2025 Optimizely Analytics release notes

  • Updated

Follow this article to receive email notifications about Optimizely Analytics and Warehouse-Native Experimentation Analytics updates for 2025.

June 16, 2025

  • Selector performance improvements – Load selectors faster, especially when working with a large number of entities such as columns, datasets, or metrics.
  • Selector search improvements – Search more easily—even with typos or partial names—to quickly find what you need.
  • Parameterized rolling window – Set rolling window lengths in moving aggregate formulas for greater flexibility.
  • Percentile aggregate function – Calculate the value below which a chosen percentage of your data falls, helping to identify distribution patterns and understand typical versus extreme user behavior. 
  • Improved unsaved changes messaging – Keep track of unsaved edits with clear, actionable messages.

Warehouse-Native Experimentation Analytics

  • Experiment on anything – Use Optimizely’s Stats Engine to analyze experiments from any source, as long as variation assignment data is in your warehouse.
  • Outlier management – Boost result reliability by capping extreme values at configurable thresholds to prevent a few edge cases from skewing experiment outcomes. 
  • Sample ratio mismatch check – Detect imbalances in experiment variation traffic, helping troubleshoot implementation or instrumentation issues early.

May 8, 2025

  • Visibility controls – Show or hide parts of your visualization without rerunning your query or changing your selection – especially useful for experiments with many metrics, variations, or crowded explorations.
  • Rounding capabilities – Edit your visualization and round large values into millions, billions, or trillions to keep data readable. For example, render 1,500,000 as 1.5 M.
  • Quarterly time grains – Use the Last tab of the date selector to group data by calendar or custom fiscal quarter and years.
  • Quarter to date, month to date, week to date time ranges – Use QTD, MTD, or WTD presets from the Since tab of the date selector to apply common look-back windows.
  • Key-pair authentication support for Snowflake – Support for key-pair authentication is now available alongside username-password login.

Warehouse-Native Experimentation Analytics

  • % of Baseline – Compare how each variation performs against the baseline in the Experiment section of Event Segmentation, Universal, and Funnel templates.
  • Conversion windows – Set a conversion window to control how long user behavior is counted after they’re assigned to a variation – for example, exclude actions that happen more than a day later to focus on immediate experiment impact.
  • Ratio metrics – Create ratio metrics by dividing one event count or existing metric by another.

March 7, 2025

Warehouse-Native Experimentation Analytics is now generally available. The integration brings the elements of warehouse-native Optimizely Analytics to Feature Experimentation and Web Experimentation. Teams can analyze experiment performance, identify winning variations, and conduct deeper analyses on experiments that ensure data security and privacy and avoid data duplication or movement.

  • Enhance experimentation results by integrating Optimizely Experimentation data with additional insights from the data warehouse. See scorecard for more information.
  • Specify key user interactions to assess engagement and evaluate impact using custom events.
  • Create specific experiment-focused metrics (such as conversion, numeric aggregation, ratio, and more).
  • Segment users by common behaviors into cohorts for precise analysis and targeted insights.
  • Create custom metrics and derived columns to transform data to gain deeper insights.
  • Use the Stats Engine to ensure reliable results and advanced analysis capabilities.
  • Use CUPED to reduce the impact of random variation and surface insights quicker.
  • Switch effortlessly between configuring experiments and conducting deep experimentation analysis from both Feature Experimentation and Web Experimentation.
  • Filter results by user segments, analyze trends over time, and track variation performance through designated funnels via Experimentation Analytics > Explore.
  • Uses the Opti ID Admin Center for user management, giving you a single login point to switch among your Optimizely products. See the Opti ID documentation to learn more about how to use it.
  • Updated the Analytics UI to match Optimizely styling.

Learn more about Warehouse-Native Experimentation Analytics.

January 22, 2025

  • Refreshed chart settings UI – Experience a smoother, more consistent chart editing process with the simplified chart settings interface with two streamlined tabs – Data and Style.
  • Slowly changing dimension(SCD) tables – Use type 2 Slowly Changing Dimension (SCD2) tables to manage historical data changes more effectively with minimal configuration. The following are the key features:
    • SCD2 datasets are modeled with three required fields:
      • The actor dataset (commonly Users).
      • The start time column, which must be a timestamp.
      • The end time column, which must be a timestamp.
    • Automatic integration – SCD2 datasets require only a many-to-one relationship with the actor dataset. Any event stream that joins with the actor dataset (for example, Users) will automatically join with the SCD2 table when configured.
    • Simplified configuration – You do not need to create direct relationships between event and SCD2 datasets; analytics handles it seamlessly.