4 things you can learn from Wikipedia’s fundraising experiments

My name’s Jeff, and I’m an unrepentant test-a-holic. (Hi, Jeff)

I’m addicted to testing—I’m obsessed with finding out what I don’t know, and I love that the internet can help us understand human behavior better today than at any other point in human history.

Today, I want to talk about everyone’s favorite website, Wikipedia. I was browsing a random Wikipedia article the other day (this one, it’s a long story), and I saw a dropdown with a fundraising ask.

No alt text provided for this image

You’ve probably seen these ads, raising money to support Wikipedia, which is totally free (and awesome). And many of us fundraisers have wondered—how much do they raise? What do they know? What are they learning?

Well, good news: they put their annual findings in a long report (on Wikipedia, no less). I combed through the last two published years, and have come out with a few interesting findings that you might be able to put to use for your own organization.

So without further ado, here goes: 4 things you can learn from Wikipedia’s fundraising tests.

1. Know where your money comes from. In Wikipedia’s world, email and desktop banners are king, providing 88% and 87% of the revenue for their campaigns. In 2015, 61% of Wikipedia’s traffic was mobile, but it produced only 12% of the revenue. Email produced more than 3x as much revenue.

Lots of sites now are designed “mobile-first”. Lots of people talk about a “mobile fundraising strategy”. There’s nothing inherently wrong with that—but it’s good to know where the money comes in, and optimize for those situations. Think about the context your donor is in when they receive your email—are they on a mobile device, sitting at a red light (hopefully, at least) where it’s hard to pull out their wallet? Or are they at a desk or at home, where they have time? Mobile is great, growing, and an area to try to optimize. But for Wikipedia, desktop traffic makes the cash register ring. Oh, and email…

No alt text provided for this image

2. Keep asking—but let the data tell you when to quit. Two interesting stats: 74.5% of Wikipedia’s donations from banners came in the first two impressions. But after the 10th banner impression, there was a big drop-off.

There are two ways to look at this—first, you might say “why show banners after the first two impressions?” The answer is that Wikipedia would forgo 25% of their revenue, which came from impressions 3-10.

The other way to look at this would be “why continue to ask people who aren’t giving?” And Wikipedia’s approach is—they don’t. After 10 impressions, most people get dropped. But not because they feel bad about asking, or because some “best practices guru” tells them they shouldn’t, because the data shows it has diminishing returns.

No alt text provided for this image

3. A bailout option might be your best friend. This is really fascinating—Wikipedia started testing the addition of a “remind me later” link that acquired an email and sent a follow-up email at a later time to ask them to give. This an awesome example of a micro-commitment, like Robert Cialdini discusses in his classic book Influence. When people make a smaller commitment, they are more likely to follow through on a larger one. Wikipedia estimates that this button will generate an extra 1.35 million dollars in the next year from this “remind me later” option.

Your organization might not have that scale, but this concept is worth the test—would a deferral link of some sort help convert the traffic that’s not converting right now? Maybe your website visitors can’t give right now—not because they don’t want to, but because it’s not convenient. Give them an option to give you permission to ask later, and see what happens. You might not have the scale to generation a million dollars from it, but you might convert some people who wouldn’t have given otherwise, and you might relieve your prospective donors of guilt along the way!

No alt text provided for this image

4. Use every bit of data you have—personalization pays off. Personalization and segmentation are something you hear about over and over. I certainly talk about them a lot—but for two reasons. First, so many organizations ignore it. And two, it works over and over again!

Wikipedia found that localizing the country where a user was visiting from increased conversion by 11%. That’s pretty basic stuff, but by ignoring it, are you turning down a 11% increase in donor conversion? It’s a big increase in cash in the door at any scale. To be fair, conversion lifts this small can be really hard to validate for smaller organizations without lots of traffic. But it’s a good reminder that you can learn from what other organizations are testing and get a head start.

No alt text provided for this image

I can’t wait to see what test results Wikipedia releases next. As a frequent user (and occasional contributor), I’m fascinated by their scale—they raise $91 million dollars with a $14 average gift! That will keep any fundraiser awake at night.

In terms of traffic, your organization is probably not in the same ballpark as Wikipedia. But take what they’re learning—and freely publishing—and use it to your advantage. Your mission demands that you use every available resource to fund it—and your donors will thank you!

If you need a place to start to optimize your email fundraising, donation pages, Facebook ads, check out the Digital Research Library. Our team at NextAfter has performed more than 2,500 fundraising experiments online to try to understand what makes donors give, and we’ve published the results, data, and analysis of each test for you to read. We’re not experts—far from it. We’re just scratching the surface—please let me know what you are testing!

Leave a Comment