The Failed World of One Size Fits All “Nudges”

April 23, 2021      Kevin Schulman, Founder, DonorVoice and DVCanvass

I don’t know how many times I’ve typed, spoken, yelled or cried in my beer (ok, whiskey..) that we have to get fundraising out of it’s one size fits all world.   Actually, I do know because I’ve just counted and its 3,247 times.  I’m clearly doing it wrong.

But, hope springs eternal and so, here goes 3,248.

People are wildly different. But they’re predictably so, and that difference typically has very little to do with what they look like and more to do with who they are.

Having said that, another powerful force is context and situation.  The exact same person can make two different choices depending on who is offering and how it’s presented.  Consider HIV testing.  The CDC has recommended setting the default “ask” to opt-out because people tend to accept the default and thus, more people get tested.  But, the CDC evidence for this is paper thin.

Studies from in-market, emergency room testing show the acceptance rate to be tested for HIV using opt-out forms ranging from 29% to 87%.  Heck, a single emergency room ran two tests and found opt-out got much higher acceptance than opt-in in Test One but the exact opposite in Test Two.

So what’s going on here?  Context and situation trump the default bias of going with whatever is put in front of you.   But, there are even more layers to this because not all emergency room patients respond the same way to the same context.  People differences matter a lot.  Context matters a lot.

These correlates are ever-present for your fundraising.  If you aren’t paying attention to context and people differences then you are leaving money on the table.  How do you know?  If you have one of everything – one control DM pack, one main donate page, one test (or 100) that goes to a random nth because you assume one audience – then you are leaving money on the table.

Back to our HIV test experiments.  Opt-in vs. opt-out is antiquated or at least incomplete thinking.  What about active choice for people at-risk of HIV, patients choosing whether to be tested or not?   The research on motivation is crystal clear, a sense of control or autonomy fosters high quality motivation, a kind of internal flame that stays lit.

And isn’t our real aim to have human behavior match the person’s true preference?   Maybe in the health domain and others of societal good and/or the safety of others we require certain behavior (wearing a seat belt, childhood immunization to go to school) but surely that isn’t where we live in the fundraising world where active choice for all choices should be our aim.

In HIV testing that was fully randomized there were three standardized scripts that all led with, “We’re offering routine HIV tests to all of our patients.  It’s a rapid test with results available in one to two hours” followed by the treatment scenarios:

  • Opt-in: You can let me, your nurse or doctor know if you’d like a test today
  • Opt-out: You will be tested unless you decline
  • Active-choice: Would you like a test today?

Imagine this was you. Which version would you prefer receiving?  The results?  Context matters a lot. The person matters a lot; in this case their medically determined risk level for being HIV positive.

Across all patients (the everyone is the same assumption) we see that Opt-Out was greater than Active Choice that, in turn, was greater than Opt-in.   But, high risk patients were far less influenced by context, with no difference between active choice and opt-in/opt-out and only a small difference between Opt-Out and Opt-In.  Compare this to huge context effects for the lower risk patients.

Lessons to apply in fundraising:

  • Context matters a lot.
  • People differences matter a lot
  • Yes, opt-out gets more “yes’s” but does offering that option result in winning the battle to lose the motivation war?   Consider active choice.
  • Your donors aren’t the same.  Demographics are largely irrelevant (this HIV study also found that)
  • Donors more vested in your cause are far less influenced by context – ask string, anchoring, opt-in/out.  Their defaults are their defaults, not yours.

Kevin