You paid for a geo. Check that the geo got it.
A media buy is a chain of promises: the network says the placement ran, the tracker says the click landed, the offer says the flow works. Each link is checkable only from inside the country you paid for.
Why proxies solve this
Placements, creative moderation and landing pages are all resolved from the visitor’s country. From your own connection every campaign looks like it is running. The countries where the creative was rejected, the placement never filled or the lander redirects to nothing are exactly the ones you cannot see without an address there.
The problems you hit without proxies
The report says delivered, the placement says otherwise
Networks report impressions, not what the page looked like. A creative can be moderated down, replaced by a fallback or served next to something you would not have bought.
Moderation is per country
The same creative passes in one market and is rejected in another, often silently. Spend keeps flowing into a geo where nothing renders.
The flow breaks between the click and the offer
Pre-lander, lander, offer. Any hop can carry a geo rule, and the break is usually in a hop rather than on the final page.
How SotaProxy handles it
Residential addresses in 220+ countries with sticky sessions that hold a whole flow, and mobile addresses in 26 where the traffic source is mobile and the placement behaves differently in an in-app browser.
Setup in 4 steps
Check the geos you buy, not the geos on the offer card
Sources deliver what they have, not what was ordered. Take the country breakdown from your own tracker and check that list.
Look at the placement, not only the click
Open the publisher page from the country in question and see the unit as a reader sees it. That is the only way to catch a fallback creative or a placement that never fills.
Walk the flow on one sticky session
Pre-lander, lander, offer, from one address, so the path you tested is a path a visitor could have taken.
Check mobile from a mobile address
In-app browsers and carrier networks render and redirect differently. Desktop checks of a mobile buy prove very little.
Fifteen geos, twice a day
Placement checks are small in volume and worth more than they cost:
The check costs less than an hour of the budget it protects. The reason it gets skipped is not price, it is that nothing on the dashboard says the geo is broken.
One geo, one flow, one address
Open the placement, then walk the flow, on a session that holds:
Checking a placement and the flow behind it
# one buyer in one country for the whole check
S="login_c_ID_s_44_ttl_50m:password@proxy.sotaproxy.com:10000"
# the publisher page where the unit should render
curl -s -x $S -H "Accept-Language: id-ID" "https://publisher.example/article" -o placement.html
# the flow behind the click
curl -sL -x $S "https://tracker.example/click?cid=77" -o prelander.html
curl -sL -x $S "https://offer.example/lander" -o lander.html
# and the same from a mobile address if the buy is mobile
curl -sL -x login_c_ID_mobile_s_45_ttl_50m:password@proxy.sotaproxy.com:10000 \
"https://offer.example/lander" -o lander-mobile.html- Save the placement page, not just a screenshot of the unit: fallbacks are visible in the markup before they are visible to the eye.
- Compare lander weight between geos. A page a tenth of the usual size is a redirect stub, not a lander.
- Record the country and the hour. Moderation decisions change during the day and so do fill rates.
Three checks that prove nothing
Checking from your own country
Everything renders, because your country is not the one with the problem. This is the single most common reason a broken geo runs for a week.
Testing the lander without the hops
The final page is fine and the redirect before it is not. A check that starts at the lander skips the part that actually broke.
Desktop checks for mobile buys
In-app browsers behave differently enough that a clean desktop test says nothing about the traffic you are actually buying.
Frequently asked questions
How is this different from ad verification?
Ad verification is usually the brand checking that its ads appeared correctly. This is the buy side checking that the traffic it paid for reached something that works. Same tooling, opposite direction.
Do you help with cloaking?
No. Showing a moderator one page and the audience another is not something we support. Checking what your own audience sees is a different thing and it is what this page is about.
Do I need mobile addresses?
If you buy mobile traffic, yes. If the source is desktop, residential is enough and costs per gigabyte rather than per day.
How many countries can I watch?
Residential reaches 220+. Practically you check the ten or fifteen that carry your spend, daily, and the rest when something looks off in the numbers.
Ready to start?
Create an account, top up, and get proxy credentials in minutes. No sales call. No monthly minimum.
Create account