Search-result and SEO research, built in — nothing to connect, billed at cost
Search-result and SEO research is now built in, so the agent can see the market your accounts compete in and not just the accounts. It could previously read your own data but not tell you who ranks for a term, who is buying ads against it, what a keyword is actually worth, or what a competitor's organic traffic earns them. That is now a built-in research family: no account to connect, no key to paste, and every call metered at what it actually costs rather than a flat guess.
What the agent can look up
- Live search results for a keyword on Google, Bing, Yahoo, YouTube, Baidu, Naver or Seznam — the organic block, the paid block (who is advertising against this term), and the People-Also-Ask tree.
- Keyword volume, cost-per-click and competition for up to a thousand keywords at a time, from Google's own planner data — without needing a Google Ads account to ask.
- Keyword research: ideas from a seed set, long-tail suggestions, the related-searches graph, difficulty, and the search intent behind each term.
- What a competitor already ranks for — every organic keyword a domain holds, with position, ranking page and estimated traffic value — plus the domains competing for the same terms and the keywords two domains both rank for.
- Backlink profiles, referring domains, anchors and broken links.
- A live page audit — status, on-page score, headings, links, Core Web Vitals timings, or a Lighthouse score.
- Technology detection, domain registration data, brand mentions with sentiment, and Google Trends.
How it is billed
Every call reports what it cost, and that measured figure is what gets charged — not a flat per-call rate that would be wrong in both directions. A one-keyword lookup costs what a one-keyword lookup costs; a thousand-row backlink pull costs what that costs.
The cost knobs a caller can turn — result depth, row limits, how many keywords per request, how many pages a crawl may touch — are clamped rather than trusted, so a single request cannot spend an unbounded amount. Checking the remaining balance is itself free.
The one thing that touches the outside world
Almost everything here reads data that already exists. One operation does not: a full site crawl actually fetches the pages of the site you name. That one is fenced behind an explicit confirmation, because starting a crawl of someone's website is an action, not a lookup.
Frequently asked questions
- Do I need to connect anything?
- No. This runs on platform credentials, so it works on the first message from a new organization with nothing connected.
- How is it charged?
- At measured cost — each call reports what it spent and that is what is metered. Checking your remaining balance is free.
- Can I get keyword volume without a Google Ads account?
- Yes. The volume, cost-per-click and competition figures come from Google's planner data through this family, so a new organization can research keywords before connecting any ad account.
- Will it crawl my site?
- Only if you ask for a crawl and confirm it. Every other operation reads data that already exists; the crawl is the one that fetches your pages, so it is fenced separately.