# RFM part 2 – segmentation goals beyond net revenue

From yesterday, we had a sample membership mailing with these results from last year, and we were going to cut any segment that didn’t get \$.33 per piece or more, since that is what the piece costs.

To do so would actually cost us dearly.  As we discussed this week, there are more goals than just the immediate gaining of revenue.  Looking at mail pieces alone with a “here’s a piece; here’s another piece” mentality can ignore what other things a mailing can do for you.  Let’s take a look at this chart of historic gross per piece by segment with two ideas in mind: groups we want to invest in and testing opportunities.

Take a look at our \$50-99.99 37-48 single donors.  They have an anemic response rate of .3% and an average gift of \$80, so they would only gross \$.24 per piece to mail to (and thus lose \$.09 for each piece sent out).  Should you mail these donors?

There are some organizations that would say no – they think that every segment should net positively in a donor mailing or that they should only do no net cost acquisition.  I’ll try to demonstrate why I think these people are wrong.

With a response rate of .3%, it would take 333 1/3 pieces to generate one donor.  At a net loss of \$.09 per piece, that’s a cost of \$30 to acquire that donor.  Chances are that that is higher than you are spending in your acquisition mailings to get new donors.

But you aren’t acquiring just person off the street.  You are getting someone who then slots into the 0-3 \$50-99.99 M segment for the next mailing.  You can see if the person got this mailing again for their next one, they would be predicted to gross \$3.62 per piece sent to them or people like them, which is very nice.

Let’s run the numbers assuming that their average gift is \$50, your retention rate per year for lapsed reactivated donors is 50%, your retention rate per year for multi-year donors is 70%, and it costs you approximately \$10 to mail your \$50-99.99 donors for one year.  To make the math easy, we’ll assume only one donation per year (it should higher) and we’ll assume that any donation is worth a net of \$40 knowing the mailing costs (in reality, you would want to look at both the possibility that someone will give multiple times per year and that you will have to mail someone even when they don’t give).

This works out to:

.5 * 40 + .5 * .7 * 40 + .5 * .7^2 * 40 + .5 * .7^3 * 40 … .5 * .7^n * 40

(Can you see why I simplified the math?)

What this basically says is that there’s a 50-50 chance of getting any future gift from this person and they have a 30% chance of lapsing every year thereafter.  We aren’t using a discount rate because interest rates now are so low.

To simplify, it’s \$20 + \$14 + \$9.80 + \$6.68 and so on.  A bit of high school calculus later and this donor will likely return an average of \$66.67 to your organization.  All for the cost of \$30.

If you had a magic box where you could put in \$30 and didn’t know what it would give you back, but knew the return would average over \$60, you’d put money in.  I myself would ask if I could put in more than \$30 to speed things up, like asking the genie for more wishes.

See the full comic and other fun stuff here

In general, your multi donors are going to be far better donors.  However, you need to communicate to single time donors in order to get those multi donors.  You also need to talk to those people whose last donation has been a while to renew them for future support.

The corollary to this is that you shouldn’t just look at this segmentation and see what to cut; you should also be adding back in.  Looking at these gross revenues per donor, you are probably (hopefully) wondering why you wouldn’t want to mail 7-12 month single donors of \$100+, or deeper into your \$1000+ donors, or more.  These are all correct thoughts.  Looking farther back into your pieces, you might see that someone has made the previous mistake – they looked at a small sample size of a \$1000+ mailing, found that no one responded, and cut the segment.

Thinking further about this, you can see that perhaps the \$1000+ donor shouldn’t get this piece, but they probably should be communicated with.  These are your best and best potential donors and there probably is a way to increase their value more so than not communicating with them.

Similarly, you’d love to renew those \$15-\$19.99 13-24 month donors, but this also isn’t the way to do it.  Now we are going to break out of yes/no segmentation and into using segmentation to create differentiated communications.  For simplicity sake again, we’ll assume that we have four treatments we are going to try:

• This mail piece, plus a pre and post email, for our “normal” donor segment (red)
• This mail piece with “lapsed” language, plus a pre and post email, for our lapsed segments (green)
• A high-touch invitation-style mailing to higher-value donors with first-class postage to invest in getting their gift (with email and higher-touch telemarketing as well) (blue)
• A prerecorded outbound voice mail campaign, coupled with an email ask to less valuable or less likely donors to attempt to renew them without high marginal costs. (yellow)

I’ve oversimplified here.  With the high-touch piece, we’d almost certainly want to test borderline segments part with the high-touch and part with the control to see if the additional investment is worthwhile.  You’d also want to test what segments telemarketing works best with.  And so on.  But for first steps, it’s directionally correct.

So let’s color in the mail plan with these four layers.

Note that you are able to contact more people with more appropriate language with this strategy.  Segments that a pure net perspective would have ignored are renewed in this new model and our most valuable donors are treated that way.

But that’s still just RFM with some embroidery on top.  It’s a fine model, but there’s more that can be done.  We will go beyond RFM tomorrow to add a few other pointers tomorrow.

# The building block of donor segmentation: RFM

Over 80 percent of nonprofit marketers know the term “communication segmenting” according to a Bloomerang survey.  Over 60 percent say they segment their files.  This means that almost 20 percent of them don’t know what segmentation is and 20 percent more know about segmenting, but don’t.

Let’s see if we can reduce those numbers.

Segmentation, as you can see from my last two posts, really is for everyone, regardless of the size of your file. We’ve talked about, obliquely, two different types of segmentation – yes/no segmentation (will someone be communicated with or not) and segmentation to help create versions or customizations.

RFM analysis can help with both of these.  It stands for:

• Recency: Usually used as how recent was someone’s last donation to the organization. You may occasionally also look at their last interaction with the organization, but we can put that aside for now.  This is perhaps the primary driver of segmentation and, if there is an answer to the “how many people do we have on file?” question, it’s when it phrased as “how many people have donated to us in the past two years” or the like – with a time horizon and action attached to it.  Do yourself a favor and add “recency” to your Microsoft Word dictionary; you are going to be using it a lot and Microsoft Word doesn’t know that word exists.
• Frequency: We go from the most often used to the most often ignored – how many times someone has given to the organization (or interacted with in non-donor contexts). This is often simplified to single v multi, as this dyad makes it easier to plan your communication.
• Monetary value. This is usually measured by another TLA (three-letter acronym): HPC or highest previous contribution – what’s been the highest amount the person has made in any one gift. This is an area of some debate, as if someone makes gifts of \$20, \$20, \$20, \$20, \$500, \$20, \$20, \$20, and \$20, it is fairly predictable that they are probably better grouped with the \$20-\$24.99 donors than the \$500-\$999.99 donors and their ask strings changes along with this (more on ask strings here).  One solution is to use a formula like 2/3 of their HPC + 1/3 of their most recent contribution or half of their HPC + half of the average of their last five contributions.  But this is something worth testing how it works best for your file.

Let’s do a yes/no segmentation with RFM.  We have a membership mail piece that has historically very well with a number of segments.  It costs \$.33 to mail and you are looking for segments that net.  Here are what your RFM response rate, average gift, and gross revenue per piece would look like from last year.

Response rate:

Gross per piece:

(these are intentionally realistic, but false, data)

To explain, the first numbers are the months (so 0-3 is someone who last gave a gift in the past three months), S & M stand for single and multi (get your mind out of the gutter), and the dollar amounts across the top are that person’s highest previous contribution.

Looking at this, to maximize net revenue, we can cut some of the segments to lower HPC groups and to one-time donors.  Anything under \$.33 per piece isn’t going to net us money.

This is a decent baseline that answers the question “I have this mail piece going out in February; to whom should it go to maximize the net revenue of the mailing?”.

However, we are going to look at it as “how do we maximize the value of this donor by treating them appropriately?” and layer in some treatment segmentation tomorrow.

# Wherefore segmentation?

Yes, wherefore.  As long as we are starting from first principles, we can go a little bit Elizabethan.  In the one and only famous “wherefore” quote, Juliet isn’t asking where Romeo is (below the balcony).  Hers are existential questions – for what reason does Romeo exist and what cruel twist of fate made him a Montague, her family’s mortal enemy?

For more of this, check out my likely-never-going-to-be-written book The Bard Does Nonprofit Direct Marketing (All’s Well that Ends with a Donation).

But wherefore segmentation – why does it exist?  We covered a lot of this in the last post, but we’re going to be going into them more granular than that as to who gets what communication when.  Why are we doing this?

The simple answer is “to maximize revenue.”  In this world, every mail piece would be opened and responded to, every phone call answered, every email and online ad clicked upon and donated to.

In this world, the ideal model would be one that gets this 100% response rate – it would read people’s minds and get them the lowest possible cost means of communication to get the maximum gift at the precise right moment.

This is not a horrid definition and, in fact, that would be a really cool (if magical) model to apply.

But it ignores two things: how people give and what your goals are.

Let’s say you have a person who, every year, like clockwork, gives to your membership mail appeal every January.  She’s on your email list, gets your e-newsletter, and a number of other mail pieces each year, but only gives to that one membership mail pieces every single year.

Do you think she would still give to you if that was the only communication she got from you throughout the year?

Probably not. I once walked each year for an organization that will remain nameless.  Every year, I started getting emails from them a couple months before the walk encouraging me to walk (whether I’d already signed up or not) and I would stop hearing from them after the final walk email for another 10 months.

Please notice I say I “once” walked for this organization, not that I still do that.

The bottom line is that even the most loyal of donors (especially the most loyal of donors!) want to hear from you.  Look at Professor Adrian Sergeant’s surveyed reasons why someone stopped giving to an organization:

The full study is here; it’s real and it’s spectacular.

Many of these involve someone not being communicated with enough (not acknowledging support, don’t recall supporting, no longer needs my support) or not being communicated with effectively (other causes more deserving, not informing how money was used).  Now look at the bugaboo of many an ED or board member: inappropriate communications is less than 4%.  More people defect because we don’t talk to them than defect because we talk to them too much.  So we can’t do just the pieces that “work” for a person without cutting the heart out of our communications.

As mentioned earlier, it also ignores other goals you have for your direct marketing program.  In a classic, Mal Warwick’s The Five Strategies for Fundraising Success articulates there are five goals you can set:  Growth, Involvement, Visibility, Efficiency, and Stability (GIVES).  He further says these are to a large extent mutually exclusive.

I’m not going to ruin the book for you, but this is just to say that there are things you want beyond maximizing short-term revenue.  You may want to get long-term revenue, volunteers, advocates, awareness of your causes, and more.

So how do we restate our goal?  How about:

The goal of the direct marketing program is to maximize the lifetime value of each of your constituents.

This isn’t just financial lifetime value if you have other non-financial goals, but it likely helpful to help quantify what you are willing to pay to get, for example, an advocate in order to put everything on the same scale.

This is important to have as a definition because it will help you transcend many obstacles.  When should your direct marketing donors get a major gift officer working with them?  When it will increase the donors’ lifetime value (and shame on you for saying “your donors” – donors belong to no individual within an organization). Should your national office or field offices do communications to donors?  Well, which mix will maximize lifetime value? These will likely need to be tested, but won’t it be nice to have an objective answer to some of these?

We’re going to initially talk about RFM analysis, which takes a look at which donors should get which communications.  This is absolutely necessary as a baseline.  However, if you are looking to maximize the lifetime value of each constituent, you will have to look at things differently.  It’s a minor difference, but you will need to think of “should this donor and donors like them receive this communication?” rather than “who should this communication go to?”.  It’s when you get to the point of thinking about donors first and make your communication vehicles reflect that rather than taking your communications and seeing to whom they should go.

# Why “mail everyone” is never the answer

I’ve had my first request – to talk about the basics of segmentation. Thanks and keep them coming!

The question will eventually come, if it hasn’t already: “how many people are on our list?”.  The answer?

For those who may not get the reference, it’s from The Hitchhiker’s Guide to the Galaxy.  When assigned to find the Ultimate Answer to Life, the Universe, and Everything, the computer Deep Thought ponders this for 7.5 million years and comes up with the answer 42.  It then says that:

“I checked it very thoroughly … and that quite definitely is the answer. I think the problem, to be quite honest with you, is that you’ve never actually known what the question is.”

The question of how many people are on our list or in our database is similarly ill-formulated.  You may have a list of X people, but are you really going to ask for a donation the person who called your executive director a [redacting] [redacter] at the gala 11 years ago and never corresponded with your organization thereafter? So your list is X-1.  How about dead people?  You need them in your database so you know not to solicit them.  And so on.

Every time you communicate with these folks, you are losing something.  Usually, for methods like mail and telemarketing, there is an additional marginal cost for each person reached.  With email, there isn’t, so email tends to be the least segmented direct marketing method.  This, however, ignores that there is a cost for not segmenting an email file; the less people open your emails (and especially in cases where you are emailing an account that no longer exists or is checked), the more likely systems are to think you are a dirty dirty spammer and cast you down into the ranks of personal attribute enhancers, Nigerian princes, and your great-aunt who thinks that you absolutely need to know about the ecstasy-lased gummy bears “epidemic.”

The bottom line is that even when you aren’t paying the bill, you have every incentive to make sure your list is as trim as possible.  That means not communicating with the deceased, opt-outs, those with incorrect communication data (although you should be doing NCOA (national change of address), eCOA (electronic change of address), and corrective phone appends on those people you would still like to talk to), and those who have opted out of the medium (e.g., email opt-outs) or message (e.g., solicitation opt-outs).

You may think that once you clear those people out of your list, you should have a defined number.  However, different people are at different stages of interaction with your organization.  Here are a few:

• Some guy (aka suspect)
• Prospect
• One-time giver
• Multi giver
• Sustainer
• Mid-major donors
• Major donor
• Planned giving donor

These are frequently presented in a pyramid because there are only so many easy-to-use graphics in PowerPoint.  Some may say it’s because suspects and prospects are the base of the program and they grow from there in smaller and smaller numbers, but you and I know the truth.

The truth is that these are like the stages of grief, in that they don’t always apply, don’t often go in order, and abstract over significant parts of the donor journey. For example, take a look at the types of retention you should be measuring and you’ll see that there are categories – first-year, but not first-time, givers and reactivated lapsed – that this pyramid doesn’t take into account.  Similarly, you will see people who are major donors on their first gift, people who you didn’t know about who leave support to you in their will, and the former sustainer who no longer wants to support your organization.

I take inspiration from Stephen Jay Gould, who critiques in many discussions of evolution the ideal of progress or, worse, inevitability:

Progress is not merely a deep cultural bias of Western thought…it is also…the explicit expectation of all deterministic theories of evolutionary mechanism that have ever achieved any popularity, from Darwinian selection to Lamarckism to orthogenesis. I do not, of course, mean progress as an unreversed, unilinear march up the chain of being; Darwin did away with this silly notion forever. But even Darwinism anticipates that an imperfect, irregular, but general ascent should emerge from all the backing and forthing inherent in a theory based on a principle of local adaption to changing circumstances.

Stephen Jay Gould, “The Paradox of the First Tier.”

In reality, the donor journey isn’t a net gradual march from suspect to prospect to one-time donor, to multi-donor, and so on.  But the general theme of this – that you should treat different types of donor different based on what you want them to do for the organization, what they want to do for the organization, and their means and interest – is a good one.  A suspect and a potential major donor are very unlikely to want the same communication in the same way.

So there are some folks you may want to get some communications, but never others.  Other such groups:

• Board members
• Organizations
• Recent donors
• The people you serve
• Public officials and opinion leaders
• People who have requested a certain number of communications each year

That’s the broadest type of segmentation – what type of people do we want to include?  But we’ll want to increase revenues and save costs by sending the most effective communication possible. So in this week of segmentation, we’ll talk about the philosophy of segmentation, then start with a basic segmentation – RFM analysis – and build from there.