Estimate Competitor Facebook Ad Spend (2027)

What the free Ad Library shows about spend, and what it hides

Start with the disappointing truth, because it saves you from chasing numbers that do not exist. For a normal commercial advertiser, the Meta Ad Library publishes no spend at all. You will not find a dollar figure, a daily budget, or a lifetime amount anywhere on a competitor's shampoo, sneaker, or software ad. Meta treats commercial spend as private, and no free tool can surface a number the platform never puts online in the first place.

What the Ad Library does give you, for every active commercial ad, is a set of signals. You see the full creative, the primary text and headline, the call-to-action button, the Page running the ad, the platforms it appears on (Facebook, Instagram, Messenger, Audience Network), and the date it started running. You also see how many ads a Page has live right now. None of that is spend. All of it is evidence you can reason from.

The disclosures widen in two situations. Ads about social issues, elections, or politics carry extra transparency: banded spend, banded impressions, and the funding entity behind them, and Meta keeps them searchable for seven years whether they are active or not. Ads delivered to people in the European Union carry their own layer under the Digital Services Act: the total number of people reached, broken down by country, age, and gender, plus the targeting parameters the advertiser chose. Those EU ads are archived for a year after their last impression rather than disappearing the moment they stop.

Read that table the right way and the method for the rest of this guide falls out of it. Commercial spend is never a readout. It is an inference you build from run time, ad volume, and the one genuinely quantitative field Meta is forced to publish, which is EU reach, then cross-check against a neutral cost benchmark. Everything below turns those proxies into a number you can defend, and shows how wide the error bars have to be.

Signal one: the active-ad count

The first number worth writing down is how many ads a Page has active. It sits at the top of the results when you search a Page in the Ad Library, and it is a rough read on how hard a competitor is testing. A store cycling forty live creatives is running a different operation from one nursing three, and heavier testing usually travels with a larger budget. It is the cheapest signal to collect and the easiest to over-read, so handle it carefully.

Two things inflate the count. First, one creative concept often appears as many near-identical ad entries, because the same ad split across placements, audiences, or language variants can each register separately. A Page showing sixty ads might be running twelve real ideas. Second, a large library can be padded with old evergreen ads that carry a trickle of budget rather than a flood. Volume tells you about testing intensity, not about dollars, and treating each listed ad as an equal slice of spend is the fastest way to a wrong answer.

Read the count as a trend and as a ratio, not as a total. Check the same competitor every week and watch the direction: a Page climbing from ten to fifty active ads is scaling and almost certainly spending more, while one drifting down is pulling back. Then look at how many distinct concepts sit underneath the count, because ten genuinely different ideas signal a more serious creative program than fifty copies of one. The count sets the scale of the operation. The next signal tells you which of those ads is actually working, which is where the money concentrates.

Two filters sharpen the read. Filter by platform to separate the Facebook and Instagram pushes, because a Page can be heavy on one and quiet on the other, and a count that looks large collapses once you split it. And distinguish a burst from a baseline: a competitor that jumps from eight to forty ads the week before a seasonal sale is loading a short campaign, not lifting its steady-state budget, so annualising that spike would badly overstate the year. The count is most trustworthy when you have watched it for a month and know the difference between the noise of a launch and the level a brand actually sustains.

Signal two: run time, the strongest free proxy

Run time is the most reliable free signal you have, and most people ignore it. The Ad Library stamps every ad with the date it started running. It does not show an end date for a live ad, so you cannot read exact length off a single visit, but the start date alone is powerful, and you can measure true length by logging the library on a schedule.

The logic is simple and hard to game. Nobody keeps paying to serve an ad that loses money. An ad that has run continuously for sixty or ninety days is, with very high confidence, profitable, because a competitor would have killed it otherwise. Longevity is the market voting with a budget. So the ads that have been live longest are both the winners worth studying and the ads soaking up the most sustained spend. A brand-new ad tells you a test just launched. An ad running since three months ago tells you where the money has settled.

Turn this into a routine. Once a week, screenshot or log a competitor's active ads with their start dates, and note which ones are still there. Over a month you build a picture no single snapshot can give: which concepts got cut inside a week (failed tests), which survived (winners), and roughly how the budget is distributed across the survivors. Ads about to pause are worth capturing before they vanish, because a commercial ad drops out of the library once it stops (outside the EU, where it lingers for a year). Pair the survivors with the active-ad count and you can say, with evidence, that a competitor is running heavy testing and concentrating spend on a handful of long-lived winners. That is a far stronger claim than any single guessed dollar figure.

Signal three: EU reach, the only published quantity you can convert

Here is the one place the Ad Library hands you real numbers on a commercial ad. Under the EU Digital Services Act, Article 39 requires very large platforms to publish, for every ad delivered in the European Union, the total number of recipients reached, broken down by member state, along with the main targeting parameters used and the beneficiary and payer. Article 39 does not require spend, so Meta still does not show it. But reach is a quantity, and a quantity you can work with.

To get from reach to spend you need two more pieces, and both are assumptions rather than facts, which is exactly why the output is a range. First, impressions. Reach counts people; impressions count views, and one person usually sees an ad several times. Impressions equal reach multiplied by frequency, and frequency is the number you cannot see, so you assume a plausible band (often somewhere between 1.3 and 3.0 over an ad's run, higher the longer it runs). Second, price. Spend equals impressions divided by a thousand, multiplied by CPM. For CPM you use a neutral, dated benchmark rather than a guess: Gupta Media's tracker put the blended Meta CPM near $8.19 across 2025, with sharp seasonal spikes, which gives you a defensible anchor.

Two cautions keep this honest. The reach shown is EU-only, so it captures nothing a competitor spends in the United States, the United Kingdom, or anywhere else, and for many brands the EU is a slice of total spend. And European CPMs vary widely by country, so a reach figure concentrated in a high-cost market like Germany implies more spend than the same reach spread across lower-cost markets. Use the EU reach as your one hard input, wrap it in an explicit frequency assumption and a sourced CPM, and you get a spend range for that ad's EU delivery that you can actually explain to a skeptical colleague.

Signal four: spend bands, and why they rarely fit a commercial rival

The Ad Library does display spend in plain view for one category: ads about social issues, elections, and politics. On those, Meta shows spend as a band rather than a precise figure (ranges such as less than $100, or $500 to $999), alongside a banded impression range and the funding entity. Meta keeps these ads for seven years, active or not, which makes political advertising unusually easy to size. If your competitor is an advocacy group, a campaign, or an issue-driven organisation, the spend bands are a genuine gift and you should read them directly.

For the commercial rival most marketers care about, though, the bands almost never apply, and the ground under them has shifted. Meta stopped allowing political, electoral, and social issue ads across its EU platforms in October 2025, in response to the EU's Transparency and Targeting of Political Advertising regulation, which the company said created operational and legal uncertainty it could not resolve. So the EU route to banded political spend, which some researchers leaned on, is closed for ads delivered there. Outside the EU, political spend bands still exist, but a shampoo brand or a fitness app has never qualified for them and never will.

The practical takeaway is to know which lever you are pulling. If you are sizing a political or issue advertiser outside the EU, read the spend bands straight off the ad. If you are sizing a normal commercial competitor, the bands are a dead end, and you are back to the three signals above: active-ad count for scale, run time for where the money concentrates, and EU reach for the only quantity you can convert. Do not go looking for a commercial spend band that the platform was never built to show.

A worked estimate you can defend

Put the pieces together with a concrete example. Say you are watching a direct-to-consumer apparel competitor. Its Page shows twenty-five active ads. Weekly logging tells you five of them have run continuously for over two months, so those five are the winners carrying most of the budget. One of those five is delivered in the EU, and its Ad Library entry reports around 800,000 people reached in the EU over its run. That reach is your one hard input.

Now convert it. Impressions equal reach times frequency, and spend equals impressions divided by a thousand times CPM. You do not know the frequency, so you run a low, base, and high scenario, and you anchor CPM to Gupta Media's 2025 blended figure of $8.19, using its Black Friday reading of $16.85 for the peak-season high case.

So this single EU ad likely cost somewhere between roughly $8,500 and $13,000 over its run in normal conditions, and could approach $40,000 if it ran through a Q4 peak at high frequency. Notice the spread: a factor of nearly five, driven entirely by two assumptions you cannot observe. That spread is the honest answer, and pretending it is a single number is where competitor estimates go wrong.

Scaling to a portfolio only widens the band. You could multiply a per-ad figure across the winners, but reach overlaps across a brand's ads (the same person sees several), so you cannot simply add reach figures without double-counting. And every dollar spent outside the EU is invisible, so a brand that is 20% EU could be spending five times your EU-based estimate globally. The defensible statement is a directional range with its assumptions stated out loud: this competitor is running heavy testing, concentrating budget on about five long-lived winners, and spending on the order of five figures a month in the EU alone. That sentence is worth more than a fake-precise total, because you can defend every part of it.

Where the estimate breaks, and the tools that hide it

It is worth being blunt about the failure points, because they are also the reasons to distrust any tool that sells you a tidy number. The reach-to-spend chain rests on two unknowns, frequency and CPM, and both move a lot. CPM is not a constant: it swings by season (Black Friday 2024 ran near $16.85, roughly double the $8.19 year average), by country, by audience, and by objective, and it drifts upward structurally, with Meta reporting the average price per ad rose about 9% across full-year 2025. Apply last year's CPM to this year's delivery and you are already off.

The visibility gaps are worse than the math. You see EU reach only, so most global spend is dark. You see active ads only for commercial advertisers, so paused winners and the budget behind them vanish from view. Duplicate ads inflate counts. And the wider measurement environment has been noisier since Apple's App Tracking Transparency, which a University of Maryland study estimated cut ad click-throughs by about 37%, thinning the very signal that made older cost assumptions reliable. Every one of these pushes your estimate toward a range and away from a point.

Which is exactly why paid competitor spend estimators deserve skepticism. Meta releases no commercial spend to anyone, so a vendor's number is built from the same public signals in this guide, then presented with a confidence the underlying data does not support. There is no private feed that makes a purchased figure more real than one you build yourself. Building it yourself has a real advantage: you know the assumptions, so you know how much to trust the answer. A number handed to you hides its own error bars, and an estimate whose error bars you cannot see is worse than no estimate at all.

The faster answer: run a counter-creative and measure the auction yourself

Step back and ask what the spend estimate is really for. Usually it is a stand-in for the question you actually care about: can I compete in this auction, and what will it cost me. You can answer that question directly, with real numbers, faster than you can triangulate a rival's budget, and the answer is about your market, your audience, and your offer rather than a guess about theirs.

Take the angle you watched survive in the Ad Library (the hook that kept running, the format that stuck) and ship your own counter-creative against it. Give it a clean test budget large enough to gather results, let it run past the learning phase, and read three of your own numbers: CPM (what reach costs you), CPC, and cost per result, which is what each purchase or lead actually costs. Those are ground truth. No competitor estimate can match a figure you measured in the same auction you are trying to win. Sanity-check them against neutral benchmarks so you know whether you are in a normal range: WordStream put the all-industry Leads CPC near $1.92 and the Traffic CPC near $0.70 in 2025, and Gupta Media put the blended Meta CPM around $8.19, on a platform whose ads could reach 2.28 billion people as of early 2025.

That measurement loop is also where competitive research pays off, because reading the winners is only useful if you turn it into a test. Research the angle already working, generate the on-brand counter-creative, launch it to Meta, then read your own CPM and cost per result and feed that into the next round. A platform like AdPlay.ai keeps that loop in one place, but the discipline holds with any workflow. Estimating a competitor's spend is a useful sketch of the battlefield. Running your own ad and reading the meter is how you find out what it actually costs to win, which is the number that was worth chasing all along.