Significance Explorer

Don’t wait too long for results: explore Bayesian & alternative significance checks right inside Optimizely.

Request connector
Built by
S
Significance Explorer
Works with
Web ExperimentationOpal

Uncover True Experiment Insights: Optimizely Opal Significance Explorer

Stop guessing. Start knowing. Validate your A/B test results with confidence, directly within Optimizely.

Many Optimizely experiments struggle to reach statistical significance, often leading users to seek validation from less rigorous third-party calculators. This "platform leakage" results in missed opportunities and a fragmented workflow. The Optimizely Opal Significance Explorer solves this by bringing advanced statistical analysis directly into your Optimizely environment.

Why the Significance Explorer?

  • Stay In-Platform: Conduct robust statistical analysis without leaving Optimizely, preventing "silent churn" and keeping your data secure.
  • Flexible Alternatives: Access powerful, in-platform alternatives to external calculators you might already use.
  • Remove Temptation: Eliminate the need to explore significance elsewhere, especially on competitors' platforms.
  • Accelerate Decisions: Gain deeper insights into your experiment results faster, enabling quicker, data-driven decisions.

How the Significance Explorer Works:

The Significance Explorer leverages a Bayesian approach to provide comprehensive confidence level assessments for your Optimizely Web Experimentation tests.

  1. Single Metric Analysis

    Dive deep into the performance of a single key metric, understanding its lift, statistical significance, and credible intervals.

  2. Multiple Metric Analysis

    Get a holistic view of your experiment's impact across all relevant metrics, identifying both positive and negative effects.

  3. Clear Explanations & Actionable Recommendations

    Receive easy-to-understand explanations of your results, along with tailored recommendations for next steps, such as implementing winning variations, further experimentation, or deep-diving into user behavior.

  4. Manual Input Capability

    If direct experiment data isn't immediately available, you can manually input key metrics to still perform a comprehensive analysis.

Interact with the Explorer:

Here are some examples of how you can use the Significance Explorer:

  • To analyze a specific experiment:
    I want to explore alternative ways of analyzing the results from an experiment with the ID as 5376433090985984. The Stats Engine says the results are not statistically significant, but I'd like to investigate further.

  • To specify metrics and variations for analysis:
    I want to look at the analysis for the metric PROD - ATC on Quick View PLP, with Original and Variation 1.

  • To get recommendations based on the analysis:
    Would you like an explanation or recommendations based on this analysis? (after an analysis is complete) what are your recommendations for the results shown (after an analysis of multiple metrics)

  • To provide manual input if experiment data is unavailable:
    I couldn't find an experiment with the ID you provided. Please double-check the experiment ID. Alternatively, if you'd like to proceed with a manual input, please provide the following information: (Opal will then prompt for details like visitors and conversions in control and variant groups)

Make every experiment count. Gain unparalleled clarity and confidence in your A/B test results with the Optimizely Opal Significance Explorer.