Data retention policy and FAQs

Optimizely Experimentation has updated the Experimentation Data Retention Policy for event data to maintain the highest quality compliance and testing standards.

Starting October 2022, Optimizely Experimentation only retains active subscription experimentation event data for 18 months.

This means that active customers with experiments with event data older than April 2021 see result page data expire on a day-to-day rolling basis. 

For example, an experiment was run on November 16, 2021. The goal was to increase conversions over Black Friday. The test was successful and paused on December 16, 2021. The 18-month time period would be on May 16, 2023.

If you visited the results page before the retention period expired, the data would reflect the whole month of data.

If you visited the results page one day after the retention period ended, the data would reflect all the data except the data from the first day, November 16, 2021. In other words, the data available would be from November 17, 2021, to December 16, 2021.

If you visited the results page two weeks after the retention period ended, the data would reflect the last two weeks of data captured or data captured from November 30, 2021, to December 16, 2021.

Because this test was over the holiday season, the days that expire can impact the results, as one day of a lot of data missing means the results are not telling the whole story.

The 18-month retention period begins when the first event is processed for an experiment.

The Experiment Results page is still accessible and continues to show data, although data after 18 months is no longer available.

The event data expires daily, meaning the data shown on the page over 18 months is statistically correct, but days are missing from the data set. The missing days can impact the value of the data shown, which is why there is a persistent message indicating there is event data older than 18 months.

Yes, Optimizely Web ExperimentationOptimizely Web Experimentation, Performance EdgePerformance Edge, Optimizely Feature ExperimentationOptimizely Feature Experimentation, and all versions of Optimizely Full StackOptimizely Full Stack are all impacted the same.

Yes, the updated retention policy applies to A/B TestsA/B Tests, Multi-armed Bandit Tests, Multivariate Tests, and Optimizely PersonalizationOptimizely Personalization experiments.

The information displayed on the Results page can be exported with the CSV Export functionality at the top right-hand corner of the page.

There is a reminder message on experiments within 30 days of the retention policy period ending.

No, resetting the results does not reset the retention time frame.

If an experiment began collecting events on November 16, 2021, the retention time frame would end on May 16, 2023. If the results were reset at any time, the retention time frame would remain May 16, 2023, as the experiment started collecting events on November 16, 2021.

No. Experimentation Event Export (E3) files are retained for 12 monthsExperimentation Event Export (E3) files are retained for 12 months and expire on a rolling basis by day. There was no change to the retention policy.

Optimizely does not have plans to offer extended retention time frames, but if it is something that would be of interest, sign up. Optimizely will inform you of changes or news about extended retention time frame options.