The number you don't have yet: zero

Open Auction Insights for a city you've never advertised in. You get nothing.

That's not a bug. The report stays blank until your impression share clears 10%. Before you spend a dollar in Calgary or Seattle, your impression share there is zero. The one native Google report built to show competitive pressure is blind to any market you haven't already paid to enter.

So the person deciding whether Edmonton is worth a test budget, the exact moment you most need to know who you're up against, is the one person Google won't tell. You guess, or you spend first and learn later.

There's a third option, and it's why I keep a geo-SERP preview open next to my planning spreadsheet: look at the live paid results in that city before you commit. That's the only pre-spend view of the competition that exists.

What the live SERP shows that Auction Insights can't

Auction Insights lists domains and six metrics. It doesn't show a single ad. No copy, no extensions, no landing pages. Google's own docs tell you to research all of that separately. That word, "separately," is the whole gap.

Load the search for your money keyword as if you were standing in the target city, and you get everything the report withholds: how many paid slots stack above the organic results, what the top advertisers actually say, which extensions they run, whether a local pack is jammed in the middle. You see the page a real searcher there sees.

Same three parameters run behind any localized Google result. gl sets the country, hl sets the interface language, and uule encodes the canonical place you're pretending to search from. Change the uule from Burnaby to Austin and the ad stack changes with it. Different advertisers, different offers, sometimes a completely different number of ads. You can compare two cities side by side and read the competitive density straight off the page.

This matters because Google decides who sees your ad partly on where the searcher is. A person in Vancouver and one in Dallas can trigger entirely different auctions for the same words. The report averages across your account. The live SERP is specific to the place.

Real competitors, not modeled ones, but only after you pay

The best thing about Auction Insights is its source. It runs on actual auction data, so you're seeing real competitors, not theoretical ones. Compare that to the scrape-based tools. Semrush and SpyFu model competitive data from scraped results and clickstream panels; Auction Insights comes straight out of Google's auction logs.

The catch is the entry fee. That first-party accuracy only switches on once you're spending enough to clear the threshold. A geo-SERP preview isn't drawn from auction logs either. It's the rendered page. But it does one thing Auction Insights can't do before launch: it works when your spend is zero, and it shows the creative.

Neither view tells you what anyone bids. You can infer relative aggressiveness, but you won't know if a competitor is at $5 or $50. A crowded ad stack tells you the fight is on. It doesn't price it. Keep that honest.

Reading the page like a bidder

Say you sell in Vancouver and you're eyeing Calgary. Pull the Calgary SERP for your top three keywords. Count the ads above the fold. Note who owns the top slot on all three, because that's your persistent competitor, the one who'll show up later in Auction Insights with high overlap. Overlap rate is just the share of times your ads and theirs appear together. You're eyeballing that figure before you have an account to measure it in.

Then read what the leaders promise. Free quote. Same-day. A price on the ad. Those offers set the floor you'll have to match or beat. This is exactly the creative research Auction Insights refuses to give you, done for free, in the target market, before your card gets charged.

One caution. A localized SERP is a snapshot, not a stakeout. Ad rotation, dayparting, and budgets that run dry mid-afternoon mean the 9am stack isn't the 4pm stack. If your budget runs out, you stop showing and lose sight of what competitors do next. Your preview is just as fragile: an advertiser who's burned through today's budget simply isn't there when you look. Check a keyword at two different times before you trust the count.

Where this fits in a Canadian rollout

Expanding across provinces breaks the "one market" assumption fast. A Toronto searcher and a Montreal searcher can see different advertisers, and if you flip hl to French, different ad copy in a different language. That's a wrinkle worth its own bilingual SERP audit. A national dashboard averages all of it into one misleading line.

The pre-spend preview and the post-spend report answer different questions, and I'd argue you need both. Auction Insights, once it's live, tells you which competitors land in the same auctions as you and who wins when you're both there. You rarely decide on that report alone, but it adds context. The geo-SERP preview is what you're allowed to look at while the report still reads zero.

Preview Google Ads competition in a new city before you spend

The number you don't have yet: zero

What the live SERP shows that Auction Insights can't

Real competitors, not modeled ones, but only after you pay

Reading the page like a bidder

Where this fits in a Canadian rollout

Sources

One email a month, when Google's geo-search changes