Effective Segmentation and Persona Development
What is Segmentation?
In marketing theory, segmentation is the separation of your audience—in our case, donors—based on common factors. Much of those common factors can be tied to what is sometimes called the ABC’s of analytics:
- Acquisition—How did you acquire this donor?
- Behavior—What did they see and do?
- Conversion—How much is the action worth?
The word segment sounds like taking a whole and cutting it up into pieces. But the actual application of segmentation is usually more about adding than dividing. For example, you would add contacts that you know are meeting certain criteria to a specific segment. Of course, personalization and segmentation conditions can quickly become overwhelming the more you drill down.
The Virtuous Way
According to Virtuous’ Rachel Specht, “Segmentation is a function of assigning Contacts to Campaign’s Segment(s) based on the results of running their associated Queries or Filters. In other words, Segmentation is used to determine who exactly receives which version of your communication.” The example she uses is of nondonors, lapsed donors, and major donors as three potential segments—based on giving frequency and giving level—to whom you plan to send out your end-of-year mailer, which will include an ask. If Contacts 1-4 have never given to your organization, they will be placed in Segment 1 and will receive the nondonor version of the end-of-year mailer. Clear and easy to understand.
The Boys & Girls Clubs of the Ocoee Region Way
Like you, when I first migrated to Virtuous, I had to make a decision about how to structure my data, including segments. The Virtuous Academy is excellent, as are the help articles like the one I mentioned by Specht. But at that time, I struggled to understand how I was going to set up our database using the Campaign>Campaign Communications>Segment tri-level hierarchy. I felt like I really needed four levels to create a trail of breadcrumbs that made the most sense for how our org was already functioning. That meant, for starters, that I would need to cut off one level either at the top or bottom of the hierarchy in order to work inside of Virtuous.
So, I started at the end and worked backwards. While giving frequency and giving levels are valuable, they don’t often tell me the story I personally want or need to know. Additionally, I really didn’t care about specific methods of communication per se. Since, for us, that would be the fourth level I spoke of, I decided to (essentially, not completely) cut that from our Virtuous data hierarchy. What I REALLY cared about in the final analysis was HOW—very specifically—each donor ends up giving. That isn’t terribly unlike how companies like Amazon and Netflix focus their efforts. Amazon cares about the items that you look at, save, put in your cart, and most of all, actually purchase. Netflix cares about what you preview, save, and most of all, stream all the way through. I cared about the fact that this donor paid board member Steve $500 for a Cash 4 Kids (C4K) ticket. I cared that this donor scanned a QR code off a flyer at a festival booth and signed up to be an Everyday Hero (monthly sustainer). And I do my best to track every single one of those engagements.
In traditional marketing terms, this would be considered behavioral segmentation, and the donor sets that his/herself by their own actions. For me, there is no guesswork involved. I simply watch what the donor does and, to the best of my ability, respond appropriately. In more traditional segmentation, typically you set the rules for who your audience is in each segment. But by using this alternate structure, the audience “naturally” enrolls themselves in different segments BY THEIR GIVING BEHAVIOR.
Since giving behavior is what I cared about most and I had three levels to work with inside the software, I just used them in the most logical way I could think of, despite their labels. Campaigns represent our organization’s generic resource development categories, Campaign Communications is the specific sub-category, and segment is the behavioral giving “trigger” and/or response of the donor.
CAVEAT: Even with such a detailed approach to tracking how gifts are triggered, I still don’t always know the behavior that accompanied a gift. A gift comes in—usually in the mail or delivered by a staff or board member—and there is no information when it crosses my desk. Why did they give? What was their journey? No clue. And sometimes, no one else knows either. So I am in the process of developing user surveys that will hopefully help me gain some insights on those things and more. But for now, I just do my best to record whatever details I can.
Now, when I create marketing communications for donors, I can create things from any of the three levels. I could send a mass letter to everyone who’s been to any Special Event, or only to those who attended C4K event, or only to those who bought C4K tickets from a specific board member. In fact, last year, I had all my board ticket sellers each write custom thank you letters, and I sent their letters to their ticket buyers using those segments. And each year, I pull the historical data of who sold to who, and I create sales packets for each board member that shows who they sold to and the contact info for that donor so that they are primarily selling to a “warm market.”
Stereotypes and Personas
With our particular setup, where do personas come in to play? Well, for us, they don’t. Since I’m focused on trackable donor behavior, I don’t have any real need for personas. But even if that weren’t the case, I’m not sure I would use them. Let me explain.
Renowned marketing expert, Jeff Bulas, says, “Customer personas are our attempt at creating archetypes for our customers that help us make meaningful decisions to improve customer targeting, conversion and sales based on preconceived generalizations, which can be based on a number of things. The problem with these archetypes is that they often tend to boil down to stereotypes.”
‘Stereotype’ feels like a dirty word today, but for decades, marketing professionals have relied on the dependability of stereotypes to appeal to mass audiences. In other words, with a good stereotype, marketing professionals had a dependable set of characteristics that they could rely on in crafting their messaging. But today, that just doesn’t work as well as it used to. In fact, people today are actively and purposely defying stereotypes. (Just look at YouTube reaction videos as one small example. Most of those shatter stereotypes!) So, if your ‘persona’ is a 56-year-old white guy who is obsessed with 80s funk (like I am), you’d never know that I equally love 80s metal. That’s why a lot of companies are forgoing persona stereotypes and instead are depending more and more on consumer behaviors.
Netflix
According to the Netflix Inc. Report 2023, Netflix stopped segmenting its subscribers based on geography in 2016. They now treat over 221 million global subscribers as one community with identical TV programming and movie needs.
“As a primary method, Netflix employs behavioral segmentation to give personalized content to over 100 million users daily. It uses Machine Learning to learn about Netflix subscribers based on how they use the Television streaming service. The information is then stored by Netflix, which uses it to divide users into groups depending on their actions, enabling the entertainment network operator to create a more tailored client experience.”
The Critical Question to Ask
In the Virtuous article, “Understanding Campaigns in Virtuous,” Meredith Matson said, “There is no right or wrong way to structure a Campaign. Just consider your reporting needs & remember to always track your activity!”
So whether you choose a simple, proven, traditional segmentation model or some alternative model like the one we use at BGCOR, the critical question I try to ask is this:
Does this specific segmentation detail significantly CHANGE the way the donor thinks and feels about the problem our organization solves?
For example, if you use demographic segmentation, you might ask questions such as:
- Does the donor’s AGE significantly CHANGE the way they think and feel about the problem we solve?
- Does the donor’s ETHNICITY significantly CHANGE the way they think and feel about the problem we solve?
- Does the donor’s INCOME LEVEL significantly CHANGE the way they think and feel about the problem we solve?
You can follow the same model for psychographic details:
- Does the donor’s VALUES significantly CHANGE the way they think and feel about the problem we solve?
- Does the donor’s INTERESTS significantly CHANGE the way they think and feel about the problem we solve?
Geographic details:
- Does the donor’s LOCATION significantly CHANGE the way they think and feel about the problem we solve?
And, of course, behavioral details:
- Does the donor’s GIVING LEVEL significantly CHANGE the way they think and feel about the problem we solve?
For those details where the answer is “no” or “probably not” (hopefully based on real data and not just assumptions), it’s probably safe to simplify your job by ignoring—or at least de-emphasizing—those criteria. For those that seem affirmative, you probably should think through the specifics of how to track and record that data and then use it in your segmentation strategy.
