Skip to content

How to see competitor ads in Claude

How-ToBlake Folgado
How to see competitor ads in Claude

Source review · Checked . Based on the linked documentation and tool definitions. Examples are illustrative unless a run receipt is provided. Editorial standards

To see a competitor's ads in Claude, connect ToolRouter, have Claude look up the brand's ad library page ID, then pull that page's active ads for your country. Ad Library Search reads the public libraries that Meta, Google, LinkedIn, TikTok and Reddit publish and returns ad copy, creative links, start dates and platforms as data Claude can compare.

Skip the ID step and you get noise. Keyword search returns any ad that mentions the word, from any advertiser.

The two-step workflow

  1. Find the advertiser. search_facebook_ad_companies returns matching pages with their page_id, Instagram handle and follower count. For Google, search_google_advertisers returns an advertiser_id, or you can pass the brand's domain.
  2. Pull their ads. get_facebook_company_ads with the page_id, a country, status: "ACTIVE" and trim: true. For Google, get_google_company_ads with domain or advertiser_id.

Then open the interesting ones: get_facebook_ad with get_transcript: true returns the script of a video ad under two minutes, and get_google_ad reads a single Google creative.

LibraryFind the advertiserGet their ads
Meta Ad Library (Facebook, Instagram, Messenger, Threads)search_facebook_ad_companiesget_facebook_company_ads
Google Ads Transparency Centersearch_google_advertisersget_google_company_ads
LinkedIn—search_linkedin_ads with company
TikTok—search_tiktok_ads with advertiser_name
Reddit—search_reddit_ads with query

Each call is billed at max($0.005, provider cost); the Meta lookups in this guide cost $0.005 each. Google's get_ad_details: true option costs more per ad, so leave it off until you need it. See Billing.

Connect once

In Claude, open Customize → Connectors → + → Add custom connector and use:

  • Name: ToolRouter
  • URL: https://api.toolrouter.com/mcp

For Claude Code:

bash
claude mcp add toolrouter -- npx -y toolrouter-mcp

A real run: Notion's ads in the UK

We ran this on 5 October 2026.

text
Use ToolRouter Ad Library Search. Find Notion's Facebook page ID, then list
its active ads in GB with start date, platforms, format, headline and call
to action. Tell me what they are selling and how.

search_facebook_ad_companies for "Notion" returned 15 pages, including a DJ, a magazine and several unrelated "Notion…" pages. The right one was identifiable by its Instagram handle (notionhq, about 492,000 followers) and page ID 825120160874613.

get_facebook_company_ads for that page in GB returned 7 active ads, all:

  • started on 19 August 2026, running on Facebook and Instagram;
  • dynamic creative (Meta tests the combinations), with the headline "Solve Work with Notion";
  • leading on Custom AI Agents that automate weekly reports and ticket triage;
  • split across two calls to action: 5 use Sign up and 2 use Get offer.

For contrast, a keyword search for "notion" in GB returned 1,292 active ads, and the first results came from a newsletter and an online course, not Notion. That is why the page ID comes first.

What you learn in two calls: Notion's UK paid message right now is AI agents rather than notes or docs, it has been running unchanged for about seven weeks, and it is testing a sign-up against an offer.

Connect your assistant, then try the task described in this guide.

Prompts that work

  • Teardown: "Find the Meta page IDs for these three competitors, pull their active ads in the US, and compare hooks, offers and calls to action in one table."
  • Longest runners: "Sort their ads by start date. The oldest ones are the ones they keep paying for. Get transcripts for any video ads."
  • Channel mix: "Check whether this brand runs ads on Google, LinkedIn and TikTok as well as Meta, and summarise where it spends attention."
  • Follow the click: "Scrape the landing page behind this ad and tell me the offer." That uses Web Scraper.

To see what the same brands post organically, read How to scrape TikTok and Instagram with Claude.

Limits

  • Libraries publish different things. Spend and reach appear for some ads and not others. Report them only when the response includes them.
  • Defaults matter. Meta results default to active ads; set status: "ALL" for history. Use trim: true and the cursor to page through long lists.
  • Empty is not absent. A domain with no Google results may advertise under another advertiser name; try search_google_advertisers.
  • Several pages share names. Check the Instagram handle or follower count before trusting a page ID.

Install the matching agent skill so your assistant knows the whole workflow:

bash
pnpm dlx skills add Humanleap/agent-skills --skill ad-library-scraper

Open Ad Library Search or connect at toolrouter.com/connect.

*Checked 5 October 2026 against the live tool schemas and calls to search_facebook_ad_companies, get_facebook_company_ads and search_facebook_ads for Notion in GB.*

B
Founder at ToolRouter
Share this article

Related Posts