Summarize results

5 min

Optimizely Web Experimentation

Optimizely Feature Experimentation

Mark (Optimizely Agent Platform) is an agent orchestration platform that helps you work smarter in Experimentation. The Mark results summary lets you move from results to confident decisions faster. Instead of manually analyzing complex data tables and metrics, Mark automatically summarizes your A/B test results in plain language.

Note

The results summary is only available for A/B test results for Web Experimentation and Feature Experimentation.

With one click, Mark does the following:

Summarizes your experiment results – Describes the impact of each variation in simple, clear language.

Interprets performance across primary and secondary metrics – Highlights key wins and any patterns in user behavior.

Recommends next steps – Suggests data-backed follow-up actions to improve user experiences.

Translates complexity into insight – Replaces statistical jargon with plain-language takeaways.

For example, Mark can do the following:

Translate results into clear takeaways – "Variant B increased conversions by 15%".

Highlight the biggest drivers of success – "The simplified form layout drove the most impact".

Suggest next steps based on the data – "Test this form design on other landing pages".

Identify hidden patterns – "This worked especially well for mobile users".

View your results summary

You can access the Mark results summary for any eligible A/B test directly from the experiment results page. But, before doing so, configure Feature Experimentation for Mark or Web Experimentation for Mark.

Go to your Optimizely Experimentation Results page.

Click Summarize Results.

The summary opens in the Mark chat and displays output similar to the following:

Experiment name.

Dates and how long the experiment ran.

A plain-language description of the experiment's outcomes.

Summary of the audience.

Summary of the variations.

Key metric impacts.

Recommended next steps.

Notable patterns or subgroup findings (such as device type or region).

If you want deeper validation, you can continue to access the full Optimizely Experiment results page and statistical analysis alongside the Mark-generated summary.

Refine the results summary

You can customize the results summary by asking Mark follow-up questions. For example, you can ask it to shorten the summary, tailor it for a specific audience (like a product manager or chief financial officer (CFO)), or highlight a particular part of the experiment.

For example, enter the following in the Ask Mark text field:

Please revise the results summary to be tailored for my company’s CFO. Keep it concise and easy to skim. Focus on outcomes and how the experiment impacted key business metrics or moved the needle financially.

Mark adjusts the output accordingly.

Click Copy to copy and paste the results summary.

Regenerate

To have Mark recreate the results summary, click Regenerate.

Best practices

Start with the summary – Use the Mark-generated overview as your initial guide, before reviewing the full results for critical decisions.

Share summaries with stakeholders – Send the plain-language results summary to help collaborators understand the results without needing a statistical background.

Act on suggested next steps – Use Mark's data-driven suggestions to guide follow-up experiments or optimizations.

Tailor the summary to your needs – Ask Mark follow-up questions to adjust the results summary, such as making it more concise or framing it for a specific audience.

Note

If you use Opti ID, administrators can turn off generative AI in the Opti ID Admin Center. See Turn generative AI off across Optimizely applications.