Live Google organic results
Position, title, URL, domain and snippet for any keyword — the real SERP as it stands today, not a model’s memory of it.
Connect SerpApi to dolv with your own API key and every brief, blog draft and Director answer is written against the results that actually rank today — cached about seven days per keyword so it barely touches your quota.
A language model’s idea of “who ranks for this” is months stale and frequently invented. SerpApi gives dolv the current page, so the work downstream starts from fact.
Position, title, URL, domain and snippet for any keyword — the real SERP as it stands today, not a model’s memory of it.
See which domains actually own a phrase before you write against it, so the brief targets a gap you can win.
Read your SerpApi plan usage from inside dolv, so you always know what a run will cost you.
Every lookup is cached per keyword for about a week. SERPs move on a weeks scale, so repeat questions spend none of your credits.
The Director’s own tool: ask “who ranks for warehouse AI?” in chat and it returns the live top results, grounded and cached.
Runs immediatelyA generated SEO brief is built against the pages that actually rank, instead of a generic outline that ignores the SERP.
Runs immediatelyThe automated content pipeline consults the live SERP while choosing and shaping a topic, so drafts start from reality.
Human-in-the-loopTurn “we should rank for this” into tracked work that auto-closes when the piece ships.
Runs immediatelyIn Settings → Integrations → SEO & Analytics, paste your own SerpApi API key (serpapi.com → Your Account → API Key). dolv is bring-your-own-key here: the calls run against your SerpApi plan, on your quota, so you keep full control of what search data costs.
It is designed not to. Every lookup is cached per tenant per keyword for about seven days, so asking the same question again spends nothing. Results are also clamped to a maximum of ten per query, and only the top few are passed into a prompt. SERPs move on a weeks scale, so a week-old cached result is still an accurate one.
Three things. The Director can look up live results in chat with its search_serp tool. Content briefs are built against the pages that genuinely rank for the target phrase. And the automated blog pipeline consults the live SERP at its keyword step, so a generated draft is grounded in the real competitive landscape rather than a plausible guess.
Generation carries on. SERP grounding is deliberately fail-soft: an unreachable provider, an API error or a timeout returns nothing rather than raising an error, so a missing SERP never breaks a brief or a draft. You simply get an ungrounded result instead of no result.
No — either one works, and dolv prefers SerpApi when both are connected because its results carry titles and snippets rather than domains alone, which grounds writing better. Everything downstream depends on the integration layer, not the vendor, so you can switch providers without changing how anything else behaves.
Start free for 14 days — bring your own SerpApi key and ground everything you publish.