What is your persona? Well, before you can even answer that question, we’d need to assume that you understand what exactly we’re even talking about—because it’s certainly a lot more than a person’s personality. Rather, in the digital context, personas are little bundles of information that are gathered about you and grouped together to create a ‘you-shaped’ digital representation that is used to try and predict what you will do in certain situations. The goal, of course, being to influence that outcome. Some version of this happens in real life too. (Have you ever been sized up by a really good salesperson?) However, the breadth and depth of creating these digital personas can have long-standing implications to nearly every aspect of our lives.
In real life (IRL) we each assume different roles as we go through our days. At an interactive session during the 2022 Mozilla Festival, we asked participants to note the roles that they play. For example, in the image above, the person using the orange avatar to represent themselves selected five different roles that they identify with:
- Sibling
- Mentor
- Volunteer
- Consumer
- Provider
These roles are examples of the “faces” that we wear in specific situations—we assume these roles to help us adapt to and be successful in these situations. They ensure that we are acting appropriately, assuming the right level of assertiveness, “looking the part” that we are assuming, and much more. You’re not a different person in these cases; you are just being selective in the moment about how you are presenting yourself.
In real life being selective about what you present works pretty well. That information might be interpreted in a continuum between two extremes:
- THE FIRST TIME MEETING: When someone is meeting you for the first time, the information that you present to them at that moment becomes their whole picture of you. Each new piece of information helps to fill in the gaps to form a more complete picture. Things that they particularly like or dislike are more likely to be remembered more strongly. As a result, you often are able to influence others’ impressions of you based on what you present, particularly when you take into account what reactions you perceive from the receiver of this information.
- THE LONGER TERM RELATIONSHIP: If someone has known you for a long time, they will take the information that you are presenting at the moment and compare it to all of the other things that they know about you. If it matches their current ideas about who you are, those ideas are reinforced, otherwise it is discounted. It would take quite a bit of unmatching information to change what they already think about you.
But, the digital world is different in two significant ways. First, you usually have limited choice in what you present in a digital transaction. And second, the information that you provide can be combined with information from other sources to try and influence your behavior in the moment. Consider the difference between buying something in person vs online.
| In Person | Online | |
|---|---|---|
| Purchasing power | You might pay by cash which indicates your ability to buy the item, but not much more. | your choice of payment may convey if you might be able to buy more based on if you used one of those convenient payment plans or purchased with a platinum credit card |
| Where you live | One can tell where you are shopping, but generally doesn’t know where you live | Your purchase must be sent to you so your address is shared. |
| What you look like | Heh – it’s in person. They can see you. | But, don’t think that online is blind to how you look. By combining your address information with generalized demographic information and purchase statistics available for that location for a small fee (that many retailers pay), it is likely that you are revealing plenty of information about your gender, race, age, etc even if you don’t explicitly provide it. |
| What you like and don’t like | If you browse a store, someone would need to follow you around and take notes about what you considered buying before you made an actual purchase. This would be expensive and unwieldy to do. | Online every search and item that you look at are captured and cataloged inexpensively and efficiently. They can be contrasted to your purchase choices, not just for this purchase, but for every purchase made at this online store and sometimes at others. |
| Non-purchase info | Maybe you shopped at a small local business where the store owner knows you and your family, and maybe you have chatted about pets, vacations, and more. | An online store may also know this information, but not because you told them any of it. Sites that capture information about you give away, buy, and sell information with other sites and data sources to get a similar depth of knowledge of you, even if the info has nothing to do with your purchase. |
Why? Personas in Marketing
In the 1950s, Wendell Smith advanced the field of marketing to incorporate the economic concept of “segmentation”, the practice of dividing your target market into approachable groups based on demographics, needs, priorities, common interests, and behavioral criteria. (Dr Smith’s paper is publicly available.) This idea took flight by the mid-1970s when an academic marketing science research group centered around Wroe Alderson started talking about market segmentation as a way of both tying marketing investments to sales revenue, and as a way of guiding approaches to target marketing strategies to specific groups of people. Today, every course on marketing includes a discussion of segmentation.
Each market segment is expressed in terms of a “persona”, a description of a fictitious “sample” person from the segment. These descriptions are used by companies to make decisions about the needs of this persona and what marketing messages would resonate. Real people are then grouped and seen through these lenses. You can see an example of a marketing persona below.
An Example: The Norah Jones Problem
There are many music services these days that will create playlists for you based on what you like to listen to. The idea is that they profile the music that you listen to, and present other music with similar profiles for your listening pleasure. Many years ago, I discovered that, no matter what I started listening to, music by Norah Jones would end up in my playlist. I have nothing against her beautiful voice – it’s just that, to me, the music profile of her songs do not match that of the songs that I was listening to. I had been assigned the “Norah Jones Persona”, and just like in middle school when you get a nickname that sticks even though your last name isn’t actually pronounced that way, I don’t seem to be able to shed this assigned persona.
When your choices are A) buy this item online because you can’t get it locally (and have the transaction be tracked), or B) do without (and avoid the tracked transaction), you really may not feel that this is a choice at all and may go ahead with the transaction. As a result, information about the transaction will be collected and stored, including things like what you bought, your credit card number or other purchase identifier, perhaps an account sign in, your computer’s IP address, maybe a history of what else you searched for during that transaction, and more. That recorded transaction may be grouped with other transactions that match similar characteristics. Sometimes that information provides insight about the object purchased such as ‘of all of the things searched, product A is the most popular.’ And, sometimes that information provides insight about you such as ‘this person usually buys blue items when there is a choice.’ A computer algorithm might assign you the “Blue Lover” persona, even if you only bought blue things because all of your favorite colors had been sold out. Alas, now you too have the Norah Jones problem.
This A Big Deal
This sticky assigned persona could just be a minor annoyance or it could have significant repercussions when used to discriminate, punish, exclude, target, surveil, or if the data are used with an unintended context or in a way that the data subject objects to.
In the 1990s, Latanya Sweeney, Ph. D. wrote and defended her PhD thesis about a then-emerging field of study, Data Privacy. (The next week she was testifying before the US congress about this topic!) Through simple experiments, she found that seemingly anonymous data that was publicly available could re-identify a person with just three pieces of general information with a high (97%) degree of certainty. If you hear Professor Sweeney speak, it’s hard to NOT want to learn more about how this significant use and aggregation of data into personas impacts civil rights, credit reporting, health privacy, equal employment, elections, and more. If you have 30 min, I highly recommend watching one of her fantastic talks.
So, what can we do?
This is another question that we posed in our interactive session during the 2022 Mozilla Festival. The participants provided GREAT answers. Look out for our “What can you do to control your persona” resources soon. In the meantime, enjoy these suggestions from the conference participants:
