Muze AI
PlatformMuze MCPLiteResultsPricingTalk to SalesStart free trial

Ad research

Every ad your competitors are running is already public.

The Meta Ad Library is free, needs no login, and almost nobody uses it properly. Eight guides to reading it for your vertical, including the compliance rules that explain why the creative looks the way it does.

In short

The Meta Ad Library is a free public archive of every ad currently running on Facebook and Instagram, searchable by advertiser page name or by keyword across ad text, with no account required. It shows you the creative, the format and the date each ad started. It does not show spend, impressions, results or, outside the EU, targeting. So competitor ad research is the discipline of reading structure and duration, then testing what you find, rather than copying ads that you have no evidence are working.

Pick your vertical

Each guide covers what advertising looks like in that vertical, the regulatory constraints that shape the creative, the angles that recur, a step by step research method, and the mistakes that make competitor research actively harmful.

The method, in six steps

This works in every vertical. Open the Meta Ad Library, then:

  1. Build the list before you open anything. Five competitors, chosen for similar price point and business model rather than similar logo. Add one adjacent player, because they test angles your direct competitors have not.
  2. Search by page name, with the country set. Page names and brand names diverge. Country matters because the same advertiser runs different creative and different disclosures in each market, and the differences are informative.
  3. Read the run dates first. Longevity is the only free performance signal in the library. An ad still live after sixty days is usually working. It is an inference, not a measurement, and brands do leave losers running.
  4. Record structure, not pixels. Six fields per ad: hook, format, offer, proof device, call to action, destination. The structure transfers to your account. The art direction does not.
  5. Follow the click. Open the landing page. The gap between the promise in the ad and the first line on the page is where most wasted spend in any vertical lives, including yours.
  6. Turn one finding into one test. One new angle, three hooks, format and audience held constant, run against your current control. Copy five things at once and you learn nothing you can reuse.

Half an hour per competitor, and the output is a list of hypotheses rather than a swipe file. The vertical guides above add the part this generic version cannot: which angles are worth stealing in your category, and which claims you are not allowed to make.

What the library will not tell you

  • Spend and results. Outside political and social issue ads there is no budget, no impressions and no conversions. Nobody outside the account knows whether an ad worked.
  • Targeting. Ads shown in the EU carry extra detail under EU transparency rules. Everywhere else you get the creative and the dates.
  • History. Non-political ads leave the library when the advertiser turns them off, so what you did not record is gone.
  • Anything about the business behind the ad: margin, sales capacity, lifetime value, or whether the campaign made money. A competitor can run a beautiful ad straight into a loss.

None of that makes the library useless. It makes it a source of hypotheses. The read that holds up is structural: which angles a category keeps returning to, how fast a competitor iterates, and what they stopped running.

Where else to look

Meta is the biggest archive, not the only one. All three of these are free and public:

  • Google Ads Transparency Center. Search, Display, YouTube and Shopping creative by advertiser, with a verified advertiser identity and date ranges.
  • LinkedIn ad library. Public, reachable from any company page, and the only good window into B2B social creative.
  • TikTok Commercial Content Library. Ads and paid branded content, with more disclosure for ads shown in the EU.

Checking two libraries for the same competitor is usually more informative than checking one library harder, because channel mix is itself a strategic decision and you can see it for free.

Doing this without the browser tabs

The research is not hard. It is repetitive, and it decays: non-political ads disappear from the library the moment an advertiser turns them off, so anything you did not record is gone. That is the part worth automating.

Muze does the collection with four tools. list_competitors returns the competitors tracked on your account, scrape_competitor_ads starts a Meta Ad Library scrape and hands back a run id, get_scrape_status tells you when it has finished, and get_competitor_ads returns what it found. You call them from ChatGPT, Claude or any MCP client, over one OAuth connection that also covers Google Ads, Amazon Ads and Shopify.

The point is not API access. The point is that the research and the change to the account happen in the same conversation, so a finding becomes a test without a handoff. Reading scraped ads is a read call. Starting a scrape is classed as a write, so it needs a paid plan. The free tier includes 25 read-only tool calls a month. Every change to a live ad account previews first and waits for your confirmation, new campaigns are created paused, and Muze never takes a percentage of ad spend.

Frequently asked questions

What is an ad library?
A public archive published by an ad platform showing the ads its advertisers are currently running. The Meta Ad Library covers Facebook and Instagram, the Google Ads Transparency Center covers Search, Display, YouTube and Shopping, and LinkedIn and TikTok publish their own. All are free and none require an account.
Is the Meta Ad Library free?
Yes. It is public, it needs no login, and it covers every active ad on Facebook and Instagram. You can search by advertiser page name or by keyword across ad text, and filter by country and platform.
Can I see how much a competitor spends on ads?
No. Ad libraries publish spend ranges only for political and social issue advertising. For commercial ads there is no spend, no impression count and no result data. Any spend figure quoted by a third-party tool is a model estimate, not a disclosure.
How do I know which competitor ads are working?
You infer it from how long an ad has been running and whether the advertiser keeps returning to the same concept. Both are signals rather than evidence. Nobody outside the account can see delivery, and treating longevity as proof is the most common mistake in competitive research.
Is it legal to research competitor ads?
Reading a public ad library is reading published advertising. The care point is reproduction: their footage, photography and copy are their property, and in regulated verticals their claims sit on substantiation and authorisations you do not have. Take the structure and write your own.
How often should I do this?
Weekly in ecommerce and DTC where creative turns over fast, monthly in B2B SaaS and healthcare where it does not, and once per intake cycle in education. Non-political ads leave the library when they stop running, so anything you do not record is gone.

Keep reading

Ad examples, Meta Ads with Muze, creative testing, free advertising calculators, or how the Muze MCP connection works.

Research, then act, in one place

Connect Meta, Google, Amazon and Shopify once over OAuth, then ask for competitor ads, performance and changes in the same conversation. Free to start, and nothing changes in your account until you confirm it.

See how Muze works