Stick with this post.  By the end –following a somewhat wonky start –you’ll feel more control over your fundraising and relatedness to your donors.

This is what the vast majority of giving formulas, albeit never expressed, look like:

Giving = solicitation + random error (difference between your budgeted number and reality)

(Remember algebra?  Don’t stop reading;  this will be worth it and if not, we’ve added a Zoom video gaffe at the end, not really…but you know you’ve been watching them.)

This is the (simplified) formula for a straight regression line (giving isn’t really a straight line, indulge us a bit) converted to our fundraising world.

Y=mX + e

Y = Giving/money raised

X = a code indicating whether which donors received the test or the control you’re comparing it against (e.g., X can be coded 1 for donors receiving the test and 0 for those receiving the control).

M = how effective your solicitation is (e.g., response rate) compared to the control

E = this is your error or how far off your projection on revenue is from reality.  It’s all noise or is it? (foreshadowing).

You know what’s missing?  The flipping donor.  You know where the donor is hidden?  In that unflattering, dismissive, ignore it, can’t control it error term.

 Why Understanding Error is Important

The error is where all the human differences exist; those differences are what cause some people to like your appeal and others to not like it.  That different reaction by different people gets lumped into the term “error” – i.e. our inability to explain why the test didn’t work.

It’s true that your test might have failed because it was a crappy idea that pretty much nobody liked, or it failed because it worked differently for different people. Those are wildly different explanations that usually get mushed together and lost because our approach to mass market fundraising is based on a single, massive assumption that everybody is the same. (Paradoxically most folks claim everybody is not the same, but their actual practices say otherwise.)

We run A/B tests with the ‘gold’ standard being that the audiences for the test and control groups are the same as each other and the same within each group.  “Everybody is the same”  is the cornerstone of our work.

How does this practice reconcile with being ‘relevant’ or ‘donor-centric’ or desiring of relationship building or aspiring to have kick-ass automation and trigger -based,  gee-whiz CRMs that claim 1:1 fundraising if our entire operation is built on the premise that everyone is the same?

A Better Formula

Here is a giving formula that you, as a fundraiser who believes in digging deeper, can get behind.

Giving = solicitation context + Identity+ (Solicitation x Identity) + “personality” + (solicitation × “personality”) + random error

Context and Identity and Personality are all about our donors.

• Identity is the donor’s personal narrative that creates connection to the organization’s mission (e.g. dog lover gives to an animal charity, person with Disease X gives to a charity fighting disease X, outdoor person gives to a conservation group).
• Context is a nod to behavioral science and the fact that how people process information is heavily dependent on context and immediate, short-term (i.e. in the appeal, on the reply form) decision making.
• Personality is another highly innate trait that is, among other things, about relevance. Can you message to me in a way that matches my disposition?

The multiplier, “x” (e.g. Solicitation x Identity) part of the equation is called an interaction effect but for our purposes, it is the essence of getting beyond a limited, everyone-is-the same world view that comes up with a single test, sent to a random nth test group (everyone-is -the-same thinking) against the control.

This formula gets us to a new, differentiated, donor-centric world view that says my “Y”( giving) can be increased if I come up with a test that is aimed at a subgroup of donors who I think exist within my mass market of donors or prospective donors.

The outcome of this approach is multiple controls but not the traditional “co-controls” that are still sent to everybody on some rotated basis but instead, different offers (controls) for different people.

Is there a practical, real example of this?  Excellent question.

A Practical Example

This testing displayed below was done for a food bank.  On the right is an “everybody is the same” control offer that was run in in the local geography where the food bank operates.  The ad on the left is a solicitation × “personality” test whose hypothesis and premise is all about the donor – it has nothing to do with the charity it all.

We know that our own personality dictates a lot of our opinions, beliefs and choices.  If the message matches my personality, I’m much more likely to pay attention.  In this case, we chose to message to Conscientious people with the very slight copy change highlighted in yellow:  Help hardworking Americans at the top of the ad and the “American families” at the bottom of the ad.

Personality also tends to show skews tied to macro demographic factors that we use as proxies to increase the efficiency (not effectiveness) of our targeting.  In this case, we know that Conscientious people (one of the Big 5 personality traits) tend to skew Conservative.

The results show that directing the Conscientious ad to cities that voted Republican in the last gubernatorial election produces a 28% lift.  To reinforce the validity of this approach, we intentionally mismatched the Conscientious message (that plays better with Conservatives on average) to a city that voted for the Democrat.  In this case, the ad did 9% worse than the control.

But, we’re not done.  This next test has even more donor-centric thinking with our giving formula focus on solicitation × “personality x Identity. 

Having done work with food charities in the past we know many people give when wearing their “Parent” Identity hat because they see helping less fortunate children as being in line with their parental values.  To further aim this ad at the audience we believe exists – Conscientious Parents – we added the copy in blue to the winning Conscientious ad.

This combination of donor-centric, theory based messaging lifted response 88% over the control (and better than Conscientious alone or Parent alone, which we also ran).

Beyond Acquisition

This exact same approach can be replicated in direct mail.  The people responding to this ad or it’s direct mail cousin are tagged in the CRM as Conscientious Parents.  The subsequent house file communications will tailor content to who they are as people and what causes their support.  This has nothing to do with the charity and everything to do with the donor.

In a nod to our favorite kind of data,–zero-party—  we do use subsequent communications to try and affirm, via direct measurement (with survey), the personality and parental status of every newly acquired donor because we always want to move beyond proxy data and mere assumption to individual level, willingly provided, zero-party data.

It’s way past time to stop pretending our practices are anything remotely close to donor-centric when we operate in a world that is built on the assumption nobody believes – everybody is the same – and treat donor differences as an inexplicable ‘error’. 

Giving = solicitation + random error  

Fortunately, there is better, more accurate giving formula that puts donors at the heart of the business.

Giving = solicitation context + Identity+ (Solicitation x Identity) + “personality” + (solicitation × “personality”) + random error

The reason this works is that theory is the driving force behind it.  Any non-theory approach to slicing and dicing your donors by channel or demographics,  or worse yet, the horrific clustering solutions that throw everything in a statistical blender, will cost you time and money and little by way of return.

Kevin