Benchmark your product

Our benchmark analysis relies on Amplitude’s behavioral database and a robust statistical methodology.

We anonymized performance and engagement data from over 2,500 companies across industries, regions, and company sizes. We excluded data from customers who have opted out of this benchmarking.

To ensure data accuracy, we ran rigorous data quality checks, avoided imputation, and applied z-score-based local baseline methods to remove outliers.

A benchmark metric is based on a group’s average performance. Benchmarks are often used to compare different companies or industries, and they can be useful for setting expectations and creating goals.

The median, or 50th percentile, represents a typical product in the selected group.

The 75th percentile represents the top 25% of products. The 90th percentile represents the top 10%.

These percentiles are calculated separately for every metric and filtered group. A product in the top 10% for acquisition is not necessarily in the top 10% for retention.

A negative benchmark reflects a group’s slower growth rate, not a decline or shrinkage. In other words, it doesn’t mean that companies are shrinking—just growing more slowly than before.

For example, if the median company saw 10% growth one month and 6% growth the next, the benchmark would show a -4% change. That doesn’t mean companies are shrinking—it just reflects a slowdown in growth compared to the previous month.

Check out our Product Benchmark Report for more data and trends across acquisition, activation, engagement, and retention.

Amplitude helps teams around the world understand how people use digital products. That gives us a broad view of how products acquire, activate, engage, and retain users. These benchmarks draw on anonymized data from more than 2,500 global companies. While the dataset doesn’t represent every product or user, we believe these benchmarks are indicative of larger trends and offer a useful reference for comparing performance across industries and company sizes.