How to check Google rankings by country and city (US, UK, Germany)

The short answer: four methods

There are four practical ways to see rankings from somewhere else. They differ in how precise they are, how far they scale and what they cost.

If you only need to know whether you rank in Germany at all, Search Console answers it for free. If a client in Dallas asks why they see a competitor above them, you need the second method. If you report on hundreds of keywords across cities every week, you need the third or the fourth.

Why rankings differ from place to place

Google tailors results to the searcher. Several of the inputs are under your control when you check rankings, and they must all point to the same place:

• IP address. Google says it estimates location from your IP “roughly based on geography”, always over an area of at least 3 km² shared by at least 1,000 users.
• Country and language parameters. gl sets the country results are for, hl the interface language. Country domains such as google.co.uk and google.de used to imply their own defaults, but Google announced in April 2025 that it is redirecting them to google.com, so set the parameters explicitly.
• Precise location. If a browser shares device location, or a signed-in account has home and work addresses, those override the IP.
• History and personalisation. A signed-in profile with search history sees a personalised page.
• Device. Mobile and desktop results differ in layout and often in order.
• Distance. Local pack and map results rank partly on how far the business is from the searcher, which is why one city over can change everything.
• AI answers. AI Overviews and AI Mode appear for some queries in some markets and push organic results down the page.

Most bad ranking reports come from one of those inputs disagreeing with the others: a German IP with an English interface, a London IP with a Manchester location parameter, a signed-in browser.

What changed in 2025–2026

Most guides still describe how Google worked before 2025. Four changes matter for anyone checking rankings today:

• 15–17 January 2025: JavaScript became required. Google began requiring JavaScript to show results, citing protection against bots. Simple HTTP fetchers stopped working; checks now run in real or headless browsers, at several times the compute.
• Around 10–14 September 2025: &num=100 stopped working. Google no longer returns 100 results per page. A study of 319 properties found impressions in Search Console fell for 87.7% of them. Checking the top 100 now takes about ten paginated requests per keyword, and many trackers reduced depth to the top 20–50.
• 11 August 2025 and 1 January 2027: official APIs are going away. Microsoft retired the Bing Search APIs; Google’s Custom Search JSON API is closed to new customers and ends on 1 January 2027.
• 19 December 2025: Google sued SerpApi. The DMCA complaint targets getting around Google’s bot protection with fake browsers and many IP addresses. In July 2026 a California court dismissed most of it, while a New York court in Reddit’s parallel case treated the same protection as an access control. Anything sold as “bypassing” Google carries legal risk; our guide to the court cases has the details.

Meanwhile Google reported AI Mode in more than 200 countries with over a billion monthly users by May 2026, so the results page you are measuring keeps changing shape too.

Method 1: Google Search Console

Search Console’s Performance report can be filtered by country. It shows your average position across everyone who saw your site in that country, which is the most honest single number you will get, because it is real searchers rather than a simulation.

It has limits. It covers only your own properties, it has no city filter, and the average blends devices unless you filter those too. It also changed meaning in September 2025: when trackers stopped loading results 11–100, many sites saw impressions fall and average position “improve” without any real ranking change. Compare periods on either side of mid-September 2025 with care.

Method 2: A clean browser and a local exit

For spot checks and client screenshots, nothing beats looking at the page. Done properly, it takes five minutes:

1. Create a fresh browser profile. Incognito alone is not enough: it still uses your own IP and may still share device location.
2. Connect the profile to an exit in the target country. With ChronoProxy that is a login such as cp_a1b2c3-cc-de-sid-check1-ttl-30: country Germany, one sticky exit held for 30 minutes so it does not change between pages.
3. Confirm where you are. Open an IP lookup and check the country, city and network. If you need a specific city, keep opening new sessions until one lands there, or use the location parameter below.
4. Stay signed out and deny location permission when the browser asks.
5. Use the market’s settings. For Germany: google.com/search?q=laufschuhe&gl=de&hl=de. Many SEOs also add pws=0 to reduce personalisation.
6. For a city, add uule. It encodes a canonical location name. It is undocumented and reverse-engineered, so generate it with a tool you trust and verify the result: Google prints the location it used at the bottom of the results page.
7. Record what you saw: a screenshot, the date, the exit IP and its city.

Why a proxy rather than a VPN? A VPN is fine for a quick look, but its exits are usually datacenter addresses shared by many users, which draw CAPTCHAs and do not resemble the people you are trying to imitate. A residential exit is an ordinary household connection in the country you are checking.

Method 3: Rank trackers

Trackers automate the second method across thousands of keywords. Before you trust one for multi-location work, check:

• How it sets location: country only, city, or ZIP and postcode, and whether it uses uule, its own exits or both.
• Depth after the num=100 change. Tools reacted differently. AccuRanker first cut its default to the top 20 and from late September 2025 tracked the top 30 daily and the top 100 every second week, SEOmonitor tracks the top 20 daily and positions 21–100 weekly, and Ahrefs restored top-100 tracking in November 2025. Ask what yours does and what it charges for depth.
• Devices. Desktop and mobile should be separate data sets.
• AI features. Whether it records AI Overviews and AI Mode citations.
• Agreement with reality. Check one keyword by hand with method 2 and compare.

Method 4: Your own tracker with proxies

Building your own makes sense when you need markets, engines or verticals that tools do not cover, or raw data for your own analysis. Since 2025 it means a headless browser, pagination for depth, careful pacing and rotating residential IPs whose country agrees with gl, hl and uule.

The cost is arithmetic. Take 500 keywords, 5 cities, desktop and mobile, the top 20 (two pages), checked daily:

So the route depends on page weight. If you can keep pages light, proxies are competitive; if you render everything, an API may be cheaper. Measure the weight of one rendered results page in your own stack before you decide; it is the number that moves the total most.

Notes for the US, UK and Germany

United States

Bing has 9.89% of US search (StatCounter, September 2026), so US reporting should include it. Desktop is still 48.5% of US page views, almost level with mobile, so track both. Local results vary by metro and even by ZIP; use a US exit plus uule for the city. Coverage and networks are on our US proxies page.

United Kingdom

Google holds 91.39% of UK search and mobile is 53.6% of page views. Remember the four nations: a client in Glasgow, Cardiff or Belfast sees different local results from London. Belfast is in the UK, so use a GB exit there. See UK proxies for cities and networks.

Germany

Use gl=de with hl=de rather than relying on google.de, which Google is redirecting to google.com; Google has 87.84% of German search, and Ecosia and Yandex are small but present. Mobile is 59.9% of German page views, so mobile rankings are the majority view. Geolocation databases use English names, so your code should expect “Cologne”, not “Köln”. Details on Germany proxies.

Which proxy type to use

Our SERP proxies page covers setup, pacing and a cost calculator.

Mistakes that skew results

• An IP in one city and a uule for another.
• Checking while signed in to a Google account.
• Old code that still asks for &num=100 and silently gets ten results.
• An English interface (hl=en) on a German exit.
• Comparing a desktop check with a mobile one.
• Solving CAPTCHAs at scale instead of slowing down.
• Treating one check as the truth. Rankings fluctuate; check twice, on different days.
• Trusting crowded VPN exits that Google already treats with suspicion.

Run a fair comparison yourself

If you want to know how much location really matters for your keywords, run a small, controlled test rather than reading someone else’s averages:

1. Pick 50 keywords that matter to the business, including some with local intent.
2. Choose two cities per market, for example New York and Dallas, London and Manchester, Berlin and München.
3. For each city: a residential exit in the country, matching gl, hl and uule, signed out, desktop and mobile.
4. Check on two dates a week apart, and save every result with the exit IP, network and city.
5. Measure how many of the top-10 URLs change between the two cities of a market, and between markets.

Fifty keywords × six cities × two devices × two dates is 1,200 checks: a few gigabytes of traffic and an afternoon of work, and the answer is specific to your business instead of an industry average.

# confirm the exit before each batch: country, city and network
curl -x gate.chronoproxy.com:7000 \
  -U "cp_a1b2c3-cc-gb-sid-test-man-ttl-30:PASS" \
  https://ipinfo.io/json

Published Oct 18, 2026. We update this guide when the platforms change; dated facts above are as of publication.