Better Online Fundraising with A/B Testing

We’ve all heard the colorful proverb about assumption and its children. As fundraisers, we make assumptions every day about current and potential donors, what factors motivate their behaviors, or what message will be most effective for a given audience. The truth or falsity of these assumptions, and the success or failure of the campaigns that follow, is usually anecdotal and a risky substitute for real data. In short, the holy grail of fundraising is valid, useful data about who is likely to donate and, most of all, why they choose to do so.


Cutting through the clutter of anecdotal evidence and broad assumptions means getting scientific about your fundraising efforts. Then, like all science, fundraising becomes a cyclical process of gathering data and testing new ideas. For a long time, data gathering was the most difficult part of the job. As a result of the constantly improving, data-gathering mechanisms of social media and internet search engines, the work now comes in using these technologies to turn data into actual discovery, and eventually new knowledge about your supporters and their motivations.


The use of A/B testing is nothing new, but the opportunity to carry out A/B tests quickly and cheaply via social media, and to solve the data collection problem with free tools to help you follow the whole donor conversion process, is revolutionary. Let’s illustrate a successful A/B test with a quick example. First, we need an organization, which we’ll call Save the Kitties, surely a nonpartisan issue if ever there was one. So, Save the Kitties wants to know what motivates donors more: feeling sorry for the kitties or wanting to adopt the kitties. Now, we’re ready for an A/B test.


For our hypothetical, ridiculous A/B test, we will use Facebook simply because they are the biggest show in town. Next, we will create two posts or ads. In one we will show the image of a poor little kitty with no home looking sad, possibly in black and white, maybe in the rain. In the other, we will show a cute kitten being held by a little kid. For both we’ll use the same message: Donate now and save a kitty. Then we’ll send one ad to one group of users and the other ad to another group of very similar users.


By adding our nifty (shameless plug) Raise The Money contribution form to your Facebook page, users can make a donation without leaving Facebook, and through our easily installed tracking software, you can track each donation and donor. Make both ads point to the donation page. Then, sit back and watch. By combining Facebook and Raise The Money tracking software, you can track the behavior of both groups and know which image yielded more donations, which image brought more traffic, and so on. Finally, you need to turn these results into something useful.
If you notice that people are more likely to respond to the image of the sad kitty (and they probably will), you’ll know that it’s a better prospect to spend your advertising dollars appealing to your audience’s empathy. Taken together, these kinds of insights inform the way your organization should engage its audience. Continued A/B testing, combined with tracking software working across all platforms in conjunction with conversion tracking codes, Google Analytics and a little javascript, will help you refine your message or split it for two distinct audiences. At the very least, A/B testing gives you something to stand on besides assumptions and experience. The proof is always in the data.