SulaHow Sula finds contact info hidden on any page
A contact page that shows you a chatbot usually still contains the real email and phone number. It's just stored where you can't see it. Here is exactly where Sula looks, in the order it looks there.
Everything below runs inside your browser, on the page you already have open. Sula has no database of contacts and no server of its own. If the information exists anywhere in the page, the scan finds it; if it doesn't, Sula tells you that instead of guessing.
1The visible page
First pass is the text you could read yourself: emails, phone numbers, and links to support pages. Sula also decodes the obfuscation tricks sites use to beat simple scrapers, like "jane (at) acme (dot) com", HTML-entity encoding, and the scrambled addresses Cloudflare produces. Numbers only count as contacts when they appear near words like "customer service" or "contact us" — that's how a tracking code avoids showing up as a phone number.
2The data published for search engines
Most companies publish their contact details in structured data (schema.org and JSON-LD) so Google can show them in search results, then leave them out of the visible page. Sula reads that layer directly. The same markup that powers a Google result panel gives up an email address, a phone number, and often the support hours.
3The JavaScript layer, including the chatbot itself
Modern sites ship their content as JavaScript state before rendering it — objects like __NEXT_DATA__ on Next.js sites. Contact details often sit in there even when no page ever displays them. This step also reads the configuration of chat widgets (Intercom, Zendesk, Drift, Crisp, HubSpot, Tidio, LiveChat, Tawk, Freshchat, Olark). Some carry a published support address in their own config. Most don't — and that absence is itself the finding: a company routing every enquiry through a chat widget with no inbox behind it is a company that has decided not to be emailed, which is exactly when the escalation path matters.
4Embedded frames
Contact forms and footers often live in iframes served from the same site. Sula scans those too, with the same rules. Frames from other domains stay off-limits.
5The fallback: fetching the contact page for you
If the page yields nothing, Sula fetches up to three of the site's own likely contact pages (/contact, /support, /help and similar) and scans the responses. These requests carry no cookies, stay on the same site, stop at the first hit, and happen once per site per session. It's what you would have done by hand, minus the clicking.
How the ranking works
Finding twelve email addresses helps nobody. Sula scores each result so the one most likely to reach a person sits on top:
support@,help@, andcare@addresses score high. Marketing, careers, andnoreply@sink to the bottom.- An address whose domain matches the site you're on outranks a stray third-party address.
- Each email also gets a quality label — a named person at a real company domain reads "Direct",
info@reads "Role inbox", and known throwaway domains read "Disposable" — so you can judge before you write.
What never happens
No page content leaves your machine. There's no account, no analytics, and no server on our side to send anything to. Every network request the extension can make, such as fetching a site's own contact pages when the page you're on has none, is listed one by one in the privacy policy.
Free. No account. Works in Chrome, Edge, Brave, Arc, Opera, and Firefox.
Add Sula to your browser