Today’s Personas Are Often “99% Bad”

I recently came across a blog entitled, “Are personas dead, or just not real people”. I got excited. I thought the author was going to discuss whether or not the current approach of developing personas was finally falling out of favor. I was mistaken. The author discussed whether or not personas should be real people or “made up” people.

For those unfamiliar with personas, they are defined (by Wikipedia at least) as “fictional characters created to represent the different user types that might use a site, brand, or product in a similar way.” Their use dates back to the 1990s and were first introduced as a concept for marketing developed by Angus Jenkinson. They were later adopted into interface design by Allen Cooper as described in his excellent and still applicable 1999 book The Inmates Are Running the Asylum. But the modern version of a persona seems to bear little resemblance to the original, thus my contention that the current practice of developing them is off target.

Sign saying: Information Overload

The Wikipedia definition describes them as “1 – 2 page descriptions that include behavior patterns, goals, skills, attitudes, and the environment, with a few fictional personal details to make the persona a realistic character [emphasis added].” But Cooper described them as “a precise description of our user and what he [sic] wishes to accomplish.” Many modern personas are anything but precise and do not stop at a few fictional personal details. They are full of extraneous details that contribute little informative value to the design problem.

Consider this first paragraph of the persona for “Scott,” developed for a Federal website.

“Scott works as a sales person for a business services company and attends the local community college part time, on a business course. He lives with a friend in an apartment not very far from his parents. Outside of work, he hangs out with his friends and plays in a softball pickup league.”

Any idea what website this persona might be associated with? It’s a website dedicated to cancer information. It’s difficult to realize this from Scott’s description other than perhaps the fact that he seems to exercise a bit. What difference does it make to the design of a cancer website the specific degree program he’s studying in college? Or that he goes to a community college instead of a four-year college? Or that he lives with one friend in an apartment? Or that his apartment is close to his parent’s house? The salient points, if they even are salient, are that he’s a young adult, not a health professional, and perhaps that he works out a bit (but this could be purely social and involve lots of beer). This first paragraph expands into an eight-paragraph biographical sketch about Scott, which also includes a picture, two subtitles, thirteen bullet points, and a diagram of where he fits on an overall information-seeking process.

Now consider the opening lines of two personas for the U.S. Postal Service.

“Marty is 42 year old, married with three children” and “Janet is 38-yrs old and is married with two kids, a 6-yr old son and a 9-yr old daughter.”

Does it really matter to the U.S. Post Office website how old Marty and Janet are, or the number, ages, and genders of their kids? And now that I know about Janet’s kids, I’m wondering why no one bothered to make up the composition of Marty’s family. Is that significant? Are her kid’s details somehow salient to the design problem where Marty’s are not? (Of course not.)

The U.S. Postal Service personas are also full pages, with photos, and go on to tell us more information about these made up people that may somehow be useful, though its hard to tell how. “Marty’s newer staff often ask him to do a spot-check or consult on a particular job.” (Isn’t that true of any company? And how does this affect the U.S Post Office website design?) We also learn that “[Marty] admits that he is more likely to go back to websites that are professional looking and easy to use.” (Good for Marty, but every site designer hopefully knows that professional looking and easy-to-use websites are our goals. I’d be more intrigued to learn there is a persona for a site for people who didn’t care how professional a site looked or if they could use it.)

Chuck Burgermeister, business traveler. A 100,000-mile-club-member who flew somewhere practically every week. Chuck’s vast experience with flying meant that he has little tolerance for complex, time-consuming interfaces, for interfaces that condescend to novices.

Ethan Scott, 9-year-old boy. He is travelling unescorted for the first time. Ethan wanted to play games, games, and more games.

Marie Dubios, bilingual business traveller. English was her second language. She likes to browse the shopping, as well as the entertainment selections.

Clevis McCloud, crotchety septuagenarian. An aging but still spry Texan, slightly embarrassed about the touch of arthritis in his hands. He was the only one of the four passenger personas who didn’t own a computer or know how to use one.

In 117 words, Cooper defined all the details needed. And he covered the range of population from 9 years old to a septuagenarian, computer savvy and computer illiterate, English as a second language, and even intolerant of condescending designs. There is no fluff, no unneeded detail, no extended history or background data not directly important to imagine this person and understand the task they want to do. Beyond that, more details are added noise.

Will Strunck Jr., reported by E.B. White in a famous book on writing, is quoted as saying:

“A sentence should contain no unnecessary words, a paragraph no unnecessary sentences, for the same reason that a drawing should have no unnecessary lines and a machine no unnecessary parts.”

The same is true for personas.

P.S. To answer the blog author’s question, personas are not typically real people, but if there happens to be a real person who represents a good persona, there’s no reason not to use them (with their permission of course). But please leave out all the unnecessary details.