# Toward a linear RFM

In addition to the many challenges of RFM already discussed, the segmentation puts up artificial barriers between segments.  Some of these include:

• Let’s say someone is one of the people we talked about yesterday that gives every November or December.  If s/he gave in November 2014, then again in December 2015, are you really going to consider them having “lapsed” in the middle?
• The distinction between frequency groups is artificial. As we discussed on Tuesday, Sandy, who gave you 100 gifts, and Miriam, who gave you two, are both considered multidonors for most RFM segmentations.
• The distinction between monetary value segments is artificial. Which donor would you prefer – a donor who donates \$10 ten times per year or a donor who donates \$50 once a year?  RFM prefers the latter; I’m guessing you would prefer the former.

But how to do you create equivalencies among every different segment?  Would you rather have a donor who gave \$100 to an acquisition package six months ago or a loyal semi-frequent \$20 donor?

The ideal would be to run a model with lifetime value as the dependent variable and your traditional RFM variables, plus as many of the ones that we’ve talked about this week, and determine what your actual drivers of value are.

But lifetime value, as you can tell from the name, takes a long time.

So let’s steal a rule-of-thumb model from the for-profit world.  Connie Bauer first (at least first to my knowledge) proposed this in an influential 1988 Journal of Direct Marketing article called “A Direct Mail Customer Purchase Model.”  Here, I’ve replaced purchases with donations; I think it works in our world with this replacement.  To get the RFM score, you multiply these three things together:

1. The reciprocal of recency of the last donation in months.
2. Number of donations
3. The square root of the total amount of donations the person has made.

• There’s a reasonable equivalence between recency and frequency.  Would you rather have someone who has given four gifts who gave their last gift a year ago or someone who has given two gifts and their last one was six months ago?  These would be roughly equivalent in this model and that looks about right.
• It mitigates the artificial distinction between months.  That 12-month versus 13-month difference that in a normal RFM analysis could be the difference between sending and not sending a communication?  In this model, it’s about an 8% difference in scoring.  Important, but not fatal.
• Because I’ve not seen the effect of the sheer numbers of gifts have a huge impact (once you get above a certain point) on retention rate, it seems intuitive that monetary value is a smaller factor than the other two.

There are some weaknesses.  Donation amounts aren’t linear: if someone has given a \$25 gift in the past, the odds that they will go from there to \$26 to \$27 is not likely.  Some time periods, like a year, are somewhat magical, especially for one-gift-per-year, seasonality-focused donors.  And in an ideal world, you would want more recent gifts weighed a bit more than more distant gifts.  A donor’s behavior tomorrow will be more like their behavior last month than their behavior in 1988.

But given that, it’s an interesting look at the topic.  I hope the week gives you the courage and the tools to take another look at your segmentation strategy and calculations.  You’ll go nuts if you try all of these simultaneously, but conscious and continuous improvement can make huge differences in the long term.

# Vive le donor difference

When I was but a wee lad, I played youth baseball.  Or perhaps more accurately other kids played baseball at me.  I excelled in three things and three things only:

• Bunting
• Getting hit by pitches, to the point that I once got hit by a pitch that was called a strike.  I had to wait to get hit by the next pitch to take my base.
• Stealing signs.

This last was where my “talent” was.  I would watch the third-base coach and when I thought a steal was coming from the signals, I would yell into the pitcher and catcher from my position in right field.  (Of course I was in right field.  There’s a chance someone might hit the ball to left field.)  I probably caused more outs with catching signs than catching balls (though still far less than I caused by batting).

The trick to stealing signs is to look for what is different from the usual.  The same is true for catching donor signals – the trick is to look for what is unusual and work from there.  Some tips:

Seasonality: Most donors are season agnostic.  They donate when an appeal touches them or strikes their fancy or they hear about you on the news or they found a \$20 in a purse in the back of a closet.  However, some will renew membership in January like clockwork.  Others believe in end-of-year giving (this is prevalent among online donors).

Like everything else, there is a way of doing a sophisticated model to determine this.  However, like only some things, there is also a fast, relatively easy, and free way to do it in Excel or similar spreadsheet:

1. Pull all of the gifts at which you want to look.  I would recommend donors with at least three years of giving history and at least four gifts, so you have a sufficient history to work with.  You want the gifts labelled by a unique donor ID number.
2. Label all of the gifts by month (1 = January, 12 = December, and everything in between)
3. Run a pivot table that summarizes the gifts by donor with the min month and the max month.
4. Subtract the min from the max.

(If you’d like a walkthrough of this in more detail, please email me at nick@directtodonor.com)

Now look at the results.  The majority of donors will likely have a wide spread of 9, 10, or 11 months.  However, you will also see some 0-3 month spreads, meaning that over (at least) a three-year period and (at least) four gifts, they have given to you only in one quarter of the year.  Thus, you can likely reduce your costs on soliciting them in the other quarters of the year (not eliminate, as you don’t want them to forget you exist).

If you want to be very thorough, add six to each month number and repeat to capture those few donors who may focus their gifting around both the end and beginning of the year, but not the middle.

This is powerful combined with seasonality analysis; if you find someone only gives at the beginning of the year to your membership campaign and has never given to a premium piece, you don’t need to send them address labels in May or the calendar in September or telemarket to them in June.  Instead, you can use lower cost (and more cultivative) pieces like donor newsletters to maintain the relationship with them.  Yes, this may only be saving \$3 per year per donor, but if there are 10,000 of those donors on your file, you are talking about real money.

Out-of-place gifts: Someone has given you \$10 times.  They just make their 11th gift to you: a \$173 check.  What should you logically ask them for next time?

HPC says you should ask them for \$173 (possibly rounding to \$175).  Common sense says that the person may not have turned from a generally smaller donor to a prospective mid-major prospect overnight.

Research indicates a better answer is to use average donation of giving for longer-term donors.  Thus, you see the anomaly, take it into account, but don’t let it drive your decision making.

Another potential treatment is to use a continuous, rather than segmented, version of RFM.  We’ll discuss that tomorrow.

In the meantime, if you are interested in more research on ask strings and amounts you should ask for, I’m working on a book/white paper/whatever it ends up being on just that topic.  Newsletter subscribers will get a free PDF copy of it when it comes out, so if you would like one, please sign up for my free weekly newsletter here.

# Using non-donor knowledge to enhance segmentation

Yesterday, we introduced you to two special people that a traditional RFM analysis would group as 4-6 month \$25-49.99 multis.  To wit:

Since Sandy first donated to your organization in 1992, she’s given over 100 gifts.  Nothing exorbitant – she’s now giving \$30 every three or four months – but she also has volunteered, come to three walks, signed up for emails, and taken almost every advocacy action you offer.

On the other hand, you acquired Miriam from an outside list in 2012.  She gave \$25, but nothing since then.  You don’t have her email or phone number, but a last chance lapsed package piqued her interest four months ago and she gave another \$25.

We talked about how their donation history can and should differentiate them.  There are additional indicators here, however, that can also enhance your messaging and segmentation:

Online interactions.  If someone is active online, it’s relatively simple to group their interests by their activity – what they click on, look at, and interact with.  (Actually, technically, interact with is the easiest, click on is slightly harder, and look at can be a bear with some online tools.)

With Sandy, she is an advocate for you and doesn’t seem to require premiums to donate – perhaps you can replace the labels in that upcoming package with a paper version of an action alert – cheaper, and likely more effective.

Other organizational interactions.  Sandy has been a walker – do you want to mention that your walk is coming up in 90 days in the PS or in a buckslip?  Similarly, you should probably customize the messaging to acknowledge that she has given her time as well as her donations.  Making her feel known will only help her loyalty.

Outside data.  Getting outside data on your donors can help you adapt your tactics.  If you find out that Miriam does all of her banking online, perhaps she’s a better target for an EFT-based monthly gift than you thought (with the right messaging).

List co-operative data may indicate that that she gives to nine other charities far more often and more generously than to you.  Perhaps she’s just not that into you and you might want to cut your losses soon than you might have thought.

You may find out she does a lot of business on the telephone and find that it isn’t your organization that wasn’t lighting her up; it was the means by which you were approaching her.

All this and more can come from data appends.  And you can try to get that email address and engage her online, so hopefully you can learn more about her.

All of this – donor and non-donor interactions – are masked by an overarching RFM category.  But what if we could dispense with RFM categories altogether?  We’ll talk about that Friday; if you don’t want to miss it, or any of our Direct to Donor posts, please sign up for our free weekly newsletter.

# Including loyalty in your beyond-RFM segmentation

Since Sandy first donated to your organization in 1992, she’s given over 100 gifts.  Nothing exorbitant – she’s now giving \$30 every three or four months – but she also has volunteered, come to three walks, signed up for emails, and taken almost every advocacy action you offer.

On the other hand, you acquired Miriam from an outside list in 2012.  She gave \$25, but nothing since then.  You don’t have her email or phone number, but a last chance lapsed package piqued her interest four months ago and she gave another \$25.

What do these two have in common?

They look the same on a traditional RFM analysis: they are both 4-6 month \$25-49.99 multis.

And if you use only a traditional RFM analysis, you will treat them the same.

That’s silly.  If you were looking only at these paragraphs to judge, Sandy would seem to be a good candidate for monthly giving, upgrade strategies, and/or planned giving.  Miriam probably has a 50-50 chance (or worse, given industry averages of lapsed reactivated retention rates) of never giving you another gift.

It’s easy to criticize this, but harder to do this analysis writ large, when you are doing five-to-seven-figure list selects.  So how do you draw these lines?  Here are a few ideas:

Lifecycle analysis.  Way back when (November 2015 – ah, those were the days), we talked about how there isn’t just one retention rate – there are several, based on where a person is in their donor journey.

This lifecycle analysis can layer on to your segmentation analysis and on to your messaging.  Some sample categories:

• New.  What it says on the tin.
• 1st year.  They have given a second gift, but it’s been less than 12 months since their first gift.
• 2nd year.  Gifts in their first two years.
• Core: Donors who have given in each of the past three 12-month periods
• Lapsed: A gift 13-24 months ago.
• Deep lapsed: A gift 25+ months ago.
• Lapsed reactivated: Someone who has given a gift in the past 12 months after a gap of at least a 12-month period

Your mileage and organization may vary – it’s more important to look at this analysis than it is to have the same precise categories.

So you may not have a separate piece for Sandy, but you might want to make sure there is language like “As one of our most loyal donors” or “You’ve stood with us for more than 20 years.” or the like in the existing piece.

As for Miriam, as a lapsed reactivated donor, you are still worried that you might lose her again.  Perhaps you want to stay close to the tactics that recruited her (or won her back or both).  She might also be worth an e-append or phone append to see if you can find a channel that further engages her.  Or maybe you want to use a less aggressive ask string – your goal for a lapsed reactivated donor is to make donating a habit again, rather than to increase their giving just yet.

Gift density.  Take a look at the number of gifts someone has made, then divide by the number of years since a person’s first gift.  This is how many gifts you will get from them in an average year (or at least what you have received).

Sandy’s number is above four.  Four is a bit of a magic number (some would say three or even two– again, having a number is more important than what the number actually is) to indicate strong likelihood of monthly giving.  When someone has a pattern of giving frequently, this ask isn’t nearly the heavy lift it is trying to get someone to go from one gift per year to twelve.

Miriam is below one.  One is a separate magic number, as below one indicates a likelihood to lapse (by definition, they’ve done it at least once)  That should trigger some of the anti-lapse efforts discussed above.

One is also a magic number in that if someone gives you exactly one gift per year (and they’ve been with you a few years), that’s the bucket they see you in.  So, if they look unlikely to upgrade and they look unlikely to increase the frequency of gifts, the only other way to increase their lifetime value (other than increasing their retention rate) is to decrease costs.  Let’s say you send an average of 14 mail pieces per year and do two telemarketing cycles.  This person probably can decrease this substantially and save costs.

Longevity.  Length of donation is something that should be honored.  Not only are milestone anniversary notes and certificates and the like a good thing to do from a moral and ethos perspective, but they will also make sure that your most loyal donors know that you know they are important to you.

Channel responsiveness.  Change your tactics to suit the terrain.

All of these are even more important when looking at borderline segments.  Should you mail the 13-18 \$15-\$19.99 multis?  Maybe just those that have been with you five years or more?  Or with previous high gift densities?  Or just mail responsive?

But there’s more to it than even that; tomorrow, we’ll talking about using other interactions with your organization to define and customize.

# RFM segmentation alone must die

Recency, frequency, and monetary value (RFM) are the ruling troika of segmentation-land.  And like one of the old Soviet troikas, they brook no challenge to their rule (e.g., Trotsky, pictured at right, was murdered on Stalin’s orders with an ice ax).

But they are simply not good enough alone anymore.  I tried to be civil about this in my post Beyond RFM.  But beyond is not good enough.  We need to let a million flowers bloom in the world of segmentation.

This means taking the “7-12 \$15-\$19.99 multi-donor” view of segment out for a date with your ice ax.

OK, not really.  It’s still going to be a decent starting point.  But it has to stop being the ending point.  Even for those of us that have to leave complex modeling to people with more letters after their names.

So this week, I’d like to take you through various different ways of figuring out the all-important question “is communicating with this donor in this way going to help achieve my goals of net revenue, quality file growth, and/or world domination?”.

And the first topic that should be layered on is listening to what a donor’s behavior is telling you.

Part of this is non-donor behavior.  You likely already have this information if you have the donor’s email.  You can potentially tell if they’ve been to your Web site, how often, how long they spent, and what they looked at.  You definitely should be able to know how they’ve reacted to emails you’ve sent them in the past.  The difference in a lapsed donor who still regularly opens your emails and clicks on the articles versus one who, according to your email records, may or may not be dead is a significant one.

If you can get robust data, so much the better, because now you can not only include people in a communication they may not have received before, but also customize it based on what they are interested in.

But some of this is donor behavior you already know, but RFM filters out.  Channel is one. Take an online donor who is reliable and frequent at donating online.  If you’ve mailed her/him 25 times over the years to try to get him/her to donate, but s/he hasn’t responded, chances are that s/he doesn’t want to give through the mail.  Personally, I’ve found telemarketing to be the most persnickety channel: those who give through it really give through it; those who don’t, really don’t.

Another is cadence.  If someone has given you ten gifts in the past ten years and all of them have been in November or December, my money is on the fact that you can ease off the gas in May.  One program of my acquaintance runs a membership campaign that starts every January.  There is about five percent of their file that will give a membership gift like a clockwork every January or February and then nothing for the rest of the year.  Should you stop trying to get extra gifts?  No.  Should you cut your cadence way down and save yourself some costs?  Yes.

These are things the donor probably thinks they are telling you explicitly with their behavior.  It’s now incumbent upon you to listen.

Because tomorrow, things get a little bit harder, as we talk about lifecycle and loyalty.

If you’d like more nonprofit direct marketing content like this, please subscribe to my free weekly newsletter. You’ll get all of my blog posts, plus special subscribers-only content.

# Microtargeting and the ABCs of customization

Microtargeting is most often thought of in the political realm, where increasingly granular models are able to predict how people are going to vote and think about various issues.  A good example is how Ted Cruz won Iowa: by microtargeting the interests and issues of voters down to fireworks regulation.

But you don’t have that type of time, budget, or modeling power.  Yet you still want to connect with your donors in ways so that they know that you know them.

Enter the poor person’s microtargeting.  We’re going to slice and dice our control letter in such a way that there’s something in it for everyone.

The important thing to remember is that the cost in customization is largely in customizing one side of a piece of paper in the mail.  Online, it’s virtually nothing.*  There can be some data costs, but while the maximum customization approach below may churn out thousands of different combinations of letter, it still is all very simple variables acting predictably.

So here goes the ABCs approach.  Try as many of these as you can on your appeals and see how different one person’s would look from another:

Age:  Does your older donor want a larger font size?  Different levels of formality?  Two spaces instead of one?  Including the Oxford comma?

Buckslip: What could you put in the envelope, based on what you know about the donor that would make them more likely to donate?  Remember, you don’t have to have it for all donors, just some…

Channel responsiveness: Don’t ask someone for their email if you already have it.  But do sent them an email that support the mail package they just received.

Donation history

Event history: “You wanted with us to cure X.  Now we need your help again.”

Frequency of giving: If someone is giving 4+ times per year, might now be the time to ask about that monthly giving program?

Giving history: “your gift” versus “your gifts.”  Also, have they given the same amount year after year?  You probably don’t need to push the upgrade.  However, if they’ve been steadily rising, go for the gusto.

History with this appeal: “As someone who supported our matching gift campaign in the past…

Initiation: “your support has helped X over these past Y years” or “since you joined X years ago.”

Jargon: J is tough, so a reminder to go through your letter and remove anything that sounds like a great buzzword to you, but gobblygook to those outside your organization.

Knowledge: How much explaining do you do?  Is it the same amount for someone who has read 50 letters as someone receiving their second?

Location: “we’re looking for seven dollars from XXCityXX willing to chip in…”  This works.

Mission area supported: tie your ask to what they want to support.

Nicknames: Does your letter sound like it was written by C3PO: “Dear Dr. Lt. Col. R. Winthrop Huntington III, MD (ret.),”?  If you tell by his checks that he actually goes by Bob, do you want to try saying “Dear Bob”?

Online activity: Mention they were a petition signer as an inducement to get them to sign an offline petition.

Postage: Send your most valuable donors’ mail first class.

Questions they’ve answered: The letter of someone whose survey said they thought it was most important you educate young people should look different from the one who said you should be advocating for better laws as a top priority, no?

Rhythm of pieces: (aka cadence, but I already had a C).  Should this person even be getting this piece or are they likely to make a gift without?

Single versus multi: With singles, you can switch up the ask string. Much harder to do with multis.

Tchotchkes: Are you sending premiums to everyone?  Even those people who have never responded to a premium?

Unique URLs: Not necessarily personalized URLs, but different URLs for different messaging so you can see what creates the greatest online response.

VIPs: If someone is a member of the “Founder’s Circle” or the “Legion of Good Deed Doers” or whatever it is you have, are you referencing that?

Wealth screening: You can do a higher-dollar treatment if you know a person has the capacity to make a larger gift.

seX: You didn’t think I was actually going to get a real X in here?  Appeal to women’s emotions in your ask and to men’s self interest

You: I’m cheating with this, because it’s not a customization.  But it does give me the opportunity to quote Jeff Brooks’ sample fundraising ask letter, which makes me happy:

Dear [name]:

You. You. You. You. You. You. You. You. You. You. You. You. You. You. You. You. You. You. You. You. You. You. You. You. Yes, you. You. You. You. You. You. You. You. You. You. You. You. You. You. You. You. You. You. You. You.

Sincerely,

[Signature]

P.S. You. You. You. You. You. You. You. You. You. You. You. You.

Instructions: Liberally sprinkle in nouns and verbs. Use adjectives and adverbs sparingly. Include specific examples of what the donor’s gift will accomplish. Include true-life stories that demonstrate the need for the donor’s involvement. Be sure to clearly and articulately ask for a gift more than once.

Someday, I’ll write a blog post that good.

Zip selects:  Increase your ask string multiplier if they are from a wealthy ZIP code.

* Get it?  Online?
Virtually nothing?  I absolutely slay me.

# Beyond RFM – doing intermediate-level segmentation

By now, hopefully, you have seen the benefits of segmentation and the value of RFM as a preliminary tool.  But you have probably intuited a few of the flaws in it already:

• The distinction between recency groups is sometimes artificial. Let’s say two people give \$20 per year, one at the end of January and the other at the beginning of February.  The former would be at 13 months for a January mailing; the latter at 12 months.  A few days difference could mean that one gets mailed and the other didn’t.  (In fact, I’d love to test a segmentation out to 13 months instead of the standard 12 because of this; please comment or email me if you’ve done this, so I can share and illuminate myself).
• The distinction between frequency groups is artificial. Your multidonors range from people who donate once a year over a series of years to people who donate to literally every communication they get.  RFM analysis gives these two people the same number of communications.
• The distinction between monetary value is artificial. You probably saw this one coming, given the first two.  Which donor would you prefer – a donor who donates \$10 ten times per year or a donor who donates \$50 once a year?  RFM prefers the latter; I’m guessing you would prefer the former.
• It carries literally no other information. The \$25 dollar donation of today could be from someone who clipped coupons to raise the money to donate or Bill Gates divesting himself of what he found in one cushion of his couch.  What a person donates is an indicator of capacity, but it’s a blunt tool when a scalpel is possible.
• You probably noticed I was talking only about the mail in examples. RFM doesn’t look at channels of donation, nor at the sensitivity of people to those channels.

Let’s address this last one first, as we look for ways to customize RFM more than using it alone.

Separate RFMs by channel. When you do your telemarketing list pull, you will almost certainly want to call deeper into your file to people who have made a previous telemarketing gift than those who you are trying to make their first gift on the phone.  Same thing in the mail: you may be willing to mail multidonors out to 48 months \$50-99.999 for people who have given through the mail, but only to 12 months or 18 months \$50-99.999 for people who have only given online or on the phone.  Separate RFMs by channel will help you make these determinations.  Remember that you are looking to make sure that you have a strong multi- or omni-channel program, but that doesn’t mean you have to be agnostic as to channel of origin or preference.

Cadence analysis.  This goes to your \$10 10x versus \$50 1x donor example above.  You have to figure out who likes/needs multiple communications to make a gift or gifts and who doesn’t.  Some ways you can do this:

• Look at how many types of offers you have going to a person in rotation; hammering the same key on a piano over and over isn’t music, nor will the same approach be music to your donor’s ears.
• Take a look at how many times you mail someone in acquisition. Sometimes chronic non-responders need a different offer (as with rotating your offers above), but you can also see whether someone after seven or 17 or 27 times of getting a message from you means they almost certainly won’t donate and you can bless and release.  The same holds true for your lapsed donors as well.
• Test consistent communication versus no communication versus resting periods to see what happens when you try different cadences.
• Look at frequency of gifts within a certain period of time, rather than ever. You can see dramatic results sometimes by decreasing your mailings to your one-time-per-year donors versus your multi-per-year donors.

Mission area: I mentioned this when discussing customization types, but tailoring your communications toward the area of your donor’s interest is also a legitimate segmentation target.  Why would you send that advocacy alert to people who care only about your work in schools or your calendar with pictures of dogs to your car people?

Income: If you don’t want to ask that millionaire for \$17, a wealth append can get you the information you need to customize your ask strings and communications strategy.

Location: Zip codes are the poor person’s income append and can be free-ish, so that’s a potential win.  More often, though, you can use zip code modeling to breathe life into underperforming acquisition lists.  Simply find your top X percent of zip codes from your current file and ask for just those zip codes from the rental list.  It will be slightly more expensive to rent per name this way, but it can provide a 20%+ lift in response rates and/or average gifts.  This is also a way to test new lists while minimizing risk.  One caution: you don’t want to do this for all lists or you risk self-limiting your acquisition strategy.  If a list works well for you across zip codes, use the whole thing – that way you give yourself a chance to be wrong about a zip code long-term.

Demographics: Some nonprofits find that different messages work better for men versus women.  Age can also be helpful, as you work to avoid sending your one millennial donor your planned giving brochure (there’s optimistic and there’s delusional…).  With demographics, backtesting makes initial sense, where you see how people of different demographics responded to previous appeals and messages, then use that data to define your strategy.

Previous responsiveness: It sounds obvious, but it’s ignored by CRM: if someone like getting your calendar three years ago, they may like getting it again.  Replace “calendar” with member card, action alert, survey, etc., and you have the makings of a profitable add on to your usual list mix.

Those are some of the things you can add to spice up RFM.

I said at the beginning that this was intermediate segmentation.  Advanced segmentation is modeling.  The hacks above will help get you many of the benefits of modeling at a fraction of the cost, but it won’t get you all of the benefits, so definitely leave yourself open to building smarter and smarter donor modeling solutions.

Any other segmentation recommendations you’ve seen work?  Please leave them in the comments.