Task completion rate calculator
Enter how many participants completed a task. Get the best estimate of the true completion rate, an adjusted-Wald confidence interval that works with small samples, and a comparison with a benchmark.
Free and private: everything is calculated in your browser and nothing you paste is sent anywhere.
How it is calculated
- The confidence interval is the adjusted-Wald interval: add z²/2 successes and z²/2 failures, then use the usual Wald formula. It keeps close to the stated confidence even with very small samples (Agresti & Coull, 1998).
- The best point estimate follows Lewis and Sauro (2006): the Wilson estimate when the observed rate is 50% or less, the Laplace estimate ((x + 1) ÷ (n + 2)) when it is 90% or more, and the observed rate in between.
- The benchmark comparison checks whether the whole interval lies above or below the benchmark.
Frequently asked questions
How do I calculate a task completion rate?
Divide the number of participants who completed the task by the number who attempted it. With small samples that raw percentage is only a rough guide, so report it with a confidence interval like the one calculated above.
Why not just report 100% if everyone succeeded?
With a small sample, 10 out of 10 does not mean every future user will succeed. The adjusted-Wald interval for 10/10 runs from about 68% to 100%, and the Laplace best estimate is about 92%, which is a more honest summary.
What is a good task completion rate?
Across many usability studies the average completion rate is about 78%. Rates well below that usually point to a serious problem with the task, although critical tasks such as checkout or sign-up should be held to a higher standard.