There are two maps in every Google Ads account.
The first is the one you drew. It lives in the settings, it is a shape on a screen, and somebody chose it once. The second is the map of where your money actually went, and almost nobody has looked at it.
They are rarely the same map, and the gap between them is not a settings problem. The settings have their own chapter. This one is about the report that tells you what the settings produced.
The Map You Drew Is Not the Map You Bought
A location target is a permission, not a prediction. It tells Google where your ads are allowed to appear. Where they actually appeared, and where the money went, is decided by demand, competition and cost inside that boundary.
So a business targeting a metro area does not spend evenly across it. It spends where the searches are, where the auctions are cheapest, and where the ads happen to win. Those places are not necessarily where the customers are.
The report exists, it is free, and it will show you the whole thing in about ten minutes. Almost every account we audit has never opened it.
Population Is a Budget Problem
Here is the pattern that shows up most often, and it is arithmetic rather than a mistake.
We audited a Southern California business running a small monthly budget across a very wide target area. The observation that mattered was not that the targeting was wrong in principle. It was that Irvine alone holds well over a million people, and the target covered far more than Irvine. A budget that size, spread across a population that size, cannot buy meaningful presence anywhere in it.
The result is a thin film of advertising across a large map. You are technically running in twenty cities and genuinely competing in none of them.
This is the same argument as impression share, viewed geographically. Owning 80% of the searches that matter in one city beats owning 8% of them in ten cities, because 8% is not a market position. It is a rounding error that costs real money.
We told one client running across several states to pick one and restructure around it, for exactly that reason. Their top impression share needed to be high somewhere, and it could not be high everywhere on that budget.
There is a rough sanity check worth doing before you read a single row of the report. Take your monthly budget and set it against the population you have told Google you want to reach. You are not looking for a precise figure. You are looking for whether the two numbers belong in the same sentence at all. When a modest local budget is pointed at several million people, no amount of tuning inside the account will fix it, because the constraint is scope rather than execution.
Two Reports, and Only One Answers the Question
This is the part that catches people, and it is worth getting right before you read a single row.
Google reports geography two ways, and they mean different things.
Matched locations shows the location Google used to match your ad. That might be where somebody physically was, or a place they had shown interest in.
User locations shows where the search was physically made from.
Those two reports can look very different on the same campaign, and the gap between them is itself a finding. If your user locations report is full of places far outside your area while your matched locations report looks tidy, you are paying to reach people who are somewhere else and thinking about you.
For a business that serves customers in person, user locations is the honest report. It answers the question you actually care about: where were the human beings who cost me money?
Matched locations still has a job. It is how you check whether Google's idea of your area lines up with yours, and it is where a mismatch between the two shows up as a pattern rather than as a single odd row. But it is a diagnostic about Google's matching, not a description of your customers, and reading it as though it were the second thing is how people conclude their targeting is fine when it is not.
A practical note on granularity. The report will go from country down to city and beyond, and most people stop at the state or metro level, which is the level at which everything looks reasonable. The finding is almost always one level further down. A metro that looks acceptable in aggregate frequently contains two or three cities absorbing most of the spend and producing none of the results.
Sort by Cost, Then Look for Conversions
The method is the same discipline as the search terms report, pointed at a different column.
Open the location report, set a date range long enough to hold real data, and sort by cost, descending. Read down from the top.
You are looking for three things.
Places taking real money with no conversions. The clearest finding in the report. A city that has spent hundreds of dollars over ninety days and produced nothing is a decision waiting to be made.
Places converting well that you were not deliberately targeting. These are the interesting ones. A neighboring town that quietly outperforms your home city is telling you something about your market that you did not know.
Places you do not serve at all. Somewhere outside your delivery radius, your license area, or your service map, absorbing budget because the boundary you drew was slightly larger than your business.
One client came to us with years of prior campaign history. Rather than starting from an assumption about where the customers were, we exported the historical performance, sorted it by location, and carried the zip codes that had actually converted into the new build. The map was not drawn from opinion. It was drawn from what had already happened.
Excluding Is Not the Same as Adjusting
Once you find waste, there are two ways to act, and they are not interchangeable.
A bid adjustment says you still want this location, just at a different price. Useful when a place performs, but less well than another, and you want to keep it at a lower cost.
An exclusion removes it entirely. Correct when you genuinely do not serve somewhere, or when it has spent real money over a real period and produced nothing.
Two cautions. First, on automated bidding strategies, manual location adjustments do not work the way they did under manual bidding, so if you are on Smart Bidding, treat exclusion rather than adjustment as your primary lever and let the strategy handle price. Second, do not exclude on a small sample. A city with four clicks and no conversions has told you nothing. Give it enough volume to be a signal before you act on it.
And a specific trap worth knowing: excluding a large region can quietly remove smaller areas inside it that were performing well. Check what sits within a boundary before you draw it.
A Radius Is a Circle. Your Customers Drive on Roads.
Distance is the assumption underneath most local targeting, and it is usually inherited rather than chosen.
We audited a wellness clinic running a 50-mile radius, and for that particular location it was reasonable. That is the point worth making. Fifty miles is not right or wrong on its own. It is right or wrong for a specific business in a specific place, and almost nobody checks which.
The awkward truth is that a circle drawn on a map is a poor model of how people travel. A river, a bridge, a state line or one badly congested highway can make somewhere fifteen miles away less reachable than somewhere thirty miles out. Distance is not the variable your customers experience. Travel time is.
You do not need mapping software to test this. The location report already contains the answer. If a town at the edge of your radius has spent money and produced nothing across a meaningful period, your customers have already told you they do not make that trip.
Where You Sell Is a Strategy, Not an Afterthought
The last point is the most commercial one, and it is easy to miss inside a report full of city names.
One therapist we spoke to had chosen her office location deliberately. She works in an affluent coastal area, does not take insurance, and her clients pay cash rates. The geography was not incidental to the business model. It was part of the business model.
That reframes what the location report is for. Waste is only half of what it shows. The other half is a description of which markets your offer genuinely lands in, and two areas can carry very different customer value even when the cost per click between them looks identical.
Which means the right question is rarely "how do I cover more ground." It is "where is the version of this customer who is worth the most to me, and am I present enough there to win." Sometimes the correct move after reading the report is to spend more money in a smaller area, which feels like a retreat and is usually a promotion.
The Twenty Minute Version
- Open the location report. Set a date range with enough volume to be meaningful, not last week.
- Read user locations, not just matched locations. One tells you where people physically were. That is the question you care about.
- Sort by cost, descending. Read the top rows. That is where the money went.
- Flag three groups. Spending with no conversions, converting without being deliberately targeted, and anywhere outside your actual service area.
- Exclude what you do not serve. Immediately, and without waiting for more data. You already know the answer for these.
- Wait on the rest. Do not act on a location with a handful of clicks. That is noise.
- Compare the map to your budget. If your budget divided by your target population is a very small number, the problem is not optimization. It is scope.
- Check what is inside a region before excluding it. Large exclusions swallow smaller areas that may be working.
Geography is one of only two decisions that govern every auction in a search account. The other is which keywords you buy. Both were set once, usually early, usually quickly, and both keep running exactly as configured until somebody looks.
We read the location report on every Google Ads audit, and the same slide keeps appearing: real money in places the business cannot serve, and thin coverage in the one place it can. Hand it to us and we will map it properly. Or open the report yourself and sort by cost. The first five rows will tell you quickly whether your map and your business still agree.
