What a Toronto ad looks like from Vancouver
Search "emergency plumber" from your desk in Burnaby and the ads that load are bidding for you, in Burnaby. That Vancouver plumber never sees the Toronto shop that outranks them downtown. Same keyword, different city, different auction. Google runs a separate ad auction per location, and the sponsored results reshuffle the instant the location does.
So planning a campaign for a market you don't physically sit in gets awkward. You can't search from home and trust the screen. Your location is one of the loudest signals Google reads, and it localizes the whole ad stack around it.
Two honest ways to see what a searcher in another city actually gets. Plus one official disclosure library that answers a different question entirely.
The uule parameter does the location work
Google encodes a location into the search URL with three parameters: gl for country, hl for interface language, and uule for the exact place. The uule pins you to a city. It's a base64-encoded string built from a canonical location name, the precise wording Google keeps in its Geotargets dataset, which they publish as a downloadable CSV of every targetable location and its criteria ID.
Get the canonical name wrong and the encoding fails without a peep. "Toronto" isn't canonical. "Toronto,Ontario,Canada" is. The dataset is the source of truth. Guess at a city string and you'll usually get the wrong SERP, or a quiet fallback to your real IP.
Once the uule is right, Google treats the request as if it came from that spot. The ads render for that auction. The local pack shifts. You see what a searcher there sees, sponsored results and all, which is the exact thing you're setting a budget against.
That's what anywhereserp does: it builds the gl/hl/uule URL straight from Google's Geotargets list and shows you the live SERP for that location. Compare two locations side by side, Vancouver against Calgary, and read the ad stack in each without leaving your chair.
The location field in Google Ads shows you the preview inside the platform
Run Google Ads and you get the Ad Preview and Diagnosis tool inside the account. Set a location and language, and it tells you whether and where your ad shows, without racking up impressions or clicks, because the preview isn't a real search. Google built it partly so advertisers stop Googling their own keywords and poisoning their click-through data.
The catch for competitor research: Ad Preview shows your ads and the results around them, so you can see who you're fighting in that city's auction. But it's tethered to your account and your keywords. Great for "am I showing in Edmonton for this term." Useless for "show me the full paid landscape in a city I don't advertise in yet."
That gap is where a raw geo-targeted SERP earns its keep. The question isn't "where do my ads rank." It's "what does the whole first page, ads included, look like to someone standing in that market."
The Ads Transparency Center shows the ads, but not the auction
Google's Ads Transparency Center is a searchable library of ads that have run on Google's platforms. Look up an advertiser, filter by region and date, see the creatives they served. For a competitor you already know by name, it's a clean way to see what they're actually running and where.
What it won't give you is the live auction. It shows ads that ran, not the ranked stack a searcher gets right now for a specific keyword in a specific city. Two different questions. The Transparency Center answers "what has this advertiser been showing?" A geo-targeted SERP answers "who wins the top of the page for this query in this market today?"
Both earn a spot. Scoping a new city, I start with the live SERP to read the shape of the fight, how many advertisers are bidding, who owns the top slot, whether a Shopping carousel is eating the fold, then use the Transparency Center to pull a named competitor's full creative history.
Why the VPN answer is shakier than it looks
A VPN swaps your IP for an exit node somewhere else, and Google reads that IP as a location signal. Set the exit to Toronto, search, get a Toronto-flavoured result. It works, roughly.
Precision and repeatability are where it wobbles. VPN exit nodes cluster in a few data-center cities, so you can usually hit a country and often a metro, but rarely a specific suburb or a mid-size town. IP geolocation is fuzzy on its own, since the IP maps to wherever the ISP or VPN registered the block, which isn't always where you think. And Google stacks your account history, past locations, and language settings on top of the IP, so two people on the same node can see different ads.
I've written before about VPN versus URL parameters. For ad research the gap gets sharper. The uule string is a named place from Google's own dataset, so "Kelowna,British Columbia,Canada" means the same thing every single run. A VPN gives you "somewhere near a Toronto data center." One is a coordinate you can drop in a client report. The other is a guess you can't reproduce next week.
The Canadian wrinkle: language changes the ad stack too
A search in Montreal isn't only a location change. It's often a language change. Set hl=fr and Google serves a French interface, and advertisers chasing French-speaking Quebec show different ads than the same query at hl=en. A national brand might run English creative in Ontario and French creative in Quebec, and you only catch both if you flip the language parameter, not just the location.
This is the part most rank checks fumble. Run one English-language ad audit across Canada and you learn what shows to English searchers while the French auction slips past you unnoticed. For any competitor working both Quebec and the rest of Canada, you're staring at two separate paid landscapes stacked on one keyword, and the only way to see both is to vary hl alongside uule.
The ads a competitor buys are a location-specific auction. Read them from the wrong city, or the wrong language, and you've learned about your own market, not theirs.