The cost of costless communication

May 18, 2017

Post by: Nick Ellinger


It costs money to send an extra person an extra mail piece.  It costs no money to send an extra person an extra email.  So it’s natural to get logorrhea online and have a default of sending more to more.

But it turns out this does have a cost: $24,522.  That’s the amount that the average nonprofit lost to spam last year according to the excellent 2017 EveryAction Nonprofit Email Deliverability Study.

In fact, the study says that for every one percent of your email going to spam, you lose $1309.

And most of this comes from indiscriminate use of email. We’ll skip over the issue of data hygiene, except to say that neither Spielberg or Scorsese have returned my calls about directing my data hygiene PSA.

Rather, we’ll look at how you can reduce the costs of spam to your organization:

Reduce your email volume. We covered this spiffingly a couple weeks ago but the TL;DR version is that you can send fewer emails and get the same revenue.

In fact, our research gnomes are polishing up a study that shows that your best, high-commitment donors are turned off by additional informational communications. While these increase retention for your less committed donors (who don’t already know your mission), telling your best donors what they already know decreases retention.

Incidentally, if you want to hear about this research in more depth when it comes out, please sign up for our email newsletter here. It’s where we’ll announce the study and related white papers, webinars, and world tours:

Sign up for tips today!

Personalize and customize your emails. Here, I’m not referring only to names, although that is important. We recently received an email here at DonorVoice Global Headquarters addressed to

Hi {!Lead.FirstName},

From an organization large enough to know and test better.

But beyond this, it’s important to know who people are (i.e., identity) and what they want from your communications (i.e., preferences). In fact, DonorVoice research shows that allowing donors to control the type and/or frequency of communications from your organization is the number one way to get people to opt in. For identity, we’ve seen significant lift in response rates from knowing information as simple as what animal someone prefers (for an animal charity) and playing that back to them in communications.

Put more simply, have you ever marked something spam that felt like it was written specifically to you? How about something that felt like it was written by a machine for a machine? Other people are like that as well.

Zig when others zag. We set off a mini-firestorm last year here and on The Agitator when we suggested that maybe let’s not everyone send out indistinguishable matching gift offers for Giving Tuesday.

So what happened on Giving Tuesday with spam rates? The deliverability report said:

“For #GivingTuesday specifically, the data showed modest growth in email sends and open rates, but the spam figures revealed something shocking: the rate at which fundraising emails were rerouted to junk folders nearly quadrupled.”


“Using our research and benchmark figures, we found that a nonprofit with a list of 100,000, the average spam rate of 36.68%, and sending the average of 3 emails loses an incredible $6,184.47 on #GivingTuesday as a result of spam.”

This isn’t surprising. If you were somehow assigned to Dante’s eleventh circle of hell (they’ve added some on in modernity) and had to personally filter emails for spam, you’d see that an email was the third from an organization in one day and the 12th from any organization with the word “match” in the subject line and you’d drop it into the trash like it was an 800-pound maggot. Same thing with the computer spam filters.

And once you have one email that triggers a spam filter for a donor, the more likely you are to go into that spam filter for that donor and to go to junk for other donors. Deliverability is like trust – easy to lose, hard to gain back.

We’re not saying not to email on December 31st. But you can benefit from being an orange in a sea of apples by asking at different times and different ways — ones that are custom to your donors.

Make it easy to unsubscribe. A person telling you they don’t want to receive your emails is a good thing: you don’t have to waste your time on them and vice versa. And they won’t take more drastic actions like marking you as spam. It’s not just allowing people to unsubscribe easily; it’s asking for feedback in your communications. If someone is hot under the collar, it’s better than they vent their spleen at you than clicking “Spam” not only for your deliveries but also to learn what you can do better.

It’s not that these weren’t always good practices, but now they are survival tools. Spam rates went from under 10% in January of last year to over 20% in June and July with no appreciable increase in volume. Thus, we must make our emails relevant to our donors and constituents or filters are going to make our emails irrelevant.


The death of average in nonprofit marketing

May 11, 2017

Post by: Nick Ellinger


There once was a time where good enough was good enough.  The difference between a top 20% medieval serf and a bottom 20% one, for example, was likely almost unnoticeable.

This is not that time.  And last month’s Fundraising Effectiveness Report shows us why.  Here is the year-over-year gain versus loss of gifts from 2015 to 2016:

Average non-profits gained only 2.2% last year in gifts

As you can see, the difference between a top 20% nonprofit and a bottom 20% one is the difference between doubling gifts and almost halving them.

But we should also take no comfort in the middle. The average nonprofit increased their giving by 2.2%. And the latest US inflation rate was 2.4%.

Average is treading water.  Even just the difference between average and above average (the 60-80% tier) is the difference between doubling your giving in 3-4 years and doubling your giving by your retirement party.

That means it’s time to shrug off good enough and “best practices” that actually are doing by everyone, including the average.

We would ask you to look at every contact you have with a potential donor and ask if there’s one thing you can do to improve:

  • The tie to a donor’s identity — who they are
  • The tie to why the donor gives to you — their commitment to your organization
  • The donor experience — making the donor happier for having talked with you

These sound like small steps, but they are still enough to move an organization above average:

  • An organization that asked two identity questions at the beginning of a telemarketing script, then played those back to the person at the end increased their giving by 15%
  • Another organization that segmented its donors by their commitment to the organization has had three five-figure gifts from donors who had been raising their hands in the past but were called on because transactional data alone hadn’t identified them as targets.
  • A third organization measured their donors’ experience on their website and, with a few tweaks, increased their donation form fulfillment from 12% to 32% in one month.

Simple steps.  But these simple steps can take you out of the average trap and start making things outstanding, for you and your donors.


A brief mediation on ask string values

May 4, 2017

Post by: Nick Ellinger


I have an urgent dispatch from the cutting edge of ask string science!

By now, you’ve read our ATIMU-award-winning* white paper The Science of Ask Strings.

But you’ve probably also read many other things since.  Thus, you may not recall the importance of fluency in ask strings.  That is, the more common and easier to process the amounts are, the more people are likely to donate.  In particular, $100 in an ask string increases both response rate and average gift — a rare feat.  There was also a discussion of what makes a fluent, easy-to-process ask amount.  One set of authors posited that banknote values are the most fluent.  That is, if there’s a bill with that number on it, it’s a good ask string value.

There may be a different definition, though, with more to teach us about good ask string values.  A study from consumer pricing found that people’s memory for prices is directly related to how many syllables it takes to say the amount.  That is, people are more likely to remember $12 than $11 and more likely to remember $173 when it’s presented as “one-seven-three” than if it’s “one hundred and seventy-three.”  This fits previous research that we humans can generally remember the number of words that we can read aloud in 1.5 to 2 seconds.  It also fits the career of Curtis James Jackson III, who became far more recognizable as the price/stage name 50 Cent and, later simply Fi’ddy.  The evidence suggests that he would not have been as successful as the rapper $17.97.

50 Cent, Behavioral Scientist

50 Cent, Behavioral Scientist

As you might guess, memory and easy-of-cognition are close cousins in the brain, so we can probably assume that this also means that shorter-to-say numbers are easier for the brain to process than longer-to-say numbers.

So what are the implications for those seeking cognitive fluency in their ask strings?  This is where we get into speculation — no one to my knowledge has tested ask string syllables, likely under the (probably correct) hypothesis that there are better things to do with one’s life.  But I promised cutting edge, so speculate we will:

  • Zeros, not fives.  $25 is often used in ask strings, but this would posit that $30 or $20 would do better.
  • Generally, reduce your significant digits.  You are looking for a single digit, followed by some number of zeros.
  • Sevens are right out.  One more syllable than either six or eight, either of which would do better by this thinking.

Has anyone experimented with this in a significant way?  Please start the discussion in the comments!


*Award That I Made Up


For our UK colleagues

Post by: Nick Ellinger


(Or, I suppose our colleagues elsewhere in the world who happen to be in England on the 17th…) We have a live event coming up in the UK in a mere two weeks called “How to Redefine Your Supporter Journey.”

Forward thinking fundraisers have invested time, energy, and money looking to re-define the supporter ‘journey’. They’ve wisely looked beyond traditional recency, frequency, and value metrics, which can only tell us ‘what’ someone did. Not ‘why’ they did it. And no amount of additional demographics, behavioral segmentation, attitudinal surveys, satisfaction, loyalty scores, you name, it will answer the ‘why’ question.

With only a handful of exceptions, no one has delivered a supporter journey that’s been proven to work.

That said the exceptions are, well, exceptional. Charities who’ve applied the science necessary to answer ‘why’ people give, why they stop, and applied that to their journey, have:

  • Increased their acquisition conversion by 15%
  • Increased average gift by £10
  • Cut first-year attrition by 50%

On May 17th, the team behind these results will host a free event, to share exactly how it’s done. You will hear how to start:

  • uncovering why your donors support you specifically
  • exploring which experiences cause loyalty, and which don’t
  • identifying experiences which are working and could be scaled
  • working out which experiences are broken and need fixing
  • finding out which messages drive value
  • developing a journey based on knowing precisely what drives loyalty/value for your supporters

Places are limited, and on a first come, first served basis, so please email to save your free place today.



The volume hamster wheel: digital edition

April 27, 2017

Post by: Nick Ellinger


Last week, we talked about three ways to get off the hamster wheel 0f mailing more to get more revenue.  Some folks noted that I hadn’t mentioned shifting to digital solicitation as a way of saving those net revenues.

This is because email has the same problem of oversolicitation as mail.

It pains me to say it, because I grew up in an online space that was the wild west: grab a piece of land, stake your claim, and keep what you find.  I paid for my first cost-per-click ads with a faxed insertion order.  And with that statement, I feel this overwhelming urge to update my will…

But the West has been won (by Google and Facebook).  And now we are as overloaded with email (or more) as with mail.

Two proof points fell into my lap about this very point this week.  The first is with a client where we have been mailing and emailing their constituents less.  When we looked at a year’s worth of numbers yesterday, they had sent 30% fewer emails to this test audience.  The drop in gross online revenues?  Two percent — within the margin of error.

A random sidenote: this is where a skeptic would be jumping into mention that email has no marginal cost and thus by mailing more, this organization was able to raise more, net and gross, by emailing more.

But you know better, no?  You are paying, internally or externally, for someone to write, design, and send those emails.  And don’t get me started on the approval process.  Imagine what you could do with the extra time you could save doing 30% fewer emails.  My guess is it’s better than “increase digital revenues by two percent.”

The other is more public.  M+R just came out with their latest online benchmarking study, which I highly recommend.  In it, they find that that fundraising emails increased by 24% from 2015 to 2016.  However, the click-through rate on fundraising emails went down 12% — nonprofits had to run faster and faster to (largely) stay in the same place.

But, you might say, click-through rates (CTRs) are down across all emails — it’s part of the game now.  If that’s the case, then CTRs should be down about equally across all types of emails.  On the other hand, if the number of emails sent is (at least partly) to blame for lower CTRs, you should see the types of emails that had the greatest increases in sends have the greatest decrease in CTRs.

Let’s go to the tape:

Type of email Increase in number of emails sent Decrease in CTR
Advocacy emails 66% increase 36.7% decrease
Fundraising emails 24.1% increase 11.6% decrease
Newsletter emails 22.7% increase 6.3% decrease

Whatdayaknow? Not a perfect correlation by any means, but a pattern nonetheless.

And I bet you will find it with your own email program as well. If you’ve been increasing the number of emails you send, have you seen a corresponding decrease in click-through rate?

That’s the hamster wheel. And all of the techniques discussed last week still work here as well.

It just takes someone willing to stand up and say “I don’t want more; I want better.”  And, if you’d like help making that case, you can probably tell it’s a passion of ours; we’d love to help.


Three off-ramps from the volume hamster wheel

April 20, 2017

Post by: Nick Ellinger


You get it.  You’ve read all the think pieces, including ours and our other one, that show that many, if not most, mature organizations overcommunicate with (and especially overmail) their donors.

You know it’s the top complaint you get: “you mail too much” and “take me off your list” fight for your #1 and #2 comment and you can’t see third place from those two.

And you know the amount of effort you personally put into crafting more, more, more appeals, scripts, emails, and more, thinking there must be another way.

But you have a goal to meet and promises to keep and miles to go before you sleep.  You don’t know how to make goal without the volume machine.

Here are three off-ramps that have worked for three very different organizations.  There may be more and I’d love to hear about your experiments in the comments.  But three should get you started…

1. Donor clustering. Our friends over at DonorTrends just released a case study from a human services organization where they were challenged to mail 15% less and hit the same revenue numbers.  The organization separated their donors into three tiers: one who could bear the most communications, the least, and the Goldilocks in between.  From there, they were able to send their more marginal pieces only to their best donors and only their best pieces to their most marginal donors.

Results: The organization didn’t just stay where they were; they increased net revenue by 14% and decreased their cost to raise a dollar by 17%.

This type of modeling won’t help you figure out if your top-end of donors are getting too much mail, but it’s a way to start down the path of decreasing volume and increasing value.  And it will get you buy-in for more radical solutions.

2.  Going where the money is. Our friends at the Agitator did a piece earlier this week about how behavioral science can impact fundraising.  In the piece, Roger talks about the US Olympic Committee.  In looking at their program, they realized that much of their donor file was premium dependent.  For a $20 donation, you could get all manner of Olympic swag in the mail.  Not surprisingly, this attracted more donors interested in the merchandise than a philanthropic gift to support Team USA.

USOC took a bold step: changing the offer to focus on a philanthropic gift and turning everyone who wasn’t interested over to their merchandise sales department.  The Sixth Ring program focused on fewer, higher-touch communications and an offer of a quarterly gift with a minimum yearly commitment of $100.  They’ve increased annual gift frequency from 1.6 to 2.4 gifts per year and substantially increased their revenue per donor, average gift, and response rates.

All because they chose to focus on those people who matter most to them.  Think how much mental energy you put into $10 donors (who are wonderful — it’s great to give something of yourself, no matter the amount — but you have to focus on increasing revenue and thus mission overall) and what would happen if you channeled some of that toward your best and brightest.

3. Pilot testing.  I should mention that the USOC example started as a pilot test and became so successful it is now the focus of their direct marketing program. DonorVoice has had the pleasure of working on their program along with a couple other pilot programs that have decreased mail quantity while increasing donor value.

But I want to talk about the Union of Concerned Scientists, whose bold move to test four communications instead of 12-15 got within five percent on a gross basis and saved a quarter of a million dollars in costs.

Or better yet, I’ll let Laurie talk about their strong results:

As we said at the beginning, conceptualizing and implementing radical change can be tough.  But hopefully by reducing quantity through modeling, focusing on your best donors, and/or running a pilot test, you can get off the volume hamster wheel and start investing your time and treasure more on getting to know your donors and letting them know you know them.


Bolding the fine print

April 13, 2017

Post by: Nick Ellinger


It is a “best practice” to put certain things in fine print.  Car dealerships run through their loan information at a speed that would make the Micro Machine man quail.  Telemarketers run through the part where they are paid solicitors as briefly as the law will allow.  And Apple’s 20,000-word opus iTunes Terms and Conditions has just been adapted into a graphic novel.

I’m not making this up.

Terms and Conditions

But there may be a reason to take at least one of these out of the footnotes and into the sunlight.  In a paper called “Do I Care if You Are Paid?”, Gneezy et al looked at a door-to-door fundraising campaign.  Some of the solicitors were paid and some were not.  Each of these groups was then split into revealing their paid/unpaid status or not.  In addition, those people who were supposed to reveal their un/paid status wore a name badge with “paid worker” or “unpaid volunteer” under their name.

In their F2F outreach, they talked to over 1000 people and found that when paid solicitors revealed that they are paid, the mean donation increased by 16%.  This was due to an increase in donation rate from 16% to 19% rather than an increase in average gift.

Interestingly, women exclusively drove this difference.  Mean donation went from $2.33 to $4.38 when women were informed that someone was a paid solicitor and response rate went from 15% to 23%.  Men’s mean donation went down when informed someone was paid, but the overall effect was positive for all people.

So the ideal solution would be to reveal paid solicitor status only to women and not to men, but if you were creating a one-size-fits-all policy, it is better to reveal and do so openly than not to reveal.

This got me thinking – what else deserves a reprieve from the fine print?

Research from DonorVoice’s own Dr. Kiki Koutmeridou found that the thing you can do to make your opt-in process the most compelling is to allow people to control the types and/or amounts of communications they get:

This is off-neglected online (partly as the result of inflexible unsubscribe solutions in some of the nonprofit industry’s biggest email solutions), but even more so offline.  One must often call a special number in fine print then do a little jig to get removed from a nonprofit list.

Like the paid solicitor information, it likely pays to have this information transparent, easy, and (gasp) legible instead of burying it.  If donors know they can change how they hear from you (or stop it entirely if they tire of you), they are more likely to begin a relationship with you.  This is for the same reason that you are more likely to go on a first date with a person than marry them.

What else deserves to be pulled from the fine print?  Love your thoughts on the comments below.