How GoFundMe Uses Data to Drive More Donations

GoFundMe, the world’s largest free social fundraising platform, is creating the “giving layer” of the internet. Regardless of whether you’re an individual, a team, or an organization, GoFundMe is providing the easiest, fastest and safest way to raise money to help people and causes.

That said, product iteration at GoFundMe could be slow at times due to a time-consuming data analytics process. A busy data analytics team had to balance other requests with writing SQL queries to create user funnels and cohorts. That was, until the company turned to Amplitude. Within months of using Amplitude to examine the performance of product experiments, GoFundMe was able to double-down on promising product bets, and as a result achieve significant gains in the number of overall donations on the platform.

Before implementing Amplitude, GoFundMe’s small team of data analysts handled internal stakeholder queries using another business intelligence platform. Due to the data team’s bandwidth constraints, turnaround time was anywhere from a day or two to a week or more. This process slowed down product ideation and iteration. “Any kind of analysis of tests took time,” Ran said. “When the process is slow, you’re just not going to be as open to trying new things.”

That changed, she said, with Amplitude. With its real-time data insights, Amplitude enabled GoFundMe to kick off a new period of experimentation as part of a shift in strategy toward optimization, going from two or three to 10 tests a month. Leveraging Amplitude’s Segmentation, Funnels, and AB Test View, Ran’s team can instantly understand their experiments’ performance. “Because it’s a much more efficient process, you are able to be more open to new ideas,” Ran observed. “As a product team, we no longer say, ‘Oh, this will take a long time. We can’t help you test this hypothesis.’”

Because Amplitude is self service, it allows product managers to answer the majority of their data questions on their own, without adding to the data team’s workload. GoFundMe’s data team can now focus more on deep dive analysis instead of spending a lot of time writing SQL queries for one-off requests.

One of GoFundMe’s most successful experiments revolved around sharing campaigns, an essential component of online fundraising. These experiments included offering users more encouragement to share and a greater variety of ways to share. When users share a campaign, it brings in more money for the campaign organizers. By testing different sharing variants and using Amplitude Cohorts to understand the impact, GoFundMe was able to increase the percentage of campaigns shared by 10 points. The result: a five percent boost to GoFundMe’s gross donation volume (GDV).

Using Amplitude’s precision tracking and flexible segmentation, GoFundMe also experimented with a dozen changes to its campaign page—precious real estate seen by anyone who comes to the site—by tinkering with the copy, design and layout of the page. The tests paid off, boosting campaign shares and donations to the tune of a four percent increase in GDV.

GoFundMe found even seemingly subtle changes—the kind GoFundMe would likely not have tested given the time and effort involved—can produce results. Ran pointed to a series of experiments with button color and style that yielded a two percent increase in the donate button click rate, and a three percent increase in the share button click rate; gains that exceeded expectations. “We’ve learned that combining a number of small changes, at scale, can generate really, really great results,” she said.

Ran and her team of product managers were the first GoFundMe employees to use Amplitude. But it didn’t take long before a large cross-section of the company was using it as well.

“We would sit in meetings where a cross-functional partner would ask, ‘What about this?’ And I’d go into Amplitude and, within a minute, it would generate a chart. I’d say, ‘Here’s the answer,’” Ran said.

The result, Ran said, was greater transparency into GoFundMe’s metrics. In turn, that has meant more productive conversations among departments. “Meetings revolve more around data now,” she said. “We often even create charts in real-time in the meeting when we are making a decision.”