December 2006
In this issue:
1. Extra: Testing matching gifts
By Dean Karlan
2. DonorSpeak: Pros and cons of Internet surveys
By Dirk Rinker
3. What's Working: Reiterating the message
By Deborah Block and Paul Karps
4. Special Report: What to do about premiums? A case study
By Peter Schoewe
By Dean Karlan
FUNDRAISERS have long relied on supposedly tried-and-true methods of soliciting funds from donors. For example, fundraising experts have often said that matching grants, when possible, are brilliant for boosting fundraising—and the higher the ratio the better. If an organization can secure a 1:1 match, great. If an organization can secure a 2:1 or even 3:1 match, even better.
Nevertheless, such rules of thumb are largely anecdotal, as little scientific study has been completed to prove such claims. More important, we know little about why matching grants may work. If we understand why they work, we can develop packages that maximize their effectiveness.
My colleague John List of the University of Chicago and I recently conducted an experiment for one charity that answered questions about whether matching grants work and how much a matching gift’s ratio matters. The charity sent out 50,083 solicitations, and 67% of those got a matching grant offer, while the rest served as a control group.
AGAINST CONVENTIONAL WISDOM
The study, conducted with a national politically oriented charity with a liberal base, found that simply announcing a matching gift campaign increased the overall revenue per solicitation by a statistically significant 19% (from $0.81 to $0.97) and raised response from 1.8% to 2.2%.
The effect worked entirely by increasing response. The average amount of each gift barely changed. Yet, contrary to conventional wisdom, when larger match ratios such as 3:1 and 2:1 were offered, they did no better statistically than a 1:1 ratio. Revenue per solicitation for the three match ratios was $0.94, $1.02, and $0.94, respectively (response rates were 2.1%, 2.3%, and 2.3%).
The experiment also found something completely unexpected. The match only mattered in red states! The offer of a matching gift boosted the revenue per solicitation in red states by 55% (from $0.68 to $1.06) through increasing the response rate from 1.5% to 2.3%. While for blue states, neither the revenue per solicitation (which stayed at $0.90) nor the response rate (which rose from 2.0% to 2.1%) changed significantly.
WHY, OH WHY?
Why did this happen? Why does the matching donation generate higher responses in red states but not blue states? We can only speculate. Perhaps liberals in red states are frustrated being in the minority in their state. They feel hopeless. And the matching offer is a sign of hope, a sign that the timing is right to reenter the fight.
Perhaps the prior donors in blue states were simply different from those in red states. While many people have noted that red states and blue states are not really that different, people are keenly aware of the political coloration of their state, and this affected their responsiveness to the match.
Further tests underway by this team will help understand why this happened, which will help fundraisers make better future decisions on how to maximize response rates and gift amounts.
This may all seem quite academic. And it can be. But it all leads to specific, usable prescriptions for organizations that help them maximize their fundraising. For instance, this charity should be considering different letters for their donors in red versus blue states. Increasing your revenue per solicitation from $0.68 to $1.06 is worth a little effort.
Many times one can “personalize” without incurring further mailing costs. It does require spending more time in the copywriting process, figuring out how to position an appeal to red-staters vs. one to blue-staters. But once a page on a letter is being personalized, it often costs nothing more to add a catch phrase or two that should appeal to particular individuals.
Studies like this help charities and fundraisers answer important questions for their fundraising. What is the best way to leverage a large, leadership gift, for example? Although it’s rare to find a silver bullet or universal rule that applies everywhere, better understanding of our donors can help charities improve their fundraising.
To view an example of a matching gifts package click here for package one and click here for package two.
Dean Karlan is Assistant Professor of Economics, Yale University, 77 Prospect Street, Room A105, P.O. Box 208209, New Haven CT 06520-8209, phone (203) 432-4479, fax (203) 432-3296, e-mail dean.karlan@yale.edu.
2. DonorSpeak: Pros and cons of Internet surveys
By Dirk Rinker
With an Internet survey, your questionnaire is programmed into a Web format using survey software and hosted on an Internet server. You invite potential respondents to visit the Internet survey using a postal mailing, an e-mail invitation, or online pop-up announcement—after which people access the survey using their Web browsers. Since the respondents answer the questions online, you don’t need to do any further data entry. This allows you to see the results in real time, as they’re collected.
Very often, charities will collect e-mail addresses from constituents, enabling them to contact their donors or customers online. This can be a valuable source of names for a survey. Opt-in Internet panels are another excellent source of respondent names.
The pros of online surveys include:
* Fielding costs are typically lower than offline surveys. Online surveys often don’t have postal costs, interviewer costs, printing costs, or phone calling charges. E-mail broadcasting is affordable: Delivery costs are mere pennies per message or less. Organizations with simple questionnaires can set up and design their surveys very cheaply with online survey companies.
* Results can be collected very quickly. Fielding can take days instead of weeks or even months. However, an online survey should stay open for at least a week to minimize the risk of bias.
* Internet questionnaires are typically easy to modify in the midst of conducting the survey. Want to add a question or restrict access to certain types of respondents? Not a big deal with most online surveys.
* Results from an online survey can also be reviewed online in real time. Graphs, tables,cross-tabulations, etc., can all be generated. This makes monitoring an active survey very easy, and allows easy sharing of data between departments and companies. In some instances, there’s no further reporting needed— eliminating even more time and cost.
* In comparison to mail surveys, Internet surveys allow a researcher to direct respondents efficiently to appropriate questions based on their responses. This tends to increase the completion rate and minimize data cleaning.
Cons for online surveys include:
* Response rates for Internet surveys have dropped consistently since their inception. It’s not uncommon to receive only 2% response.
* Some groups are naturally under-represented on the Internet. For example, only 8% of Americans online at home are seniors, though they make up 16% of the population (and often a much larger proportion of a nonprofit’s donors).
* It may not be possible to obtain e-mail addresses for the specific groups you’re trying to reach. Instead, it may be necessary to mail the survey links. However, response rates for mail to Internet surveys are lower than e-mail to Internet surveys, since mail pieces force potential respondents to go online and manually type in the link to participate.
* Internet technology is still young and continuously changing. Smaller organizations may find it difficult to maintain the technological knowledge needed.
* E-mail address lists typically cost more than mailing address and phone number lists.
* Security? Well, it’s easier to crash an online survey than steal someone’s paper survey out of the mail and fill it out. There’s also the problem of keeping respondents from answering your survey multiple times to get more incentives or affect the results.
Dirk Rinker is President, Campbell Rinker, 25600 Rye Canyon Rd., #202, Valencia CA 91355, phone (888) 722-6723, fax (661) 775-5926, e-mail rinker@campbellrinker.com.
By Deborah Block and Paul Karps
LONG-TIME READERS of this newsletter have seen many an article over the years about how to say “thank you” in response to a gift. Strategies and approaches,we’ve noted, can vary widely. For example,whether you should include a direct Ask for another contribution or just drop in a replyslip—or choose to be even more subtle by enclosing only a reply envelope.
There are a number of reasons to like this thank-you package from Planned Parenthood Federation of America (New York NY). The #10 outer includes the elegantly simple “Thank you!” teaser, handwritten in a bold blue font. The one-page, personalized letter— which features the gift amount embedded in the first paragraph of copy—is warm and engrossing.
While letter copy doesn’t try to solicit another gift directly, a reply panel is attached to the bottom of the letter. It reads, “YES! I will help Planned Parenthood Federation of America advance reproductive freedom and make sure every child is born wanted and loved.”
However, the real reason this package stands out is the copy shown here, which is typeset on the back of the letter.
Because rather than leaving the back side empty, it’s being utilized in a smart, compelling fashion to reiterate the group’s message. Call it a statement of principles or, as PPFA terms them, “our core beliefs.” But, at least from the perspective of the pro-choice donor who received this mailing, call it inspiring and uplifting.
The intent is to create a list of values, goals, and a vision of the future that the donor can identify with and feel a sense of pride for being a part of. Not to mention feeling the urge to do even more to advance the cause . . . as in making another gift.
Politics and religious beliefs aside, what more could you ask for in a thank-you package?
To view the entire package click here.
4. Special Report: What to do about premiums? A case study
By Peter Schoewe
ABOUT FIVE YEARS AGO, Mercy Home for Boys and Girls (Chicago IL) did something odd, but very exciting. After a decade of heavy use of premiums to acquire new donors, the group decided to go cold turkey. It stopped mailing premiums in acquisition and,instead, focused on acquiring high-value, non-premium donors.
I was fortunate to be a part of Mercy Home’s direct mail program during that period. And I’m proud to report how— after much trial and error, many false starts, and anxious moments— Mercy Home’s grand experiment resulted in a much more productive and valuable direct mail program.
Mercy Home has one of the oldest nonprofit direct mail programs in the country, dating back to the year 1900. Its long history of communicating with donors through the mail means that, even though it’s a local Chicago organization, 75% of its 250,000 active donors live outside the Chicago area.
HIGH RESPONSE, LOW VALUE
Throughout the 1990s, these donors were acquired using a high-response greeting card package that resulted in very low-value donors. After years of explosive growth using the card package, the bottom fell out in 1999. Response rates stayed as strong as ever (at about 4%), but the average gift fell so low—to below $8—that it became impossible to break even on the cost of the acquisition mailing, let alone the subsequent marketing costs, in any justifiable length of time.
At that point, we realized we were in a deep hole. We had built an acquisition program that was acquiring thousands upon thousands of low-value donors who would never pay off their initial acquisition cost. To make matters worse, a file audit validated our deep fear: The donors responsible for almost all of the net revenue in our direct mail program had a profile opposite that of the donors we were acquiring. Instead of being lowvalue premium donors, they were high-value donors who gave mainly to non-premium appeals.
To dig out of the hole, we knew we needed some aggressive testing. And in 2000, we thought we hit pay dirt. We tested a nonpremium package that did everything we wanted it to do. It had a response rate well below that of the greeting card package. But the donors we acquired gave over triple the average gift—and we paid for our initial acquisition investment in less than a year’s time.
Those early results were all the encouragement we needed to take the leap—and remove premiums from of our acquisition program for good. But, of course, it was all too good to be true. Our emancipation from premiums didn’t last over a year.
TOO GOOD TO BE TRUE?
The problem was that we needed to acquire a certain number of donors—even at the higher average gifts we were experiencing from our non-premium package— to meet our net-revenue growth projections. And, to meet our new-donor goals, we needed to roll out with non-premium acquisition mailings every other month, with each mailing containing at least three quarters of a million names.
This heavy acquisition schedule burned through our entire list universe in several months’ time. It wasn’t long before our once promising non-premium package was performing at the same break-even level as the reviled premium package it had replaced.
It was a humbling realization. As much as we hated premiums and felt they had led us down the wrong acquisition path, we simply had to use them to meet our goals for net-revenue growth. During this same period, Mercy Home was expanding its programs to double the number of children in its care. Because direct mail was responsible for the majority of the organization’s net revenue, we had to meet our goals for growth.
And that meant we needed to include premiums in our acquisition mix. But it wasn’t enough to begin mailing premiums again: We needed to mail them smartly.
Our first order of business—in both the premium and non-premium realm—was to reduce our reliance on a single control package. We began testing extensively in each mailing to build stronger controls and to gain more of them. And, in a ground breaking shift of thinking for us, we changed how we judged a package’s success. Instead of looking solely at cost to acquire a donor, we began to measure how much we could afford to invest in donors—package by package—based on their long-term value.
NEW BENCHMARKS
Using this new measure, we discovered we could resuscitate many of the old control packages we had dropped in favor of greeting cards. For example, while we could only afford to pay $6 to acquire a donor through a greeting card package and make back our money in a year’s time, we could afford to spend up to $11 using address labels to achieve the same 12-month break-even. And for a package containing a paperback book written by our President, we could afford to spend $16—while for a nonpremium package, we could afford to invest up to $35 per donor.
Another challenge was to optimize our results by matching offer and package to list. Not surprisingly, we found that many of the lists in our plan could only be mailed premiums. But we also had a good number of lists— mainly smaller, exchange-only lists—that we could use to gain higher-value donors through non-premium mailings. Through package testing, we replaced our initial non-premium control with two new packages. One was intended exclusively for a Catholic audience and the other designed to appeal to more secular donors, which further reduced list fatigue in our more frequent non-premium drops.
Of course, all these changes in acquisition had an effect on our housefile mailing strategy. Through our years of premium acquisition, we had clung to the idea that we needed to maintain a separate premium track of housefile mailings to maximize the value of our premium- acquired donors.
IN-DEPTH ANALYSIS
But a deeper analysis showed us that, contrary to our assumption, those donors who gave to only housefile premium mailings were barely pulling their own weight. For premium acquired donors to be valuable, it was critical to convert them to give to both premium and nonpremium appeals. For example, after a year on the file, the average total lifetime giving of a premium-acquired donor who gave only to premiums was $22. For a premium-acquired donor who gave to both premium and non-premium appeals, the average total lifetime giving was $105—a staggering difference.
And, in the same way, the most valuable non-premium donors weren’t those who gave only to non-premium appeals. They also began giving to our premium housefile mailings, as well. After a year on the file, non-premiumacquired donors who gave only to non-premium mailings had an average lifetime giving of $124, compared to $202 for those donors who gave to both premiums and non-premiums.
This discovery solidified a shift in our strategy. Instead of immediately trying to segregate new donors into premium and non-premiums tracks, we made sure that each type of new donor was exposed to as many different mailings as possible. It was only after several years of giving history that we would dub a donor as a “premium only” or “non-premium only” giver.
After all of these changes to our acquisition and housefile mailings—which took several years to put into place—we ended up with a much more robust direct mail program. At the start of the period, we were mailing 2.6 million names in acquisition a year, over 80% of them with a greeting card package. We were acquiring well over 100,000 donors every year, but only recouping about 50% of our acquisition investment in the subsequent 12 months.
At the end of the period, we were mailing five different control packages, with 57% of our acquisition volume in premiums and 43% in non-premium packages. We were acquiring fewer donors than before—around 80,000 a year—but our average new-donor value at the end of 12 months had more than doubled. This meant we were consistently growing Mercy Home’s direct mail net revenue year after year—while recouping our initial acquisition investments in less than 12 months’ time.
In the end, the daring step of striking premiums from our acquisition program— which so many nonprofit direct mailers have dreamt about—didn’t quite pan out. But it paved the way toward building a direct mail program with a much more solid footing. In that sense, in spite of all the anxious moments and failed expectations, it was a grand experiment after all.
Mercy Home New Donor Acquisition Trends
|
Year |
Response Rate |
Average Gift |
Net Cost to Acquire a Donor |
% of Initial Cost Recovered after 12 Mos |
| 2000 | 2.9% | $9.17 | $13.35 | 63% |
| 2001 | 2.5% | $9.46 | $14.73 | 69% |
| 2002 | 1.1% | $15.70 | $26.00 | 77% |
| 2003 | 1.5% | $14.94 | $18.50 | 93% |
| 2004 | 2.2% | $12.43 | $11.45 | 105% |
| 2005 | 2.0% | $13.10 | $11.92 | 110% |
Peter Schoewe is Senior Consultant,Mal Warwick Associates, 2550 Ninth Street, Suite 103, Berkeley CA 94710-2516, phone (510) 843-8888, fax (510) 843-0142, Web www.malwarwick.com, e-mail peter@malwarwick.com.
Biggies
Who sends the most direct mail in the country? A survey by Target Marketing reports that 17 of the top U.S. mailers by volume are nonprofits. These not-for-profit leaders, in alphabetical order, are: AAA, AARP, American Diabetes Association, American Heart Association, American Red Cross, Consumers Union, Cystic Fibrosis Foundation, Disabled American Veterans, Leukemia & Lymphoma Society, Memorial Sloan-Kettering Cancer Society, National Geographic Society, National Wildlife Federation, Paralyzed Veterans of America, Republican National Committee, Salvation Army, Smile Train, and U.S. Fund for UNICEF.
Perhaps you’re asking yourself, “Who cares?” If so, consider this: For a relatively low cost, you can contribute your way onto some or all of these organizations’ donor lists. In return, you’ll get a ready source of professionally prepared direct mail samples.
Volume
In the past year, over 114 billion pieces of direct mail were sent to Americans, according to the U.S. Postal Service, as reported in The New York Times. That's an increase of about 15% from five years earlier. What's more, the volume of bulk mail (meaning all direct mail) surpassed first-class mail for the first time. Interestingly, the newspaper also says direct marketers are finding that many people prefer the mail to e-mail spam or phone solicitations.
Givers
A study conducted by the Center of Philanthropy at Indiana University, as reported in The NPT Weekly, found that high net worth households - comprising just 3.1% of total households in the U.S. - donated an estimated $126 billion to charities in 2005. That represents more than half of all individual charitable giving. What motivates these wealthy donors? They say they're more interested in meeting critical needs and giving back to society than in leaving a legacy.
Tops
Donations to the nation's 400 largest charities rose 13% in 2005, according to the latest Philanthropy 400 survey condcuted by The Chronicle of Philanthropy. And which not-for-profits did the best? In order, here's the Top 10: United Way of America, Salvation Army, AmeriCares Foundation, American National Red Cross, American Cancer Society, Fidelity Charitiable Gift Fund, Gifts In Kind International, YMCA, Feed the Children, and Tulsa Community Foundation.
Sign Up