Competitor ad monitoring alerts: the routine, not the hack

There is no alert. Neither the Meta Ad Library nor the Google Ads Transparency Center will email you when a competitor launches a campaign, changes an offer or kills a creative — no digest, no webhook, no notification setting. Every guide on this subject concedes that, then either points a page-change watcher at a JavaScript app that defeats it, or pitches a tool. Nobody publishes the thing you need: an operational routine, with a cadence and a definition of signal versus noise. That's this page — grounded in 50,982 ads we've AI-qualified across 158 UK brands (Google 30,466, Meta 15,365, TikTok 5,151; read 4 November 2026), plus the undocumented Google cap that explains why eyeballing a rival's page silently misses launches.

Why no native alerts exist

Both libraries exist to satisfy political-advertising transparency law, not to serve marketers. They're built to answer "who paid for this specific ad?" for a journalist looking at one ad, once — not "what changed across my six rivals since Friday?", which is the question every marketer actually arrives with. So the capability is missing by design, on both:

• Meta Ad Library — no email digest, no Slack integration, no notification option anywhere. There is an Ad Library API, and for UK and EU audiences it does return commercial ads rather than just political ones, thanks to the Digital Services Act. But it needs government-ID identity verification, and it is a query endpoint: you ask it a question, it answers. Nothing in it watches for change or tells you when something new appears, so the alerting is still yours to build.
• Google Ads Transparency Center — no alerts, no API of any kind, no export. You get a search box and a grid of creatives.
• TikTok Commercial Content Library — a CSV export for EU-targeted ads, but again no subscription or alerting layer.

Why the "point Visualping at it" trick disappoints

The most-repeated free workaround is to aim a visual page-change monitor at a competitor's Ad Library URL. It's a reasonable instinct and it mostly doesn't work, for three reasons worth knowing before you spend an afternoon on it:

• The results are rendered by JavaScript after load. Change monitors generally compare fetched HTML or a screenshot of initial render. The ad grid often isn't in either.
• The grid is not stably ordered. A pixel-diff monitor fires on a reshuffle and stays silent on a genuinely new creative buried further down — alerts that are noisy and incomplete at once, which is how you stop trusting them and then stop reading them.
• It watches one platform, one brand, one page. Six competitors across three platforms is eighteen monitors to maintain, and the thing you most want — "which rivals started discounting this week?" — is a question no single page answers.

For one small advertiser on one platform, a page monitor beats nothing. The failure modes above scale with the advertiser's size and the number of things you watch — they bite hardest where the stakes are highest.

The real reason a routine beats eyeballing: the 900-creative cap

We only found this by crawling at scale. For large advertisers, the Google Ads Transparency Center reports a large total then quietly refuses to show most of it: pagination stops at roughly 900 creatives. No error, no warning — the grid just ends, and it's natural to conclude you've seen everything.

We first measured it on Pets at Home: the Center reported about 3,000 ads and stopped paging at around 900. Across the UK brands whose Google ads we track, 48 now hold more Google creatives in our database — accumulated over repeated captures — than the Center will page through in a single session.

The implication for alerting is the point. If a competitor is big enough to hit that ceiling, manual checking cannot detect their new launches reliably — a new creative may never be reachable in the pages you can load. You aren't being careless; the tool is capping you. That's the structural argument for a recorded routine: it compares what you captured last time against what you can capture now, so "new since Tuesday" is a diff rather than a feat of memory. Our guide on why the Transparency Center isn't showing all ads covers the cap and five other causes; the Meta equivalent does the same for the Ad Library.

The cadence table

The principle: match checking frequency to how fast the thing you're watching can change, and never check faster than the library updates. Both lag — practitioner measurements put Google's at roughly one to three days between an ad serving and appearing. Daily checks for new creative mostly re-read yesterday's cache.

What is actually a signal

The hard part isn't collection, it's refusing to react to most of what you collect. Base rates separate signal from noise — if something happens constantly across 50,982 ads, a rival doing it isn't news.

Which platforms to watch for a given competitor

Generic advice says "check all the libraries". That's wasted work, because most brands don't run everywhere. Across our 158 brands with qualified ads, 63 appear on one platform, 52 on two and 43 on all three of Google, Meta and TikTok — so roughly 60% need multi-platform watching and 40% don't. The three-platform brands are where single-platform monitoring misleads most: you can watch their Meta activity all quarter and miss where the volume is.

The shapes differ wildly. Pets at Home puts 41% of its creative into TikTok, so a Meta-only watch sees under a tenth of their output; Cancer Research UK inverts that, Meta-led with Google a distant second; ASOS is overwhelmingly Google. There is no default platform mix, which is why the first job in any monitoring setup is a one-off audit of where each rival actually runs — our guide on which platforms competitors advertise on covers doing it by hand.

Setting the routine up: a worked weekly check

Thirty minutes, once a week, for up to six competitors. The discipline that makes it work is recording, not looking — you cannot diff against a memory.

1. Fix the slot and the sheet. One row per competitor per week: date checked, platform, creative count seen, headline offer, offer mechanic, promo code, stated deadline, notes. "None stated" is a value — leave nothing blank, or next week you won't know whether you saw nothing or forgot to look.
2. Check only the platforms that rival uses, per your one-off audit. Don't re-audit weekly.
3. Record the count before reading anything. It's your crude new-creative detector — note it even when it looks unchanged.
4. Read the newest creatives only. Sort by most recent where the library allows, looking for a changed offer, a new product, a new audience or a deadline.
5. Diff against last week and write one line — what changed, or "no material change". That habit is what turns a spreadsheet into something you'll actually read back in three months.
6. Escalate only approval-level changes — new mechanic, dated offer, new platform. The rest waits for the monthly review.

Two caveats. This routine detects presence, not performance: the libraries show what's running, never what's working, so a creative you see fifty times may be a winner or a badly-managed budget — inferring which ads work is a separate problem. And for rivals large enough to hit the 900-creative cap, your manual count plateaus and stops detecting new creative, so lean on offer changes rather than counts.

When to stop doing this by hand

The routine above is sufficient for two or three competitors on one or two platforms. It stops being sufficient predictably: when competitors × platforms passes roughly ten combinations, when rivals are large enough that the Google cap hides their launches, or when you need history — "when did they start discounting last year?" — which a manual sheet answers only for the weeks you filled in.

That's the gap we built AdRival for. We track live ads across Google, Meta, TikTok and LinkedIn, qualify every creative with an AI pass extracting the offer, discount, hook, audience, urgency and fourteen other fields, and capture on a schedule so the diff — what's new, changed or stopped — is computed rather than remembered. Accumulating across repeated crawls is also how we get past the 900-creative ceiling.

To be straight: alerting is a service we run and deliver, not software you log into — no self-serve signup, no API. We agree the competitor set and cadence with you, then send the reports and flags. A one-off brand report is £99 and ongoing tracking of a single competitor is £99 a month; broader category coverage costs more. Current tiers are on our pricing page, and enquiries go through the form there or on our landing page.

If you never buy anything from us, the two tables above are the useful parts — the operational layer the rest of the internet leaves out, and they work with the free libraries exactly as they are.

Related reading: tracking competitor promotions and discounts for the offer-monitoring detail, and ad spy tools compared for an honest look at the paid alternatives.