Fundraising, without the guesswork

Continuously. We run over 50 experiments each year, with 2-10 tests running at any given time, depending on their scale, across different parts of the donation experience.

We measure more than 20 key metrics, such as Average Revenue Per User (ARPU), conversion rate, average donation amount, and many others. Each of these key metrics is further broken down by factors like operating system, country, and browser, providing a deeper understanding of donor behavior. Additionally, we measure test-specific metrics that are unique to each experiment. In total, every decision we make is based on 100+ indicators, ensuring a comprehensive and data-driven approach to optimizing the donation experience.

Only statistically significant findings drive action, meaning every improvement is backed by real data. If a test delivers better results, we roll out the change across the platform, ensuring all nonprofits using Fundraise Up benefit from the latest optimizations.

For results to be trustworthy, the test setup must be rock solid. We ensure fair traffic distribution, typically split 50/50 but sometimes adjusted based on specific needs. Test duration is carefully calculated to reach statistical significance, ensuring decisions are driven by data, not guesswork.

Moreover, to eliminate bias, we use a sophisticated distribution algorithm that assigns users to test groups randomly. This ensures a 0% probability of skewed results.

Statistical significance in an A/B test means that the difference in results between version A and version B is unlikely to have happened just by chance. It helps to determine whether one version is truly better than the other based on the data.

A key indicator of significance is the p-value. A low p-value (typically under 0.05) signals a real effect, giving us the statistical evidence to act.

We include p-values in experiment summaries to help you understand just how meaningful the results are.

We learn from every outcome. When a test shows negative results, we quickly end it and use those insights to guide future development.

Every test we run is designed with security and compliance in mind. We follow strict best practices to protect donor data and ensure a risk-free testing environment.

• Risk-mitigation protocols: We handle sensitive tests with extra safeguards to balance innovation with security.
• Data protection: Donor privacy and security are never compromised, no matter how advanced the testing methodology.

When conducting tests, we always consider the current settings of each nonprofit organization. This means that elements like minimum donation amounts or required address fields remain unchanged to ensure the test does not interfere with an organization’s existing fundraising strategy. Our tests are designed to enhance, not disrupt, the donor journey.