Information about users supplied by stakeholders and subject matter experts (SMEs)
Market research data such as focus groups and surveys
Market-segmentation models
Data gathered from literature reviews and previous studies
However, none of this supplemental data can take the place of direct user interviews and observation. Almost every aspect of a well-developed persona can be traced back to a user statement or behavior.
Personas are represented as individual people
Personas are user models that are represented as specific, individual human beings.
They are not actual people but are synthesized directly from observations of real people. One of the key elements that allow personas to be successful as user models is that they are personifications.2 This is appropriate and effective because of the unique aspects of personas as user models: They engage the empathy of the design and development towards the human target of the design.
Empathy is critical for the designers, who will be making their decisions for design frameworks and details based on both the cognitive and emotional dimensions of the persona, as typified by the persona’s goals. (We will discuss the important connections between goals, behaviors, and personas later in this chapter.) However, the power of empathy should not be quickly discounted for other team members. Not only do personas help make our design solutions better at serving real user needs, but they also make these solutions more compelling to stakeholders. When personas have been carefully and appropriately crafted, stakeholders and engineers begin to think about them as if they are real human beings and become much more interested in creating a product that will give this person a satisfying experience.
We’re all aware of the power of fictional characters in books, movies, and television programs to engage viewers. Jonathan Grudin and John Pruitt have discussed how this can relate to interaction design.3 They note, as well, the power of method
09_084113 ch05.qxp 4/3/07 6:02 PM Page 82
82
Part I: Understanding Goal-Directed Design
acting as a tool that actors use to understand and portray realistic characters. In fact, the process of creating personas from user observation, and then imagining and developing scenarios from the perspective of these personas, is, in many ways, analogous to method acting. (We’ve even heard our Goal-Directed use of personas referred to as the Stanislavsky Method of interaction design.)
Personas represent groups of users
Although personas are depicted as specific individuals, because they function as archetypes, they represent a class or type of user of a specific interactive product. A persona encapsulates a distinct set of behavior patterns regarding the use of a particular product (or analogous activities if a product does not yet exist), which are identified through the analysis of interview data, and supported by supplemental quantitative data as appropriate. These patterns, along with specific motivations or goals, define our personas. Personas are also sometimes referred to as composite user archetypes because personas are in a sense composites assembled by grouping related usage patterns observed across individuals in similar roles during the Research phase.4
Personas and reuse
Organizations with more than one product often want to reuse the same personas.
However, to be effective, personas must be context specific — they should be focused on the behaviors and goals related to the specific domain of a particular product. Personas, because they are constructed from specific observations of users interacting in specific contexts, cannot easily be reused across products even when those products form a closely linked suite.5
For a set of personas to be an effective design tool for multiple products, the personas must be based upon research concerning the usage contexts for all of these products.
In addition to broadening the scope of the research, an even larger challenge is to identify manageable and coherent sets of behavior patterns across all of the contexts.
Clearly, it is a fallacy to believe that just because two users exhibit similar behaviors in regard to one product, that those two users would behave similarly with respect to a different product. Thus, as focus expands to encompass more and more products, it becomes increasingly difficult to create a concise and coherent set of personas that represents the diversity of real-world users. We’ve found that, in most cases, personas should be researched and developed individually for different products.
Archetypes versus stereotypes
Don’t confuse persona archetypes with stereotypes. Stereotypes are, in most respects, the antithesis of well-developed personas. Stereotypes represent designer
09_084113 ch05.qxp 4/3/07 6:02 PM Page 83
Chapter 5: Modeling Users: Personas and Goals
83
or researcher biases and assumptions, rather than factual data. Personas developed by drawing on inadequate research (or synthesized with insufficient empathy and sensitivity to interview subjects) run the risk of degrading to stereotypical caricatures. Personas must be developed and treated with dignity and respect for the people whom they represent. If the designer doesn’t respect his personas, nobody else will either.
Personas also bring issues of social and political consciousness to the forefront.6
Because personas provide a precise design target and also serve as a communication tool to the development team, the designer much choose particular demographic characteristics with care. Ideally, persona demographics should be a composite reflection of what researchers have observed in the interview population, modulated by broad market research. Personas should be typical and believable, but not stereotypical. If the data is not conclusive or the characteristic is not important to the design or its acceptance, we prefer to err on the side of gender, ethnic, age, and geographic diversity.
Personas explore ranges of behavior
The target market for a product describes demographics as well as lifestyles and sometimes job roles. What it does not describe are the ranges of different behaviors exhibited by members of that target market regarding the product and related situations. Ranges are distinct from averages: Personas do not seek to establish an average user, but rather to express exemplary or definitive behaviors within these identified ranges.
Because products must accommodate ranges of user behavior, attitudes and aptitudes, designers must identify a persona set associated with any given product.
Multiple personas carve up ranges of behavior into discrete clusters. Different personas represent different correlated behavior patterns. These correlations are arrived at through analyzing research data. This process of identifying behaviors is discussed in greater detail later in this chapter.
Personas must have motivations
All humans have motivations that drive their behaviors; some are obvious, and many are subtle. It is critical that personas capture these motivations in the form of goals. The goals we enumerate for our personas (discussed at length later in this chapter) are shorthand notation for motivations that not only point at specific usage patterns but also provide a reason why those behaviors exist. Understanding
09_084113 ch05.qxp 4/3/07 6:02 PM Page 84
84
Part I: Understanding Goal-Directed Design
why a user performs certain tasks gives designers great power to improve or even eliminate those tasks yet still accomplish the same goals.
Personas can also represent nonusers
While the users and potential users of a product should always be an interaction designer’s primary concern, it is sometimes useful to represent the needs and goals of people who do not use the product but nevertheless must be considered in the design process. For example, it is commonly the case with enterprise software (and children’s toys) that the person who purchases the product is not the same person who uses it. In these cases, it may be useful to create one or more customer personas, distinct from the set of user personas. Of course, these should also be base
d upon behavior patterns observed through ethnographic research, just as user personas are.
Similarly, for many medical products, patients do not directly interact with the user interface, but they have motivations and objectives that may be very different than the clinician using the product. Creating a served persona to represent patients’
needs can be useful in these cases. We discuss served and customer personas in greater depth later in this chapter.
Personas and other user models
There a number of other user models commonly employed in the design of interactive products, including user roles, user profiles, and market segments . These are similar to personas in that they seek to describe users and their relationship to a product. However, personas and the methods by which they are created and employed as a design tool differ significantly from these in several key aspects.
User roles
A user role or role model, as defined by Larry Constantine, is an abstraction, a defined relationship between a class of users and their problems, including needs, interests, expectations, and patterns of behavior.7 As abstractions (generally taking the form of a list of attributes), they are not imagined as people, and do not typically attempt to convey broader human motivations and contexts.
Holtzblatt and Beyer’s use of roles in consolidated flow, cultural, physical, and sequence models is similar in that it attempts to abstract various attributes and relationships abstracted from the people possessing them.8
09_084113 ch05.qxp 4/3/07 6:02 PM Page 85
Chapter 5: Modeling Users: Personas and Goals
85
We find these methods limiting for several reasons:
It is more difficult to clearly communicate human behaviors and relationships in the abstract, isolated from people who possess them. The human power of empathy cannot easily be brought to bear on abstract classes of people.
Both methods focus on tasks almost exclusively and neglect the use of goals as an organizing principle for design thinking and synthesis.
Holtzblatt and Beyer’s consolidated models, although useful and encyclopedic in scope, are difficult to bring together as a coherent tool for developing, communicating, and measuring design decisions.
Personas address each of these problems. Well-developed personas describe the same type of behaviors and relationships that user roles do, but express them in terms of goals and examples in narrative. This makes it possible for designers and stakeholders to understand the implications of design decisions in human terms.
Describing a persona’s goals provides context and structure for tasks, incorporating how culture and workflow influence behavior.
In addition, focusing on user roles rather than on more complex behavior patterns can oversimplify important distinctions and similarities between users. It is possible to create a persona that represents the needs of several user roles (for example, in designing a mobile phone, a traveling salesperson might also represent the needs of a busy executive who’s always on the road), and it is also possible that there are several people in the same role who think and act differently (perhaps a procurement planner in the chemical industry thinks about her job very differently from a procurement planner in the consumer electronics industry). In consumer domains, roles are next to useless. If you’re designing a Web site for a car company,
“car buyer” is meaningless as a design tool — different people approach the task in very different manners.
In general, personas provide a more holistic model of users and their contexts, where many other models seek to be more reductive. Personas can certainly be used in combination with these other modeling techniques, and as we’ll discuss at the end of the chapter, some other models make extremely useful complements to personas.
Personas versus user profiles
Many usability practitioners use the terms persona and user profile synonymously. There is no problem with this if the profile is truly generated from ethnographic data and encapsulates the depth of information the authors have described.
Unfortunately, all too often, the authors have seen user profiles that reflect Webster’s definition of profile as a “brief biographical sketch.” In other words, user profiles often consist of a name and a picture attached to a brief, mostly demographic
09_084113 ch05.qxp 4/3/07 6:02 PM Page 86
86
Part I: Understanding Goal-Directed Design
description, along with a short, fictional paragraph describing the kind of car this person drives, how many kids he has, where he lives, and what he does for a living.
This kind of user profile is likely to be based on a stereotype and is not useful as a design tool. Although we give our personas names, and sometimes even cars and family members, these are employed sparingly as narrative tools to help better communicate the real underlying data. Supporting fictional detail plays only the most minor part in persona creation and is used just enough to make the persona come to life in the minds of the designers and the product team.
Personas versus market segments
Marketing professionals may be familiar with a process similar to persona development because it shares some process similarities with market definition. The main difference between market segments and design personas is that the former are based on demographics, distribution channels, and purchasing behavior, whereas the latter are based on usage behavior and motivations. The two are not the same and don’t serve the same purpose. Marketing personas shed light on the sales process, whereas design personas shed light on the product definition and development process.
However, market segments play a role in persona development. They can help determine the demographic range within which to frame the persona hypothesis (see Chapter 4). Personas are segmented along ranges of usage behavior, not demographics or buying behavior, so there is seldom a one-to-one mapping of market segments to personas. Rather, market segments can act as an initial filter to limit the scope of interviews to people within target markets (see Figure 5-3). Also, we typically use the prioritization of personas as a way to make strategic product definition decisions (see the discussion of persona types later in this chapter). These decisions should incorporate market intelligence; an understanding of the relationship between user personas and market segments can be an important consideration here.
When rigorous personas aren’t possible:
Provisional personas
Although it is highly desirable that personas be based upon detailed qualitative data, there are some occasions when there simply is not enough time, resources, or corporate buy-in to perform the necessary fieldwork. In these cases, provisional personas (or, as Don Norman refers to them, “ad hoc” personas) can be useful rhetorical tools to clearly communicate assumptions about who the important users are and what they need, and to enforce rigorous thinking about serving specific user needs (even if these needs are not validated).
09_084113 ch05.qxp 4/3/07 6:02 PM Page 87
Chapter 5: Modeling Users: Personas and Goals
87
Market segments
Pool of interviewees
Behavior patterns emerge
1
Kate & Sara are
in segment 1
2
Bob overlaps 2 & 3
3
Ann is in segment 3
Figure 5-3 Personas versus market segments. Market segments can be used in the Research phase to limit the range of personas to target markets. However, there is seldom a one-to-one mapping between market segments and personas.
Provisional personas are structured similarly to real personas but rely on available data and designer best guesses about behaviors, motivations, and goals. They are typically based on a combination of stakeholder and subject matter expert knowledge of users (when available), as well as what is understood about users from existing market data. Provisional personas are, in fact, a more fleshed-out persona hypothesis (as described in Chapter 4).
Our exp
erience is that, regardless of a lack of research, using provisional personas yields better results than no user models at all. Like real personas, provisional personas can help focus the product team and build consensus around product features and behaviors. There are, however, caveats: Provisional personas are called this because they should be recognized as stand-ins for personas based on definitive qualitative data. While provisional personas may help focus your design and product team, if you do not have data to back up your assumptions you may:
Focus on the wrong design target
Focus on the right target, but miss key behaviors that could differentiate your product
09_084113 ch05.qxp 4/3/07 6:02 PM Page 88
88
Part I: Understanding Goal-Directed Design
Have a difficult time getting buy-in from individuals and groups who did not participate in their creation
Discredit the value of personas, causing your organization to reject the use of personas in the long term
If you are using provisional personas, it’s important to:
Clearly label and explain them as such
Represent them visually with sketches, not photos, to reinforce their provisional nature
Try to make use of as much existing data as possible (market surveys, domain research, subject matter experts, field studies, or personas for similar products)
Document what data was used and what assumptions were made
Steer clear of stereotypes (more difficult to do without field data)
Focus on behaviors and motivations, not demographics
Goals
If personas provide the context for sets of observed behaviors, goals are the drivers behind those behaviors. A persona without goals can still serve as a useful communication tool, but it lacks utility as a design tool. User goals serve as a lens through which designers must consider the functions of a product. The function and behavior of the product must address goals via tasks — typically, as few tasks as absolutely necessary. Remember, tasks are only a means to an end; goals are that end.
Alan Cooper, Robert Reinmann, David Cronin - About Face 3- The Essentials of Interaction Design (pdf) Page 15