Company’s B2B Ads

Check what ads a company is running — give it a company domain, get back the B2B ads that company has live.

What it does

• Searches B2B ads running for a company, via LeadMagic.
• Returns normalized ad objects — creative, targeting and performance data — plus resolved company details.
• Takes company_domain or company_name. Pass at least one; the other is optional.
• Domain lookups are more reliable than name lookups. A name has to be resolved to a company first, and common or ambiguous names resolve badly. Pass the domain whenever you have it.
• Cost is per ad scraped — 0.3 credits each — so a heavy advertiser costs more than a quiet one, and a company with no ads costs nothing.

Running ads is a spend signal. A company buying B2B ads has a budget and an active demand-gen motion — useful for prioritising an account list.

Use cases

• Prioritise an account list by spend. A company paying for B2B ads has budget and an active demand-gen motion. Run the list, keep the advertisers at the top of the queue.
• Read a prospect’s positioning before you write. The creative tells you the promise they are making to the market, so your first line can echo it instead of guessing.
• Qualify a marketing-tooling ICP. Ad activity is the cheapest proxy for “they have a marketing team with money” when firmographics alone leave the account ambiguous.
• Time a competitive displacement. A sudden ad push usually follows a launch or a raise — both are windows where a vendor evaluation is already open.

How to use it

Search B2B ads for the company domain acme.com.

For each of these 30 domains, search B2B ads. Flag every company that is
running ads, and note what the creative is promising.

For acme.com, summarise what their ads promise and who they target. Draft a
one-line opener that references it without sounding like I stalked them.

Example prompts

Best prompting practices

• Name the tool. Say b2b_ads_search and “use the Sync GTM MCP server”. “Check what ads they run” is not a tool name, and a client that answers instantly with no credits charged has described the company’s marketing from memory rather than scraping anything.
• Pass the domain, not the name. This is the single biggest driver of hit rate on this tool. company_domain: "stripe.com" — root domain, lowercase, no www. and no protocol.
• Pass both when a name is all you were given. Resolve the domain first if you can, then send both.
• Use it as a filter, not an enrichment. “Flag the companies running ads” is a fast qualification pass across a list — and non-advertisers cost nothing, since billing is per ad scraped.
• Number the steps when it runs over a list. Step 1 normalise the domains, step 2 b2b_ads_search on each, step 3 keep only the accounts with live ads. In one sentence the model guesses the boundaries and enriches everything it touched.
• Ask for the raw ad fields. “Return a table: company, ad headline, body copy, format, first seen, last seen.” Left to itself your client writes “they advertise about AI” and you lose the copy you paid to scrape.
• Say what an empty result should look like. “If no ads are found, mark the company none and move on — do not substitute homepage or blog copy.” Padding an empty response is the common failure here.
• Read the creative, do not quote it back verbatim. A prospect noticing you copied their own ad headline reads as automated.
• Check Meta separately. This tool covers B2B ads; Facebook and Instagram page ads come from meta_ads_analysis.

Output

{
  "company_domain": "acme.com",
  "ads": [
    {
      "headline": "Close the books in two days, not two weeks.",
      "targeting": "Finance leaders, US",
      "landing_page": "https://acme.com/close"
    }
  ]
}

Credits and limits

0.3 credits per ad scraped, not per call. A domain running ten ads costs 3 credits; a domain running none costs nothing. Spend therefore scales with how heavily the account advertises, not with how many domains you check. Run check_credits (free) before a large batch.

Parameters

Related tools

Next steps

• All MCP tools — the full catalogue with credit costs
• Research & Web Scraping — workflows that chain this tool into account prioritisation

Keep learning

Two courses take these tools past the reference page — clustered by GTM job, then chained into workflows.