Why Your Segmentation Probably Stinks (and What to Do About It)

Suppose you were asked to separate your donors into Pile A and Pile B. What criteria would you use?

Before answering, recall the only reason to segment is to raise more money than one-size-fits-all. Creating and executing tailored content and experiences come at a cost, which at the very least is more work for staff.

To increase the chance the juice is worth the squeeze, you’d probably want to set up some rules like:

• Pile A donors should want different things from your organization than Pile B
• The differences among the people Pile A should be smaller than the differences between Pile A and Pile B people.

Do you segment like this? Most don’t. And often those that say yes to the first question are potentially using an approach that is conceptually on the mark, but fatally flawed on methodology.

Most segmentations are poor proxies for donor needs, like:

RFM segmentation. No charity is segmenting on RFM to create tailored journeys and different experiences for their $25 to $49, zero- to six-month recency versus the $50 to 99, seven- to 12-month recency. These mindless, seemingly endless slices of the data file are used, if at all, for selection (who is in or out for the one-size-fits-all journey) or maybe, just maybe, to send nicer paper stock to the higher value people.

But mostly this is “used” for reporting on performance. The reaction from any sane human looking at the Excel sheet with all these random, endless rows of mindless RFM splits is, “What the hell do I do with this?” Answer: nothing. This is faux segmentation that gives the term a bad name.

Demographics. Are men from Venus and women from Mars, and do all men want something from your charity that is uniquely different from all women? Are Millennials—all 75 million of them—a tight, well-defined marketing segment that want fundamentally different things from older people? These hypotheticals underscore violation of the two rules: not all people in a demographic pile A are ever the same or even close to the same nor are they different enough, consistently enough from Pile B.

The evidence that demographics are crap is significant and you can find more here and here.

Personas via lifestyle/psychographic/demographic clustering are often justified by the “persona” and “clustering” terms and the price tag—you’d think nothing that expensive can be that bad. But this type of segmentation only gives you the most generic segments. Whether it’s “Pools and Patios,” “Frugal and Frequent” or “New and Noteworthy,” these are usually demographics and/or transactions repackaged, with the weaknesses of each. Most importantly, these aren’t organization-specific, so there’s little actionable data to be gleaned. You want something that will raise more money, not pretty Powerpoint slides or a big invoice. Two hints on your persona clustering being nothing more than vague stereotypes: 1) The lovely profile for Group A doesn’t fully describe a single person in that segment, or 2) the differences between the segments have nothing to do with why they support your organization.

So that’s a lot of bad news.

The good news is that there is a better, proven alternative: segmenting by donor identity.

When you segment your appeals and communications by donor identity, using messages that speak to those identities, you dramatically increase the commitment/loyalty and lifetime value of those donors.

You also change your economic engine from volume—of both donors and communications—and all of the diminishing returns it entails to one based on donor understanding

The more understanding you have about core motivations, needs and preferences, the more segments that get created and the more customized and relevant the interaction becomes. In turn, the more profitable those segments become.

When you put higher lifetime value and a new revenue/growth model, together the improvement in net is dramatic (15 to 25 percent) and can repeated year after year. Anybody can grow gross or net in a single year; sustaining that level of growth is another matter.

Essential First Steps

To use donor identity to segment your donors, you first need to gain specific information from those donors. You must also put in the work to validate identities that reflect materially different needs and preferences. Let’s assume we have three distinct identities for an organization combating Type 1 (juvenile) diabetes: those who have Type 1, those who have someone with Type 1 diabetes in their immediate family and those who don’t have a direct or indirect cause connection.

These questions can be asked in an onboarding phone call, an online survey or a mail-reply device. And, not shockingly, most donors will have a connection. Those that do will be of higher value than those who don’t have this direct connection, and they will want different things from the organization.

There are also likely subgroups within this; not everyone with Type 1 has the same values and goals or is looking for the same thing from the organization. Folks in the early stages might be looking for help with diet and lifestyle and insulin options. Others who have been living with the disease are looking for research to find a cure.

These identities (e.g. person with Type 1) and the subgroups defined by what they want from the organization would logically get different touch-points and journeys.

Changing copy so that it says, ”Because you’re affected by Type 1 diabetes…“ is necessary, but not sufficient. Donor identities aren’t just about copy points. It’s about meeting the donor’s need, to the point that the content looks like it came from the services side of the house. Yes, you’ll still ask for money, but only after you meet the need that their identity demands.

So, to end the hypothetical, if you were the fundraiser for ABC Juvenile Diabetes and had this information, what type of file segmentation would you be using?

Would you still be segmenting by transactional information hoping that the donor in the under 14-month bucket with the most recent gift of $25 responds to your generalized message. Or would you tailor your appeals—and the entire donor journey—around what donors recently diagnosed with Type 1 want?

And if it’s true for this hypothetical organization, what does it mean for your organization? After all, every organization has different identities of donors in their donor file. Using these identities gives you the significant advantage of knowing why donors do what they do, not simply what they’ve done.

There’s no getting around the cold reality of today’s fundraising: identity with your organization is ingrained—or not—in the donor. And the donor has thousands of options. If you aren’t catering to their identity, someone else will.