You search a competitor’s brand terms and see nothing unusual. Their landing page looks unremarkable. Meanwhile they are running an aggressive campaign in three markets you compete in, with offers you have never seen, and your competitive picture is quietly two quarters out of date.
Two different mechanisms produce that blind spot, and they are worth separating because only one of them is adversarial. Geo-targeting means the ad was never meant for you, since you are not in the target market. Cloaking means someone is deliberately showing different content to different viewers. Both make competitor activity invisible from a single vantage point, and both are addressable the same way.
Geo-targeting: not hidden, just not yours
Most advertising is targeted, and location is the primary axis. A campaign running in Germany, Brazil and Japan is simply not served to a request that appears to come from your office.
That is ordinary and not deceptive, but the consequence for competitive intelligence is the same as if it were: you cannot see it. If your competitor tests aggressive pricing in one market before rolling it out, or runs a campaign against your brand in a region where you are weaker, none of that reaches you. Worse, its absence reads as inactivity, which is a comfortable and wrong conclusion.
Beyond geography, delivery also varies by device, language, time of day, and often by inferred audience, so a desktop check in one country is a narrow slice of what is actually being served.
Cloaking: the adversarial version
Cloaking is the deliberate practice of serving different content depending on who is asking, specifically to show one thing to some viewers and something else to everyone else.
The classic use is evading review: a compliant page for the platform’s automated checks and human reviewers, and the real, often non-compliant offer for ordinary users arriving from the ad. The detection logic keys on whatever distinguishes a reviewer from a user, which usually means the IP address and its network, since traffic from known hosting ranges and platform-owned infrastructure is easy to identify, plus signals like user agent, referrer, and how the session behaves.
There is a competitive version too, where advertisers filter out traffic that looks like it belongs to a rival, so competitors browsing from their corporate network or from cloud infrastructure are shown something bland while real prospects see the actual campaign.
The relevant point for you is that both variants filter on the same thing: does this look like a real user from a real place. Which is precisely the question a residential vantage point answers.
Why your usual checks miss it
Three habits produce a false negative.
Checking from one location. Everything geo-targeted elsewhere is invisible, and you cannot distinguish “not running” from “not running here”.
Checking from an office or cloud IP. Corporate ranges are identifiable, and datacenter ranges more so. Cloaking specifically filters those, so this is the case where you are most likely to be shown the sanitised version and conclude nothing is happening.
Checking as a logged-in, cookie-laden browser. Your own history and account state affect what serves, so you are measuring your personalised view rather than a prospect’s.
Seeing what a real user sees
The method follows directly from the mechanism.
Request from inside the market. A residential exit in the target country makes the check look like an ordinary local visitor, which is the same requirement that underpins residential proxies for ad verification. Where the campaign is city-targeted, so is the check, per city-level targeting.
Match every other signal to that persona. Device, language, timezone and locale should agree with the market you are claiming to be in, because a contradiction is both a detection signal and a reason to be served something different, per matching geo, timezone and locale.
Arrive clean. Logged out, with a fresh cookie jar, so you see the acquisition experience rather than a personalised one.
Follow the whole path. Cloaking often lives after the click, so the ad and the first page may look fine while the destination differs. Capture the full redirect chain and the final page, not just the impression.
Vary and repeat. A single check is a data point. Cloaking is often probabilistic, rotating, or time-boxed, so sample repeatedly across different exits and hours before concluding anything, and treat one clean result as weak evidence.
What to capture, and what to do with it
For each market, record the creative and copy, the offer and pricing shown, the landing page and full redirect path, and a screenshot with a timestamp and the market used. The screenshot matters because the interesting finding is usually a difference between markets, and a difference is only demonstrable if you captured both sides.
Comparison is where the value is. The same competitor showing different pricing in two markets is a strategy signal. A campaign appearing in one region and not others is a test in progress. And an offer that differs between what a reviewer would see and what a user sees is a compliance matter you may want to report to the platform rather than merely note.
Run it as a repeating sweep rather than an occasional exercise, since the point is to notice change. The broader competitive programme this fits into is in residential proxies for competitor ad intelligence.
Staying on the right side of it
Worth being clear, because this area attracts sloppy advice.
Observing publicly served advertising as an ordinary user does is legitimate market research. Clicking competitor ads to drain their budget is click fraud, not intelligence, and it is both illegal in most places and trivially detectable. Do not click through on paid placements as part of a monitoring sweep, and if you need to see the landing page, reach it directly rather than through the ad.
Beyond that, the usual discipline applies: keep to public content, pace politely, respect terms of service, and do not attempt to access anything gated. The wider framing is in are residential proxies legal.
The bottom line
Competitor advertising you cannot see splits into two cases: geo-targeted campaigns that were never aimed at you, and cloaked campaigns deliberately hidden from viewers who look like reviewers or rivals. Both filter on the same signal, whether the request looks like a real user in a real place, so both are addressed by checking from residential vantage points inside each market, with device, locale and timezone matched, arriving logged out, following the full redirect chain, and sampling repeatedly rather than once. Capture screenshots and offers per market, because the finding is almost always a difference between markets rather than any single page. And keep it observational: monitoring what is served is research, clicking competitor ads is fraud.
The vantage points are the part you cannot improvise, and that is what residential proxies for ad verification provide, running on residential proxies with country and city targeting and per-GB pricing that suits a recurring multi-market sweep.