RESEARCH UPDATE: Making Your Match Less Bad

September 6, 2018      Kevin Schulman, Founder, DonorVoice and DVCanvass

I’m temporarily giving up.

We’ve talked about how:

Yet when the end of the year comes around, my inbox will still look like this:

(Why do some of these say Anita or Anita Sue?  I signed up for the top 100 revenue US charities email lists.  I’m crazy, not stupid, so I set up a different account to collect them.  I needed a pseudonym for the project; Anita Sue Donim was born.)

So today I’m temporarily reconciling myself to this matching inevitability.  Players gonna play, haters gonna hate, Taylor Swift gonna shake, and digital fundraisers gonna match.

In that (crushed) spirit, let’s talk about some tips making your match campaign as successful as possible:

Test a hold-out group.  One thing we haven’t talked about yet for matches is how they suppress giving in the long-term.  Researchers matched charitable giving in a randomized experiment.  Not surprisingly, matches won in the short-term (I never said they weren’t effective in the short-term; I just said they were less effective than other things you could be doing with the same money).

What is perhaps surprising is that…  well, I’ll let the researchers tell it:

“in the periods after the experiment, when matching donations have been stopped, the contribution rate declines for the treatment group. The matching mechanism leads to a negative net effect on the participation rate. The field experiment therefore provides evidence suggesting that the willingness to contribute may be undermined by a matching mechanism in the long run.”

Match now, get a gift.  But potentially sacrifice long-term value.  This could be for any number of reasons.  My hypothesis is that we train our donors to respond when there’s a match, meaning they are less likely to give at other times.  And there’s a crowding out effect – if you can have the same impact with $50 as you would have had with $100, why not give $50 instead of $100?  (This is a real effect – see this study as an example)

This may or may not be true for your organization and your donors.  But you won’t know unless you test.  So I encourage you to send a test group a set of emails (and corresponding donation forms) that doesn’t mention a match, then watch them over the next 13 months.  (Usually, you’d follow for a year, but you want to see what happens next year-end, don’t you?)

Try a smaller match.  Surely, more match power = more response rate = better campaign, no?

No.  And don’t call me Shirley.  (I’m Anita!)

A randomized test of direct mail pieces to 50,000 donors found match ratios over 1:1 had no additional positive impact.

OK, you say, I’ll keep it at 1:1.

But I’d like you to think smaller for a second.  What about a 1:2 match?  That is, someone matches one dollar (or pound or Euro or whatever) for every two dollars the person gives.

The Bavarian State Opera House tested this.  They found a 50% match had the same response rate as a 100% match, but higher average gifts.  This is likely because of the crowding out effect mentioned earlier.

Try a different structure.  The same study from the Bavarian State Opera House found that you could have different effects with different matches.  If you are looking to increase response rate, try a fixed gift match (e.g., any gift of any size will be matched by 20 Euros).  While this had lower revenue per piece, it might be worth a try with lapsed givers from whom any gift would be a blessing (and for whom traditional matches don’t work).

Or, try a contingent match.  This is when funds will be matched in and only if X% of donors respond; otherwise it’s just a donation.  The research on this found that such a match, when set at 75% on an online donation form significantly increased the number of people who upgraded to monthly gift over “no match” or a “standard match”. The 75% contingent match also worked for one-time donations.

So the U.S. Olympic and Paralympic Foundation tested this in the mail at year-end.  They found a 26-27% increase in revenue per piece in both current and conversion donors, a similar increase to what the academic research found online.

Get behavioral science in your corner.  Previously, we’ve talked about goal proximity – the idea that people give more as you get closer to a goal.  Think the sprint you give at the last part of a marathon. (Note I say you, not me.  IMHO, If God had wanted us to run 26.2 miles, S/He wouldn’t have invented/let us invent the wheel.)

What if you started your thermometer at 50%?  That is, if people give enough, the first half of the goal is already achieved – we just need you to fill in the second half?  Certainly worth a try given the evidence.

The bottom line to all of this is you need not be a prisoner to last year.  You can experiment with different structures, different formats, and different amounts to get different results.

Maybe you could even do something other than a match that would be more effective, wouldn’t crowd out donations, wouldn’t hurt results in the long term, and maybe even focused on the donor and her/his reason for giving!

Nick

2 responses to “RESEARCH UPDATE: Making Your Match Less Bad”

  1. Jen says:

    Oh, Anita! You’re supposing the HiPPO’s mind is swayed by research and facts….. (highest paid person’s opinion)

  2. Nick Ellinger, VP of Marketing Strategy, DonorVoice says:

    It often isn’t enough. In which case, I recommend http://agitator.thedonorvoice.com/losing-donors-in-the-sea-of-sameness/, which shows an inbox on GivingTuesday — sometimes for the HiPPO who doesn’t go by research, a good story will persuade.

    Failing that, I recommend getting a new set of HiPPOs. 🙂