QA and localisation teams
Checking their own sites for the language, redirects and legal notices a market is served, none of which is visible from an office in a different market.
REGION TESTING
Geo testing proxies: see a page the way a local visitor does
Sites change what they serve by country: prices, stock, currency, language, consent banners, whole redirect chains. Testing that from your own desk measures your own market. A static address in the target country turns a guess into an observation you can repeat.
The catalogue covers 77 countries and names the provider behind every block. It lists what the catalogue carries and makes no claim about delivery tests in any of them.
In short: Geo testing here means testing by geography, not generative-engine optimisation: checking your site from inside a country, so that what you record is what a visitor there is served. Unless a visitor grants location permission, a site has to infer where they are, usually from the IP address. A static address in that country, kept between runs, is what makes a week-on-week comparison mean something.
The address you test from is on an internet provider’s network in the country picked at checkout, and nobody else uses it while the period runs. Prices, currencies, stock, delivery options and legal notices all change by country, and none of those changes is visible from an office in a different market.
How a site decides where you are
A browser shares a precise location through the Geolocation API, and the W3C specification requires express permission from the user before any location is shared. A site that never asks falls back on the address, looked up in a geolocation database. Networks can publish their own data for such databases in the geofeed format of RFC 8805. Language is a separate signal: the browser states its preference in the Accept-Language header (RFC 9110, section 12.5.4), so a test that moves the address but not the language setting can show local prices in the wrong language. The advice here follows from those mechanisms; we have measured no particular site.
WHO THIS IS FOR
Who runs these tests
Pricing and e-commerce teams
Checking their own sites for the prices, currencies, stock and delivery options that change by country.
Anyone comparing markets
Comparing what your own site offers in different markets, from a fixed address in each so that a difference between two runs points to a change.
STEP BY STEP
A repeatable regional check
List the pages to check and what should differ by market
The table of what varies by vantage point is a starting set.
Buy one address in each market
Keep it for as long as the comparison runs.
Start each run from a clean browser profile
With the language and time zone set to the market’s. Step-by-step guide: how to view your website from another country.
Record a baseline
A screenshot, the final URL after redirects, and the price and currency as shown.
Repeat on a fixed schedule from the same address
Compare each run with that baseline.
Confirm the address before trusting a difference
Load what is my IP through the address and confirm it still geolocates to the market. If it has moved, ask for a replacement and do not file the result.
WHAT VARIES BY VANTAGE POINT
What a site can change about itself once it knows where you are
| What changes | How you notice it is wrong | |
|---|---|---|
| Price and currency | The number, the symbol, and whether tax is included | Your screenshot does not match the customer's invoice |
| Stock and availability | Which items and plans are offered at all | Support answers a question about a product nobody there can see |
| Language and locale | Copy, date order, decimal separators, address forms | A form rejects a valid local postcode |
| Consent and legal notices | Which banner appears and what it must ask | A market complains about a notice you never saw |
| Redirects and routing | Which domain or subfolder you land on | Analytics shows traffic entering somewhere you did not build |
| Search and recommendation | Ranking and suggestions tuned to the market | Two people describe the same page differently |
WHY A FIXED ADDRESS
Why one address per market
A repeatable measurement
A test is only useful if a difference between two runs comes from the site. Arriving from a different network each time adds a second cause. One address per market keeps the vantage point constant, so it drops out of the comparison.
An address outside hosting ranges
Some sites treat addresses registered to hosting companies differently from other connections. An address on an internet provider's network is the same kind of connection as a local visitor's on that check, so that variable matches what the visitor has.
State that survives the test
Carts, sessions, A/B assignments and cookie choices all live against an address. Rotation resets them mid-run, and a test that keeps resetting ends up testing the proxy.
WHAT YOU CONTROL
Properties that matter for this job
Country selection- Chosen at checkout and fixed for the period; 77 in the catalogue
-
Provider selection
- Named at checkout, so you choose it; 19 carriers in the United States
Rotation- None, so a repeat run starts where the last one finished
Protocols- HTTP, which carries HTTPS, or SOCKS5, switched on the order; scripted (headless) and ordinary browser runs leave from the same address, so they match
Access control- Username and password, or an IP whitelist for the server that runs your automated tests
Parallel checks- 100 sessions per proxy, 10 past 50 GB in a calendar month (Terms of Service, section 2.4)
RELATED
Other scenarios, and the product itself
FAQ
Questions about region testing
Can I choose a city or a region inside a country?
No. The static residential catalogue is sold by country, provider and plan, and no offer in it today is tied to a city or region. Each address carries whatever registration location its provider recorded for the block. You cannot choose it, and we do not promise one.
How many countries can I test from?
The catalogue carries 77, each with the provider named. Testing several markets means one address per market: they are priced per IP per period, so a five-market test is five addresses. No plan carries a country switch.
Do I need a browser, or is a script enough?
It depends what varies. Prices, stock and redirects are usually visible to a plain request. Consent banners, client-side geolocation and anything rendered after load need a real browser, and a browser will also ask for permissions and read a time zone that should agree with the country of the address. The address is one input among several; making the rest agree with it is your side of the job.
Why not use a VPN?
A consumer VPN puts you on an address shared with everyone else using that endpoint, and one that a site's geolocation or risk data may already mark as a VPN. That is fine for reading a news site and poor for measuring one, because you cannot tell whether what you are seeing is the local experience or the experience that kind of address gets. A dedicated address removes the ambiguity.
Will the address change between test runs?
Not during the period you paid for unless you ask for a replacement, and not when you renew, so the only thing that changed between runs is the thing you were testing.
What if a market's address stops resolving there?
An address that starts resolving to a different country is explicit grounds for a free replacement, up to 20 % of the IPs in the order per calendar month ([Terms of Service](/terms-of-service/), section 2.5). For this scenario it is the failure that matters most, because a test bench quietly moved to another country produces wrong answers that look confident. Terms of Service
When this is the right tool
Use it when the question is what your own site serves in one market and whether that has changed. Using an address to reach content a site withholds from your country is a different thing, and section 5 of the Acceptable Use Policy prohibits it. It is the wrong tool when the difference you care about is between cities, which the product cannot target, or when you need many independent samples at once, which is a rotating pool’s job. Start with one ISP proxy per market you sell into, priced per IP per period, and add markets as the test plan grows.Sources
- W3C Geolocation API: a browser shares a location only with the user’s express permission
- RFC 8805, the geofeed format: geolocation data networks publish for their own address blocks
- RFC 9110, section 12.5.4, Accept-Language: the browser’s language preference, a signal separate from the address