Note: This offering is currently in closed beta, available to select Pendo customers for testing. The information that follows isn't a commitment, promise, or legal obligation. The development, release, and timing of any features or functionality described here are subject to change at the discretion of Pendo, which can occur without notice. If you're interested in getting early access and providing feedback, contact your Pendo account representative.
AI tagging uses Leo to scan your application, identify untagged pages and features, and suggest tags for you to review and apply — reducing the time and manual effort required to build out tag coverage.
Overview
AI tagging is available inside the Visual Design Studio (VDS). When you launch it, Pendo navigates through your application, captures the DOM of each page, and uses AI to generate tag recommendations. Each recommendation includes a proposed tag name, rule, description, product area, and page-to-feature link. You review all suggestions before anything is applied to your subscription.
AI tagging can also identify tags that are broken or no longer in use and suggest them for deletion, helping you keep your tag library accurate and well-maintained.
Use cases
AI tagging is most useful when:
- You're starting with a largely untagged application and need to build out coverage quickly.
- You're auditing an existing tag library for gaps, broken rules, or stale tags.
- You need consistent naming and structure across tags without applying rules page by page.
Better tag quality benefits your whole subscription — more accurate analytics and more precise guide targeting flow from well-maintained tagging.
How AI tagging works
AI tagging captures the DOM of your application's pages through the Visual Design Studio and analyzes the page structure to generate tag suggestions. Suggestions are organized by page and grouped by action type: tags to create, update, or delete.
AI tagging runs up to five pages simultaneously. Additional pages are queued and processed after the initial set completes. Suggestions stream in as they're generated — you don't need to wait for all pages to finish before reviewing.
If you've added a naming convention, Leo applies it when generating suggestions. Otherwise, Leo uses its own naming defaults. If you've dismissed a suggestion, it won't be offered again in future sessions.
Leo uses a prioritised rule set to determine tag quality: custom HTML attributes first, then element IDs, then weaker selectors. If only a weak selector is available, the suggestion is flagged — you can still accept it, but it's worth asking your engineering team to add a proper identifier.
Before you begin
To use AI tagging, you need:
- Access to at least one AI provider: Gemini is available to all data regions; Claude is available to the US data region only.
Tag with Leo
Use AI tagging with Leo to automatically identify and tag pages as you browse your application in the Pendo Designer.
Open VDS
- Go to Product > Pages.
- Select Tag Pages.
- In the dialog, select your preferred tagging mode:
- Tag side-by-side: Opens the VDS in a new window alongside your application.
- Tag in an overlay: Opens the VDS on top of your application.
- Select Launch Designer. This opens your application in a new tab with VDS open.
- You can revert to classic tagging at any time by selecting Switch to classic mode.
- Select Add/Edit your naming convention to instruct Leo on how to name pages and features. This is recommended if you already have a lot of manual tags. This opens the Naming convention page in VDS.
- In the text box, describe your team's naming convention, including instructions on how to handle ambiguous tags.
- Review your instructions, then select Confirm to apply changes.
- The Heatmap, which is also available in classic tagging, is enabled by default. It highlights popular elements in your application that should be tagged.
Configure your session
Before you start capturing tags, you can adjust how Leo analyzes your pages and features.
- In the Manage tags with Leo window, select Settings.
Set your preferences:
- Suggest Product Areas: Leo assigns Product Areas to your pages and features. On by default.
- Fast mode: Sets AI to be most performant. On by default.
- Multi-page capture: Lets you capture multiple pages in a single session. Off by default. Enable this if you want to tag more than one page before reviewing Leo's suggestions.
- AI model: Choose the AI model Leo uses: Claude or Gemini. If you're in a data region outside of the US, you can only use Gemini.
- Dry run: Dry run is always on — no tags are created or modified until you confirm your selections.
- Debug: Capture error details to help troubleshoot tagging issues.
- After completing the settings, select anywhere on the VDS to close the settings.
Tag pages and features
- Select Start capturing pages.
- If you enabled Multi-page capture in settings, browse through the pages of your application you want to tag.
- When you're done, select Stop and tag with Leo. Leo begins to populate the VDS with suggested tags.
Review suggested tags
In VDS, you see a list of suggested new tags or updates for existing tags. You can view the element in the image preview box to understand which feature Leo has tagged. Suggestions are grouped by page, with site-wide features listed first, then subsequent pages as collapsable sections in the VDS. When you expand a section, the page is listed first, then any suggested features follow.
New tags are indicated by the green band and are under the Create heading. Tags that require updates are indicated by the yellow band. Each suggestion shows the proposed tag name, element type (button, link, or input), rule, description, and product area, if enabled in settings.
- Expand a tag to see Leo's AI-generated description and rule.
- For each feature, you can review:
- The element type.
- Leo's suggested description.
- Leo's suggested rule.
- Select the checkbox next to any page or feature you want to tag. You can also use Select all to select everything in the list.
- When you're happy with your selections, select Accept selected.
- To discard your selections without saving, select Reject selected. To start over, select Clear selection.
Review tags for deletion
AI tagging surfaces tags that are broken or no longer in use alongside creation and update suggestions. These appear as deletion recommendations.
- Review deletion suggestions.
- To accept a deletion, select the tag, then select Accept selected.
- To keep the tag, select it, then select Reject selected.
Deleting a tag removes it from your subscription and any reports or guides that reference it. Review deletion suggestions carefully before accepting.
Frequently asked questions (FAQ)
Does AI tagging work with tag maintenance?
Yes. AI tagging builds on tag maintenance. If Leo detects a broken tag or one with no recent DOM match, it surfaces it in the suggestions panel as an update or removal recommendation alongside suggestions for new tags.
Will Leo only suggest tags with strong selectors?
No. Leo uses a prioritised rule set: custom HTML attributes first, then element IDs, then weaker selectors. If only a weak selector (such as an nth-child) is available, Leo flags the suggestion rather than skipping it. You can still accept it, but it's worth asking your engineering team to add a custom HTML attribute or element ID for a more reliable rule.
Can I choose which suggestions to apply?
You choose. Use the checkboxes next to each suggestion to select what to accept or reject, or use Select all to apply everything. Nothing is committed to your subscription until you select Accept selected.
Does AI tagging work on extension apps?
Not yet. Extension support is on the roadmap as a future milestone.
Will Leo's suggestions appear in the heatmap?
Yes. The heatmap shows a green/red overlay of tagged versus suggested elements, with an optional usage data layer on top.
Can I update my naming convention without changing existing tag rules?
Not in the current version. Adding or changing a naming convention may also trigger updates to descriptions and other tag metadata, not just names. Scoping naming convention changes independently is on the roadmap.
Is AI tagging the same as Leo?
Leo is Pendo's intelligence layer. AI tagging is one capability within it — specifically, the feature that suggests and maintains tags in the Visual Design Studio.
Does using Gemini require a separate data processing agreement?
No. Gemini is used within Pendo's existing data processing framework. If your subscription already covers Pendo's standard data processing terms, no additional agreement is needed.