If there’s one thing that I’ve learned in my 10+ years working in the charitable sector, it is:

You are not your donor.

It can be a hard truth, I know! But at the end of the day, it doesn’t matter what I like, what you like, or what your CEO likes; it matters what donors respond to.

And if there’s another thing I’ve learned, it’s that donors will often surprise you! What works for one organization (or even one group of donors within an organization) may totally bomb for another. So, how do you figure out what makes your donors tick?

Simple: you test it!

If you’ve read my blog, “3 Helpful Tips for Super Awesome Emails,” you know I’m a sucker for a good test. I love tinkering with copy, creative, and gift arrays to find that perfect combination that leads donors to take action and drive revenue. A well-designed test reduces risk in your program, removes external factors from your results, gives you the data to make solid decisions

Plus, testing elements of a fundraising campaign can help you get more bang for your fundraising buck. And in today’s economy, who doesn’t want that?

But testing takes time, energy, and resources. And you are super busy! So, when you invest in a test, how do you make sure it’s going to yield solid results? I’ve got you covered.

What should I test in my fundraising

We don’t run any old test for any old reason. The key to making your tests work for you is to look at what’s not working in your fundraising program – or as we like to say, identify the holes in your bucket. Then, you can develop testing hypotheses around those pesky problem KPIs, and test strategically to drive up ROI on the whole!

For example: Is your direct mail response rate lower than the sector benchmark? Invest in some testing around the outer envelope, letter length, or the stories you’re telling.

Is your Facebook campaign driving a ton of traffic, but the donations just aren’t there? Consider testing your donation page to see if more or less copy, different imagery or messaging, or removing friction from the donation process can increase those conversions. (And don’t forget that integrated campaigns have integrated results – look at the big picture!)

No matter what you’re testing to fix, make sure you’ve laid out what success looks like out of the gate. If your gift matrix test lifts average gift, but suppresses response, will you consider it a win? Is there a threshold of comfort you have? What’s the primary success metric, and are there secondary ones? If you’re struggling with any of these questions, go back to your strategic plan (if you have one!), and look for where your test intersects with your larger program goals.

Strategic testing for fundraising is all about figuring out what your end goal is first, then working backward to see what tests might get you there.

How to structure a fundraising test to get actionable results

two hands with yes and no written on them

The gold standard, of course, is an A/B test. In this test, you take one version of something (A, sometimes called the control) and pit it against a different thing (B, or the test) to see what performs best. In a true A/B test, you should only change one element of the package/email/ad between versions A and B to determine if that change was the factor that increased or decreased results.

You can also do an A/B/C test, or even an A/B/C/D test, where you have 3 or 4 versions running against each other. As you can imagine, this gets a little bit more complicated when it comes to analyzing and interpreting results. Stay tuned for my brilliant colleague Angie Caunce’s blog on that next month!

The key to the success of testing for fundraising is the data. If you serve your Control to one audience, and your Test to another, you’re not comparing apples to apples and your results mean nothing! Ideally, you’ll take a single audience and split it with equal representation (that is, make sure all your most generous donors don’t happen to end up in the same group!).

Pro tip: You don’t have to split 50/50 here! If a test is more expensive or riskier, 60/40 or even 70/30 can enable you to roll out your reliable Control and get the benefit of a test. But no matter how you split, you need to ensure your Control and Test audiences are big enough to give you enough results to provide actionable outcomes.

You also need to make sure you’ll be coding the results correctly. Whether you’re in a small shop and responsible for your own database, or working with a larger data team, it’s wise to make a plan before you test about how results will be inputted, so you can extract them effectively!

This is especially true with tests that are long-term. Testing the OE creative on a single direct mail package is one thing – but testing to see if a phone call to a new donor at 3 months increases second-gift conversion rate is a 24-36 month game, and you don’t want to learn 3 years in that your data is unusable and you wasted all that time and energy! Do this work up front, and thank yourself later.

Got that testing data! Now how do I use it for my fundraising program?

So, you’ve tested out teasers, envelope sizes, social post copy, reply coupon formats, stories, and imagery and you know exactly what motivates your donors to give. Congratulations! You’ve reached the most important step in testing: implementing your findings.

Next month, Campaign Strategist Angie Caunce will take you through the next steps in putting your data to work so you can delight donors and maximize results.

In the meantime, if you’re looking for testing ideas, you can check out Print Production Specialist and Team Lead Nadine Cheney’s blog for more details on things to test in your DM packages, or another blog I wrote specific to Facebook campaigns. And any blog about testing would not be complete without a shoutout to the folks at NextAfter. I will always and forever sing their praises for their work testing everything all the time and then making the results available for all to benefit from – for free!