Spying on a competitor's ads: what you can actually see

Let us start with what most articles on this subject hide from you: nobody can see a competitor's advertising budget. Not you, not the paid tools, not the agencies that promise it. That number exists nowhere outside the advertiser's own account.

What is visible, on the other hand, is far more than people assume, and entirely legally. Since transparency obligations were imposed on the large platforms, active ads are publicly viewable, with their visuals, their copy and their run dates. That is already a great deal, provided you know what to draw from it.

Here is the exact border between the two, and what you can deduce while staying on the right side of it.

What you can see, for free

The creative currently running

The ad libraries of the major platforms list the active ads of any advertiser. You see the image or video, the hook, the body copy, the button and the landing page. Most of them require no account at all.

This is the most useful raw material, because it shows you the argument the competitor chose to lead with, rather than the one displayed on their website.

The first run date

Every ad carries the date it went live. It is the single most underrated field in the whole library, and I come back to it in detail below, because on its own it replaces half the numbers you wish you had.

The number of variants

One campaign often appears several times over, with different visuals or hooks. That count tells you how much the advertiser is testing, and therefore how seriously they work the channel.

The pixels installed on the site

This one does not come from the libraries but from the site itself. Tracking tags are loaded by the browser, so you can see which ad networks are wired up, even when no ad is running right now. A pixel that has sat on a store for a long while, on a store advertising nothing today, tells a story all by itself.

What you will never see

Four things stay out of reach, and no serious tool will claim otherwise.

The budget. Not the amount spent, not the cost per click, not the acquisition cost. Those figures live in the ad account and do not leave it.

The precise targeting. You will not know which audiences are aimed at, nor which exclusions were set. Some platforms show age brackets or countries for political ads, never for commerce.

The performance. No click-through rate, no return on spend. A highly visible ad can be an expensive failure; a discreet one can be the most profitable in the account.

What has been switched off. Libraries mostly show what is live. The full history of past campaigns, and the reasons they stopped, is not public.

Hold on to that limit, because it defines everything else: you are observing decisions, not results.

The most reliable signal: how long it has run

Why time replaces budget

Nobody keeps paying for an ad that loses money. It is the one economic rule that holds in this field, and it turns a date into first-rate information.

An ad that has run for three months without interruption is almost certainly profitable. An ad launched four days ago tells you nothing: it is a test, and most tests fail.

How to read the dates

Sort a competitor's ads by age rather than by the order they appear in. The ones that have survived past sixty days form their foundation: these are the arguments validated by real money. The recent ones form their test queue, useful for seeing where they are looking, not for imitating.

If a single creative has been running for a very long time and nothing else moves, you are dealing with an advertiser who found their formula and stopped exploring. That is a comfortable position and a vulnerable one at the same time.

Reading the creative without copying it

The hook names the problem

The opening words of an ad reveal which problem the competitor believes is most painful for their customer. Two stores selling the same product with two different hooks are not addressing the same people, and that gap is often where your place is.

The format names the funnel stage

A long demonstration video does not serve the same moment as a product carousel or a single image with a discount code. Spotting which formats coexist shows you whether the advertiser works the whole journey or only its end.

The landing page names the strategy

Follow the link. An ad that leads to the product page is chasing the immediate sale. An ad that leads to an article or a quiz is building an audience before selling. The second approach costs more upfront and holds up better over time.

What the pixels add

A pixel is a tag the advertiser installs to measure and to retarget. Its presence is a declaration of intent far more reliable than any single ad.

Several networks wired at once signal diversification, so a budget and a team. A single network signals dependence, and an opening for you elsewhere. No pixel at all on a store that has traffic means it lives on something else: search, organic social, or an existing audience. That last case is the most instructive, because it proves a market can be taken without an ad budget.

And an old pixel on a store running nothing today points to a campaign that was stopped. Often because it did not pay.

The method, in one hour

Take ten competitors, not one. The point is not to dissect a store but to see what recurs across all of them.

For each, note three things: how many ads are active, how long the oldest one has been running, and which networks are wired on the site. Then add the hook of the oldest ad, word for word. In an hour you get a table that shows, with no dubious estimates at all, who is investing, for how long, and with what argument.

That table beats a paid report handing you an invented budget.

What you do with it next

The temptation is to copy the best ad. It is the worst possible use of this work, for one simple reason: you would be copying the visible output of a system you cannot see. The same creative, served to a different audience with a different page and a different price, will not produce the same effect.

Look for the absences instead. Which argument is nobody using? Which format is nobody trying? Which objection is everybody avoiding? In a market where ten competitors have said the same thing for six months, the profitable position is rarely the eleventh repetition.

And always weigh what you see against the rest: a very visible advertiser on a store with a two-year-old catalogue and no fresh reviews is not necessarily healthy. Advertising is one signal among several, to be crossed with traffic and catalogue, as we set out in our article on analysing a competitor's store.

The mistakes that distort everything

Confusing visibility with budget. An advertiser can flood your feed because you are exactly their target, on a modest but well-aimed budget.

Taking a recent ad as proof. Until it has cleared several weeks, it proves nothing at all.

Ignoring the country of delivery. The same brand often runs different messages per market. Check that you are looking at yours.

Believing tools that display a budget. That figure is always a derived estimate, never data. A tool that states it without an error margin is selling you confidence, not information.

Watching a single competitor. One advertiser can be wrong for months. Ten advertisers doing the same thing for six months is another matter.

Frequently asked questions

Is it legal to look at competitors' ads?

Yes, unambiguously. These libraries were created for transparency and are open to everyone. You are consulting what the advertiser broadcasts publicly, much like a poster in the street.

Does the competitor know I am looking?

No. Browsing an ad library is anonymous and nobody is notified.

Can you see ads that were stopped?

Rarely, and incompletely. Which is why it pays to record what is running on a regular basis: your own history will end up richer than what the platform keeps.

Do you need a paid tool for this?

Not for the creative, which is public. A tool earns its place for what requires cross-referencing: detecting installed pixels, tracking change over time, and comparing ten stores in the same table.

Where to start this week

Three actions, in this order, and the first takes ten minutes.

Open the ad library of your main network, look up your three closest competitors and note only the date of each one's oldest active ad. You will know immediately which of them has found something that works. Then widen to ten competitors and add the number of active variants, which separates those who test from those who coast. Finish by recording which networks are wired on each store, to see who depends on a single channel.

By hand this survey eats most of a morning and goes stale in a fortnight. Evidio does it from the browser, on any store: ad pixels and active networks, estimated traffic and its sources, installed apps, featured products and technical make-up, with an export so your ten competitors sit in the same table. The extension is free, and our plans exist only to track that change over time.

To go further into what those stores actually sell, our method for finding a winning product takes over.