personas descriptions that either depict super humans or stereotypes that is difficult to engage in. In this phase you must remember that the whole purpose of personas is not to describe users as such, but to create solutions that use the needs of the persona as a starting point.
Drawing on knowledge from fictional writing of characters 5 areas need to be present in the description, not mentioned specifically, but possible for the reader to deduct from the description.
- Body (a photo or a description of how the person looks creates a feeling of the person as a human being, posture and clothing tells a lot about the person)
-Psyche (we all have an overall attitude towards life and our surroundings which also influence the way we meet technology e.g. is the persona introvert or extrovert)
-Background (we all have a social background, education, upbringing which influence our abilities, attitudes and understanding of the world)
-Emotions and attitudes towards technology and the domain designed for
-Personal traits. This one is tricky, in fictional writing there is a distinction between flat characters and rounded characters. The flat character is characterized by having only one character trait which is reflected in all actions the character does and creates a highly predictable character close to the stereotype. The flat character is difficult to engage in. The rounded character has more than one character trait, is not predictable and easier to engage in. When writing personas is becomes essential to avoid the stereotype and create descriptions that the project team members can engage in. Therefore it is advisable to look for information that repeats the same trait. In a case I had, the persona to be described liked to feel in control, from this the team members writing the description made her work for the tax authorities, this came to reflect her attitude to life, she became overweight and with few friends. For them the information of being in control created a negative attitude towards the persona that was repeated in all information.
The fifth step is also a step that can ensure that can enhance buy-in. In my experience it is few organizations who allow for team members to be part of the writing process instead they use consultants or the usability department to write the descriptions. The personas method should rather be perceived as a process where everybody should understand how the descriptions came about and what they can be used for. If you allow different team members to be part of the writing process they feel ownership for the personas. They can be rewritten be a single person to ensure homogeneity in writing and presentation, but it pays off later to include more in the writing process or as we did in a project, to let the participant choose the pictures for the personas.
Click to view high resolution image
Click to view high resolution image.
I am fully aware that not everybody can be part of the process, newcomers arrive, a row of companies might be involved, but if the personas are not disseminated to participants they are not worth anything. It is not only the personas that need to be distributed to everybody, but also the data behind (the foundation document as Grudin, Pruitt, Adlin calls it) and not least how and for what you are to use the personas. Many projects forget to inform and teach developers and designers how to use the personas, how to think in scenarios or how to use them in the use-cases.
Step 6: Defining Situations
As mentioned earlier the real purpose of the personas is to create scenarios from the descriptions. This step is a preparation for the scenarios where it is described in which situations the persona will use the system/site or which needs the persona has that will lead to a use situation. Each need or situation is the beginning for a scenario.
Step 7: Validation and Buy-in
To ensure that all participants agree on the descriptions and the situations two strategies can be followed: ask everybody their opinion and let them participate in the process. Often the persona method is viewed as mean for communication users to developers and others, but is as much a process that ensures a user-centered development. Having a process view helps create sessions where as many stakeholders as possible can be involved in the developing the personas and in using them for design.
Step 8: Dissemination of Knowledge
I am fully aware that not everybody can be part of the process, newcomers arrive, a row of companies might be involved, but id the personas are not disseminated to participants they are not worth anything. It is not only the personas that needs to be distributes to everybody, but also the data behind (the foundation document as Grudin, Pruitt, Adlin calls it) and not least how and for what you are to use the personas. Many projects forget to inform and teach developers and designers how to use the personas, how to think in scenarios or how to use them in the use-cases.

Step 9: Creating Scenarios
As mentioned earlier, personas are nothing in themselves, it is when a persona enter a scenario they prove to be valuable. A scenario is like a story, it has a main character (the persona) a setting (somewhere the action takes place), it has a goal (what the persona wants to achieve), it has actions that lead to the goal (interactions with the system/site/device), and -not least- it has obstacles that blocks the way to the goal. I have seen quite a number of what I call happy scenarios, where a device solves all problems. Try to read this description of Mrs Tahira Khan and how she overcomes her diabetes and you will see what I mean. It is not a very realistic or convincing example that a 65-year old woman, who recently traveled to UK, who has undiscovered diabetes, hardly any understanding of the English language and relatively poor literacy in her own language overcome her diabetes with an electronic device.
Step 10: Ongoing Development
Lastly I recommend updating information on the personas. This can be done if user tests suddenly show new results or if something changes in the personas environments. It is crucial that not everybody is able to change the information, but knows whom to contact. I often recommend having a personas ambassador, who looks into the descriptions now and then, and who project participants can contact if they find irregularities in the descriptions. And as Adlin and Pruitt recommend in 'The Personas Lifecycle' to let the personas die, when they have outlived their purpose.

Lene Nielsen wrote her Ph.D. thesis "Engaging Personas and Narrative Scenarios" in 2004.
She is part-time assistant professor at Center of Applied ICT, at Copenhagen Business School
and part-time usability consultant at Snitker & Co.
The illustration for the article is a poster made by Snitker & Co
Contact information:ln.caict@cbs.dk

Feedback at
Contribute Articles To
Copyright 2007-2008