Data freshness
- Optimizely Web Experimentation
- Optimizely Performance Edge
- Optimizely Feature Experimentation
- Optimizely Full Stack (Legacy)
Data freshness is the time it takes Optimizely Experimentation to collect, process, and deliver data via the Results page or API. Because data freshness is critical, Optimizely Experimentation has a fast results pipeline that helps you make validated decisions at scale. Results are typically available within 5 to 10 minutes of Optimizely Experimentation receiving the data.
To handle several billion events per day, Optimizely Experimentation’s distributed event infrastructure queues and processes events, then holds the processed events in a low-latency data store. This infrastructure ensures that events are ready for queries when you visit the Results page. After attributing events to sessions and users, Optimizely Experimentation applies statistical tests to the data. Events display as metrics for business decisions.
Optimal data freshness offers these advantages:
- Early validation – Confirm experiment setup and mitigate risk using first results.
- Support for time-sensitive decision-making – Use this information for things such as headline testing, high-traffic testing, holiday sales, and testing during critical news cycles.
- QA without waiting – Send events to confirm your experiment setup.
Determine your results data freshness
Check the freshness of your current data on the Results page. The Last update and Last event headings at the top of the page list the date and time of the results you are currently viewing.
A banner appears at the top of the page when fresher results data are available. Click the link in the banner to reload the page and view the new results data.
Data freshness throttling
When an experiment or campaign experience reaches a sample size of 1 million users, Optimizely Experimentation may throttle the refresh rate to up to one hour in order to maintain the highest accuracy of results.