A few years ago while at a leading tech financial firm in Silicon Valley, a team of marketers from my division were presenting their new personas to our VP of marketing. They had assembled a group of eight personas, complete with life-size cutouts they felt represented various customer segments. The team divided the personas into groups with descriptive names such as “Credit-Savvy Carl” and “Budget-Conscious Betty.” After their presentation, the VP asked a few questions and then stormed out of the room.
“I don’t ever want to see these again!” he shouted as he punched poor Carl, sending him tumbling to the floor. Eventually, the team found out they focused too much on the raw data and spent very little time uncovering audience intent. Great intention, lousy execution.
So what's the relationship between profiles and personas, and how can we use both to improve our marketing strategies?
The History of Buyer Personas
Alan Cooper is widely recognized as the “Father of Visual Basic” and the inventor of personas as a way to create more user-friendly design. Cooper’s work developing software in the ‘80s led him to a startling conclusion: software engineers were tasked with both developing and designing interfaces. He realized this process needed better input for engineers to create more useful products.
While designing an early version of a Business Intelligence software, Cooper realized the best way to understand how people might use the product was to interview them. From these conversations he created Chuck, Cynthia, and Rob – the Adam and Eve of personas. Each had distinct uses for the software. Once he identified each user, Cooper explained individual user needs. The end result was a much more intuitive product. Cooper first wrote about personas in his 1997 book, The Inmates Are Running the Asylum: Why High Tech Products Drive Us Crazy and How to Restore the Sanity.
Profiles vs. Personas
Today, profiles and personas are often used interchangeably, but they are really quite different. Profiles are compilations of data, including:
- Age Group
- Race or nationality
- Geographical Location
- Technical Ability
Personas, on the other hand, are based off of research into who the customers actually are and their intentions with the product or service you are selling. Effective personas define individual behaviors and go beyond profile data to extrapolate intent. A few sample questions around intent include:
- How do you use the product?
- What do you intend to do with it?
- What are your goals for using the product?
The point here is that two users with identical demographics can be motivated and influenced in different ways through their interaction with your product. As a result, marketers will need to persuade them differently.
Cooper breaks out the behavior variables for online shopping by service, price, necessity and entertainment. These behavioral variables should be formulated into questions that guide your development of the personas. By asking all users the same questions in the same way we start to see patterns that can help us as we develop the personas.
I caught up with Cooper the other day, and I asked him if anything had changed since he first developed personas. His response shouldn’t surprise us.
“I think it is unfortunate that the design community has so widely abused the very useful notion of design personas,” said Alan Cooper. “While there have been many positive changes in UX over the last 20 years, much of the thinking in 'The Inmates' is still fresh and relevant."
At Groove, we believe intent is the key to effective personas. We want to understand user motivation when they visit our clients' sites. Knowing why someone is visiting your site is the key to creating a rewarding experience for them.
Speak to your customers. Listen to customer service calls. Hold focus groups. Use profile data to assemble what you believe are segments and interview random members. Create a standard list of questions to probe their intent. Then use that information to create your personas. But most importantly, do the primary research - don’t just take the profile data and give them names.