2026 Optimizely Web Experimentation release notes
9 min
Follow this article to receive email notifications when new Optimizely Web Experimentation content is added for 2026.
September
Added a Build agent to the Idea builder, which turns a saved or draft idea into a complete A/B test: Opal creates the variation, then suggests a primary metric and audience so you can review and start the experiment without configuring each component yourself. See Idea builder for Optimizely ExperimentationIdea builder for Optimizely Experimentation.
August
- Added Click event creation in the new visual editor, so you can select an element in the Visual Editor and turn it into a metric without leaving your experiment. Choose Create as experiment-level event to keep the event scoped to this experiment, or leave it clear to create a project-level event other experiments can reuse. See Configure a click event from an experimentConfigure a click event from an experiment.
- Added a Metrics tab to Experiment Config in the Visual Editor, where you pair an event with a winning direction and a measurement to define what an experiment tracks. The first metric you add becomes the experiment's primary metric, which decides whether it wins or loses; any additional metrics become secondary metrics that measure downstream effects. See Add a metricAdd a metric.
July
- Added a JavaScript Library Version setting under Settings > Advanced that lets you run your snippet on ES6 or ES11. Selecting ES11 reduces snippet size and improves performance, but removes support for legacy browsers.
- Upgraded all code editors to the Monaco editor as part of the Code Editor release, giving you line numbers, syntax highlighting, and find and replace. The new Visual Editor also uses the Monaco editor. See New Visual EditorNew Visual Editor.
- Released the Code Editor, which gives you a code-first path to write custom JavaScript and CSS without loading the Visual Editor. Changes sync between the Code Editor and Visual Editor, so you can switch between them without creating separate copies. See Code EditorCode Editor for details.
- Released the Optimizely Web Companion, a Chrome extension that opens the Visual Editor directly from your browser. It lets you launch the editor on sites that strip the URL parameters the editor normally relies on. See Optimizely Web CompanionOptimizely Web Companion for details.
- Added click event creation to the new Visual Editor. Turn any element on your live page into a reusable metric by selecting it, instead of hand-writing the selector. You can view, edit, and archive a page's events from the Attributes panel. See Configure click eventsConfigure click events for information.
- Added the Filter User Agents and IPs setting under Settings > Advanced, which blocks bots and unwanted visitors from being tracked. Matching visitors send no events, so they consume no monthly active users (MAU) and never appear in your results. See IP Filtering: Exclude IP addresses or ranges from your resultsIP Filtering: Exclude IP addresses or ranges from your results.
June
- Expanded Experimentation Events Export to EU-based accounts. EU customers now have data export options with Google Cloud Storage (GCS). See Get Started with Experimentation Events ExportGet Started with Experimentation Events Export for steps.
- Improved the Contextual Bandit Results page to show attribute importance per variation and relative improvement per variation. Each variation now lists its most influential attributes and the values driving performance and you can see how each variation compared to the holdback. See Contextual Bandit Results pageContextual Bandit Results page for details.
- Added one-click variation creation to the development agent, which uses your existing variations to suggest what to build next. Approve the suggested changes instead of re-prompting for each variation. See AI variation development agentAI variation development agent for details.
- Released custom fields, which add required or optional metadata to experiments and campaigns beyond properties such as name and description. Consistent metadata improves filtering, reporting, and collaboration across your experimentation program. See Custom fields in Web ExperimentationCustom fields in Web Experimentation for details.
- Added experience template creation to the Visual Editor, so you can build reusable components without switching to the implementation page. Template creation works the same as before, and templates stay reusable across your experiments and campaigns. See Add templatesAdd templates for details.
May
- Added device emulation in the new Visual Editor, which lets you preview variations at mobile, tablet, and custom viewport sizes. Predefined device profiles, custom dimensions, and viewport rotation are available. See Emulate devices in the new Visual EditorEmulate devices in the new Visual Editor for details.
- Added the ability to duplicate selected element in the new Visual Editor. The action copies the selected element so you can edit and save the copy without rebuilding it. See Duplicate element for details.
- Added inline text editing in the new Visual Editor, which lets you edit text on the page directly. The HTML text field in the Element Change window is no longer required for text-only changes. See Inline text editInline text edit for details.
- The AI variation development agentAI variation development agent is now an Optimizely-managed agent, which locks select configuration fields and scopes the agent to only the tools it needs. This makes tool calling more reliable, and applies changes faster. You can invoke the Optimizely-managed variation development agent only from the visual editor for now.
April
- Released the Optimizely Experimentation MCP serverOptimizely Experimentation MCP server, which connects AI clients to your Experimentation data using the Model Context Protocol (MCP) open standard. Supported clients include Claude Desktop, Claude Code, Claude.ai, Cursor, Visual Studio Code with GitHub Copilot, and Windsurf. Authenticate with OAuth 2.0 through Opti ID and interact with your experimentation data through natural language.
- Released Idea builder for Web ExperimentationIdea builder for Web Experimentation, which uses Opal to generate experiment ideas based on your page, goals, and research. Idea builder helps you skip manual brainstorming and move from page context to a testable hypothesis. Inputs include a target URL or page screenshot, a stated objective, and optional heatmaps, analytics notes, and up to four previous tests. Each idea includes a title, description, problem statement, hypothesis, and source references. Refine ideas in Opal chat or save them to the Saved tab.
July
- Added a JavaScript Library Version setting under Settings > Advanced that lets you run your snippet on ES6 or ES11. Selecting ES11 reduces snippet size and improves performance, but removes support for legacy browsers.
- Upgraded all code editors to the Monaco editor as part of the Code Editor release, giving you line numbers, syntax highlighting, and find and replace. The new Visual Editor also uses the Monaco editor. See New Visual EditorNew Visual Editor.
- Released the Code Editor, which gives you a code-first path to write custom JavaScript and CSS without loading the Visual Editor. Changes sync between the Code Editor and Visual Editor, so you can switch between them without creating separate copies. See Code EditorCode Editor for details.
- Released the Optimizely Web Companion, a Chrome extension that opens the Visual Editor directly from your browser. It lets you launch the editor on sites that strip the URL parameters the editor normally relies on. See Optimizely Web CompanionOptimizely Web Companion for details.
- Added click event creation to the new Visual Editor. Turn any element on your live page into a reusable metric by selecting it, instead of hand-writing the selector. You can view, edit, and archive a page's events from the Attributes panel. See Configure click eventsConfigure click events for information.
- Added the Filter User Agents and IPs setting under Settings > Advanced, which blocks bots and unwanted visitors from being tracked. Matching visitors send no events, so they consume no monthly active users (MAU) and never appear in your results. See IP Filtering: Exclude IP addresses or ranges from your resultsIP Filtering: Exclude IP addresses or ranges from your results.
June
- Expanded Experimentation Events Export to EU-based accounts. EU customers now have data export options with Google Cloud Storage (GCS). See Get Started with Experimentation Events ExportGet Started with Experimentation Events Export for steps.
- Improved the Contextual Bandit Results page to show attribute importance per variation and relative improvement per variation. Each variation now lists its most influential attributes and the values driving performance and you can see how each variation compared to the holdback. See Contextual Bandit Results pageContextual Bandit Results page for details.
- Added one-click variation creation to the development agent, which uses your existing variations to suggest what to build next. Approve the suggested changes instead of re-prompting for each variation. See AI variation development agentAI variation development agent for details.
- Released custom fields, which add required or optional metadata to experiments and campaigns beyond properties such as name and description. Consistent metadata improves filtering, reporting, and collaboration across your experimentation program. See Custom fields in Web ExperimentationCustom fields in Web Experimentation for details.
- Added experience template creation to the Visual Editor, so you can build reusable components without switching to the implementation page. Template creation works the same as before, and templates stay reusable across your experiments and campaigns. See Add templatesAdd templates for details.
May
- Added device emulation in the new Visual Editor, which lets you preview variations at mobile, tablet, and custom viewport sizes. Predefined device profiles, custom dimensions, and viewport rotation are available. See Emulate devices in the new Visual EditorEmulate devices in the new Visual Editor for details.
- Added the ability to duplicate selected element in the new Visual Editor. The action copies the selected element so you can edit and save the copy without rebuilding it. See Duplicate element for details.
- Added inline text editing in the new Visual Editor, which lets you edit text on the page directly. The HTML text field in the Element Change window is no longer required for text-only changes. See Inline text editInline text edit for details.
- The AI variation development agentAI variation development agent is now an Optimizely-managed agent, which locks select configuration fields and scopes the agent to only the tools it needs. This makes tool calling more reliable, and applies changes faster. You can invoke the Optimizely-managed variation development agent only from the visual editor for now.
April
- Released the Optimizely Experimentation MCP serverOptimizely Experimentation MCP server, which connects AI clients to your Experimentation data using the Model Context Protocol (MCP) open standard. Supported clients include Claude Desktop, Claude Code, Claude.ai, Cursor, Visual Studio Code with GitHub Copilot, and Windsurf. Authenticate with OAuth 2.0 through Opti ID and interact with your experimentation data through natural language.
- Released Idea builder for Web ExperimentationIdea builder for Web Experimentation, which uses Opal to generate experiment ideas based on your page, goals, and research. Idea builder helps you skip manual brainstorming and move from page context to a testable hypothesis. Inputs include a target URL or page screenshot, a stated objective, and optional heatmaps, analytics notes, and up to four previous tests. Each idea includes a title, description, problem statement, hypothesis, and source references. Refine ideas in Opal chat or save them to the Saved tab.
March
- Released a new version of the AI variation development agentAI variation development agent which helps you modify and update existing website elements, create new ones, and generate and apply enhancement suggestions, while maintaining brand consistency by automatically retrieving page styles.
- Improved the look and feel of selecting the parent or child element of a selected element. See Edit elements with the new Visual EditorEdit elements with the new Visual Editor.
- Introduced a hotkey (keyboard shortcut) to edit the items behind a hover. See Edit interactive or dynamic elements.
- Released a more reliable way to dispatch events and rewritten some of the logic to collect events that use redirectsredirects. This ensures a more accurate count.
- Added two new utilities to the Optimizely Web Experimentation utils libraryOptimizely Web Experimentation utils library to support experimentation on dynamic websites. To access these utilities, you must initiate a snippet rebuild by making any change to your project. See Snippet order of activationSnippet order of activation for information on when the snippet updates.
- `onUrlChange`onUrlChange – Added a utility that runs a callback function whenever the current URL changes, without requiring a full page reload. This is useful for re-applying experiment changes during page navigation on Single Page Applications (SPAs).
- `persistHtml`persistHtml – Added a utility that re-inserts and persists HTML onto a page if the site's own scripts remove or re-hydrate it. Given a CSS selector and placement position, the function monitors the DOM and automatically re-injects the supplied HTML whenever it is removed. This is useful for maintaining banners, notifications, or other injected elements on dynamic websites during page navigation.
February
- Enhanced the new Visual EditorVisual Editor with smoother highlighting.
- Released drag and drop re-arrange in the new Visual Editor.
- Released the following Web Experimentation system tools in OpalWeb Experimentation system tools in Opal to help you use the AI variation development agentAI variation development agent:
- exp_ve_apply_change – Applies a pending change to the Visual Editor.
- exp_ve_element_glob – Finds elements in the Visual Editor DOM that match a glob pattern.
- exp_ve_element_grep – Searches for elements in the Visual Editor DOM whose content or attributes match a regular expression.
- exp_ve_element_read – Reads the content or attributes of a specific element in the Visual Editor.
- exp_ve_element_tree – Retrieves a partial DOM tree for a specified element and its descendants in the Visual Editor.
- exp_ve_get change_schema – Retrieves the JSON schema for a specific type of Visual Editor change.
- exp_ve_get_selector – Generates a robust CSS selector for a given element in the Visual Editor.
- exp_ve_list_pending_changes – Lists all pending Visual Editor changes for the current experiment or personalization.
- exp_ve_revert_change – Reverts a specific pending Visual Editor change, undoing its effects.
- exp_ve_validate_change – Validates a Visual Editor change object against its schema to ensure it is well-formed.
January
- Added the Experimentation Program OverviewExperimentation Program Overview agent to the Opal Agent DirectoryOpal Agent Directory which creates a report on how your Optimizely Experimentation program is performing using top experiment and win rate data.
- Released the following system tools in Optimizely Opalsystem tools in Optimizely Opal to help you with your Web Experimentation projects. See System tools for Web ExperimentationSystem tools for Web Experimentation for the full list of available Opal tools, including when to use them, variables, and example prompts.
- exp_get_schemas – Retrieves detailed schemas for various Optimizely entities relevant to Web Experimentation. This is a mandatory first step before using the exp_execute_query tool to retrieve actual data about these entities.
- exp_execute_query – Executes a template-based query to fetch specific data from your Optimizely Web Experimentation instance. It lets you retrieve detailed information about various entities.
- exp_program_reporting_top_experiments – Helps you identify and analyze your highest-performing experiments (A/B tests, multivariate tests, and so on) within Optimizely Web Experimentation. It provides a program-level view of success by ranking experiments based on their lift values.
- exp_program_reporting_underperforming_experiments – Helps you identify A/B tests that are not yielding significant or positive results, letting you quickly decide whether to stop, re-evaluate, or iterate on them.
- exp_program_reporting_win_rate – Gives you a high-level overview of the success of your experimentation program by calculating the win rate over a specified period. It is a great way to track the overall effectiveness of your A/B tests.