Open the page you are interested in, open the side panel, and capture. What happens next depends on which site you are on, because the two use different techniques.
LinkedIn capture is AI-driven, not selector-based.
The extension sends the page's title and visible text to the AI agent on the server, which extracts structured lead data and saves it. There are no CSS selectors involved.
That is a deliberate design choice: LinkedIn rotates its class names, so anything selector-based breaks regularly. Sending text to a model that understands it survives a redesign.
Three page types are recognized automatically from the URL:
| Page | Detected as |
|---|---|
| A person's profile | linkedin |
| A search results page | linkedin_search |
| A company page | linkedin_company |
You do not choose — the extension works it out from the path.
Capture reads the text that is actually rendered. On a search or list page, results you have not scrolled to are not in the page yet and will not be captured.
Google Maps
Maps capture is DOM-based, reading the fields Maps exposes directly:
- Business name
- Address
- Phone
- Website
- Rating
Two modes:
One business — open a listing and capture its details.
Visible businesses — capture the results currently shown in the list, for working through an area.
Maps is stable enough in structure that reading it directly is faster and cheaper than sending text to a model.
Where leads go
By default, capture means saved. A captured lead goes straight into your CRM with your configured pipeline, stage and source.
If you turn auto-sync off, captures are held as pending leads and pushed in a batch when you choose. See settings.
With auto-sync off, captures sit in extension storage until you sync. Removing the extension, clearing its data, or moving to another machine loses them. Auto-sync is on by default for exactly this reason.
After capture
Captured leads are ordinary CRM leads. Everything in the CRM applies — enrichment fills in what the page did not show, scoring prioritizes them, pipelines move them through stages, and campaigns can target them alongside every other lead.
When a page will not capture
Layouts change. The extension includes a CSS selector helper, toggled with Ctrl/Cmd + Shift + S, for identifying elements on a page when extraction needs re-pointing. It works on any URL.
That matters mostly for Maps, since LinkedIn does not depend on selectors.
Staying within the rules
LinkedIn and Google both set terms on automated collection and access rate. LeadSearch captures what is on screen in a session you signed into yourself, but staying within those terms — and within data-protection rules for the people whose details you store — is on you and your organization.