You Set Your Location - Google Interpreted It Generously

You set your location carefully. Your town, maybe the county, the area you actually cover.

Then somebody two hundred miles away, who is never going to hire you, saw your advert and clicked it.

That is not a bug, and nobody has hacked your account. It is a setting; it is switched on by default, and most business owners have never seen it.

The setting has a name

Every new campaign defaults to a location option usually described as presence or interest.

It reaches three groups. People physically in the area you targeted. People who regularly visit it. And people who’ve shown search interest in it, from wherever they happen to be sitting.

That third group is the one that costs money.

Somebody in Cornwall searching for a roofer in Leeds matches a Leeds roofing campaign under that default. They’re not moving to Leeds. They are probably comparing prices, researching for a relative, or simply curious.

You paid for that click at the same rate as a genuine local one.

The stricter alternative reaches only people in your area or regularly visiting it. It is one dropdown, and it is not the default.

For a lot of local service businesses, that single change is the most valuable ten seconds available in the whole account.

WHY THE DEFAULT EXISTS AT ALL

Most articles on this treat the default as a trap. It is worth being fairer than that, because the reason matters for whether you should change it.

For some businesses the wider setting is correct.

A hotel in York wants people searching for York from anywhere in the country. They are not in York yet. That is the entire point.

A holiday let, a conference venue, a wedding photographer taking bookings from out of town, a university, a removals firm handling moves into the area. All of these have genuine customers who are physically elsewhere while searching.

For those businesses, restricting to presence would cut real demand.

Where the default is wrong

It is wrong when your customer has to be near you to buy.

An emergency plumber. A commercial cleaner. A dentist. A domiciliary care provider. A skip hire firm.

If somebody physically distant can’t become a customer, serving them is a waste by definition. However interested they seemed.

The honest test

Ask one question about your business. Can somebody who is not in my area, and is not coming to my area, ever buy from me?

If the answer is a clear no, switch to presence and stop paying to be interesting to strangers.

If the answer is yes, keep the wider setting and manage it with exclusions instead. More on that shortly.

HOW TO FIND OUT WHAT IT COST YOU

Don’t take a general warning as evidence about your own account. There’s a report that answers it precisely.

It is the geographic report, and like most of the useful screens in Google Ads, almost nobody opens it.

The twenty-minute version

Open the geographic report and set the date range to the last ninety days. A short window will not show you enough to be sure.
Segment or view by the location where the user actually was, rather than by the location they showed interest in. Those are two different columns, and confusing them is the classic mistake here.
Sort by cost, highest first. You are asking where the money went.
Go down the list and mark every area that cannot become a customer. Not areas that look unfamiliar, but areas that genuinely cannot buy from you.
Add up the cost of everything you marked. That total is your answer, and it is specific to you rather than an average from an article.
Check the conversions column against those same rows. Occasionally a distant area converts, which is worth investigating before you exclude it.

Most local service accounts find something. Some find a great deal.

The two columns people confuse

This trips up almost everyone the first time, so it is worth being explicit.

One column tells you where the person physically was. The other tells you which location they were interested in.

Under the default setting, those diverge constantly, and reading the wrong one tells you the opposite of the truth. Somebody in Cornwall interested in Leeds appears as Leeds in one column and Cornwall in the other.

The one you care about, for waste, is where they actually were.

THE FIX THAT DOES NOT WORK

Here is the part most articles get wrong, and it matters because the wrong fix feels like doing something.

The instinct, when an area performs badly, is to reduce your bids there. Set a location bid adjustment of minus fifty per cent and move on.

That worked when everyone bid manually. It does not work the way people expect on an automated bidding strategy.

Why it fails

Target CPA, Target ROAS and Maximise Conversions set bids using their own signals at auction time. A percentage adjustment applied to a location is not a reliable brake on those strategies the way it was on manual cost per click.

So you apply the adjustment and feel the problem is handled. The traffic keeps arriving.

Three months later, somebody notices the area is still spending. They decide the adjustment wasn’t aggressive enough and make it minus ninety. That does very little either.

What to do instead

If an area cannot become a customer, exclude it. Exclusions are a boundary rather than a preference.

Exclusions override targeting, so an excluded area will not be served even if it sits inside a radius you targeted. That makes them the reliable tool for carving an unwanted patch out of a broad target.

On an automated bidding strategy, the choice is in or out. Bidding down is a comfort, not a control.

This applies to more than location, incidentally. It is the same reason that time-of-day bid reductions rarely do what people expect on automated strategies, and why turning something off beats turning it down.

WHAT GOOGLE ACTUALLY KNOWS ABOUT WHERE SOMEBODY IS

Worth understanding, because it explains why location targeting is never perfectly precise and stops you from chasing an accuracy that does not exist.

Location is inferred, not read off a label. It comes from a mix of signals. The device, the network, previously visited places, and what the person has been searching for.

That works well most of the time and imperfectly some of the time.

Where it gets fuzzy

Somebody on a work network can appear to be at their head office rather than their desk. Somebody using a privacy tool can appear somewhere else entirely. Mobile signal can place a person in the next town over.

None of that is common enough to ruin a campaign. It is common enough that a handful of odd rows in your geographic report are normal rather than sinister.

What this means practically

Do not build very tight radius targeting on the assumption of street-level accuracy. A three-mile radius in a city is a reasonable instruction. A one-mile radius is asking for a precision the system cannot reliably deliver.

And don’t panic at a few impressions from somewhere unexpected. Look at the cost. If it is a few pounds across ninety days, that is noise. If it is a meaningful share of the quarter, that is a signal.

The distinction matters. Businesses sometimes tighten their targeting into uselessness, chasing a few stray rows that were never costing them anything.

MULTI-LOCATION AND MULTI-AREA BUSINESSES

If you serve more than one distinct area, the structure of your account matters as much as the settings.

The common mistake is one campaign covering everywhere, with all the areas thrown into a single location list.

Why that causes problems

You lose the ability to control budget by area. One strong area quietly absorbs it, and a weaker one gets nothing, even though the weaker one might be the growth opportunity.

You lose the ability to write for the area. A single set of ads has to serve everywhere, so the copy drifts towards the generic.

And you lose clean reporting. Everything is aggregated, so you cannot easily tell which area is actually profitable.

The better structure

For genuinely distinct areas, separate campaigns are usually worth the extra management. Each gets its own budget, its own ads mentioning the area, and its own reporting line.

That also lets you serve areas differently. A city centre might justify a higher budget and tighter targeting, while an outlying town runs modestly.

When not to split

If your budget is small, splitting it across five campaigns leaves each too thin. They can’t gather data or leave the learning phase.

There is a floor below which more structure makes things worse. If you are spending a few hundred pounds a month, one well-targeted campaign usually beats five starved ones.

Split when the budget can support it and the areas genuinely behave differently. Not because the structure looks tidier.

REPORTING ON LOCATION PROPERLY

Once the settings are right, this becomes a monthly habit rather than a project.

Three things worth looking at, and none of them take long.

Cost by actual location

The same report as the audit, run monthly rather than once. New areas appear over time, particularly on broad match or automated campaigns.

Ten minutes a month keeps it clean once the big offenders are already excluded.

Enquiries by area, counted by you

This is the one the platform cannot do for you.

Keep a simple note of which area each genuine enquiry came from. Over a few months you will see which parts of your patch actually produce work, and it is frequently not the parts people assume.

Businesses are often surprised. A neighbouring town outperforms their own postcode, or a suburb they never think about produces the best jobs.

Distance against value

If you travel to customers, the useful comparison is not just whether an area converts. It is what a job there is worth after the travel.

A job forty minutes away at the same price as one ten minutes away is a worse job. Your account cannot know that. You can.

That calculation is what should drive your radius, more than any best practice about mileage.

RADIUS OR NAMED AREAS

The other question people ask, and the answer depends on how the work is delivered rather than on best practice.

Radius targeting draws a circle around a point. It suits businesses that travel to the customer from a base, because a circle roughly matches a drive time.

Named area targeting selects towns, boroughs, counties or postcode districts. It suits businesses whose customers think in those terms, and it is far easier to report on afterwards.

When radius is right

A mobile trade working from a yard. A domiciliary service. Anything where the real constraint is how far the van goes.

Set the radius to the distance you would actually travel for a job, not the distance you would travel for an exceptional job. Aspiration in a radius setting is expensive.

When named areas are right

A business whose customers say the name of a place. A solicitor covering a borough. A dentist whose patients search by town.

It also makes the geographic report easier to read. The rows match how you think about your patch.

Using both

Plenty of accounts target a radius and then exclude specific areas inside it. A river, a tolled crossing or a congestion zone can make a nearby area commercially unattractive, even when it’s geographically close.

The map does not know about the traffic. You do.

THE OTHER LOCATION SETTING, ARRIVING THIS AUTUMN

Worth flagging, because it belongs in this conversation and it is new.

Locations of interest is one of the named components of AI Max, and it operates at the ad group level. AI Max is the layer Google is applying to Search campaigns, with migration for certain older settings running through September 2026.

Which means location behaviour is one of the things changing for accounts moving over.

The practical advice is the same as it is for everything else in that migration. Find out whether it is on, decide deliberately rather than by default, and record what your geographic report looked like beforehand so you have something to compare against.

If you have just done the work of tightening your locations, it would be a shame to have that quietly widened again without noticing.

WHAT ABOUT LANGUAGE AND COUNTRY

A related trap, because it produces the same symptom and has a different cause.

If your ads appear in another country entirely, check two things beyond the location setting.

First, whether an additional country was ever added to the targeted locations. It happens more often than you would think, usually during setup or when somebody duplicated a campaign.

Second, the language setting. Language targeting works on the person’s browser and account settings, not on the words they typed. So a very broad selection can widen who’s eligible in ways that surprise people.

Neither is exotic. Both are worth two minutes when the geographic report shows somewhere unexpected.

A WORKED EXAMPLE, START TO FINISH

A hypothetical to make it concrete. A domestic electrician covering one city and about fifteen miles around it, spending seven hundred pounds a month.

They target the city by name and a thirty-mile radius, and they have never touched the location options. So the campaign runs on presence or interest.

They open the geographic report, ninety days, by where the user actually was, sorted by cost.

The top rows are their own city, as expected. Then a run of towns well outside the radius. Then a scattering of cities two and three hundred miles away.

Adding up everything that cannot become a customer comes to a meaningful share of the quarter’s spend. Not catastrophic, but the sort of number that would have paid for a month of advertising.

What they change

The location option goes from presence or interest to presence. That is the ten-second fix, and it removes most of the distant traffic on its own.

The radius comes down from thirty miles to eighteen, which is the distance they would genuinely travel on a normal weekday.

Two areas inside the radius get excluded. One is across a river with no nearby crossing, so a twelve-mile job is a fifty-minute drive. The other is a town where they have never once won work.

They do not apply any bid adjustments, because the campaign runs on Maximise Conversions and adjustments would not reliably hold anything back.

What happens next

Impressions fall noticeably. Clicks fall. The reported click-through rate barely moves, because the traffic they removed was not clicking much anyway.

Enquiries hold roughly steady, on lower spend, which is the entire point.

The freed budget goes back into the same campaign rather than being cut, so they buy more of the traffic that was already working.

Nothing sophisticated happened. Somebody read a report and told the account the truth about where the customers are.

WHEN YOUR SERVICE AREA CHANGES

One habit worth building, because location settings are set once and then forgotten for years.

Your service area isn’t fixed. It changes when you hire or lose staff, when you buy a second van, when a competitor opens nearby, or when fuel costs shift what a distant job is worth.

The settings rarely change with it.

The moments worth a review

When you take on staff or lose them. Capacity is what a radius really represents, and it moves.

When you open or close a location. Obvious, and still frequently missed on the advertising side for months.

When you change what you charge. A price rise can make a longer drive viable. A price cut can make it pointless.

And seasonally, if your work is seasonal. A gardener’s sensible radius in June is not the same as in December, when there is less competition for their time and fewer jobs around.

The five-minute version

Put a note in the calendar twice a year. Open the location settings, look at the radius, and ask whether it still matches how far the business actually goes today.

Most years the answer is yes, and it costs you five minutes. The year the answer is no, it will have been costing you money for a while.

This is the least glamorous advice in the article and one of the more valuable, because nothing in the account will ever prompt you to do it.

THE MISTAKES THAT WASTE THE MOST MONEY

Leaving every campaign on the presence or interest default when your customer has to be local to buy.
Reading the interested-in column rather than the where-they-actually-were column, and drawing the opposite conclusion.
Applying a location bid reduction on an automated bidding strategy and assuming the problem is handled.
Setting a radius based on the furthest job you would accept rather than the distance you normally travel.
Never opening the geographic report, so the whole thing stays theoretical.
Excluding a distant area that was quietly converting, without checking the conversions column first.
Forgetting that an extra country or a very broad language setting can produce the same symptom.
Tightening locations carefully and then letting a migration widen them again without noticing.

FREQUENTLY ASKED QUESTIONS

What is presence or interest in Google Ads?

It is the default location option on new campaigns.

It serves people physically in your targeted area, people who regularly visit it, and people who have shown search interest in it from elsewhere.

The stricter option, presence, covers only the first two groups.

Why are my Google ads showing in the wrong area?

Almost always the default setting, which makes somebody hundreds of miles away eligible if they searched about your town.

Open the geographic report before assuming, and read the column showing where the person actually was.

If that still does not explain it, check whether an extra country was added to your targeting.

How do I stop my ads showing outside my area?

Change the location option from presence or interest to presence.

Then add explicit exclusions for anywhere that still appears in your report and cannot become a customer.

Exclusions override targeting, so they will carve an area out of a radius reliably.

Do location bid adjustments work with Smart Bidding?

Not the way they did under manual bidding.

Automated strategies set bids from their own signals at auction time, so a percentage reduction on a location is not a dependable brake.

If an area is not worth serving, exclude it. On an automated strategy, the real choice is in or out.

Should I use radius targeting or city targeting?

Radius suits businesses that travel to customers from a base, because it approximates drive time.

Named areas suit businesses whose customers think in terms of towns, and the reporting is much easier to read.

Using both, with exclusions inside the radius, is common and sensible.

How do I exclude a location in Google Ads?

In the campaign’s location settings, add excluded locations alongside your targeted ones.

Because exclusions take precedence over targeting, an excluded area stays unserved even if it falls within a radius you selected.

That precedence is what makes exclusions the reliable tool rather than bid adjustments.

Will tightening my locations reduce my leads?

Your impressions and clicks will fall, sometimes sharply.

Enquiries should hold roughly steady, because the traffic you removed was largely not enquiring.

If genuine enquiries drop in proportion, you cut too hard. Widen the radius slightly rather than reverting the presence setting.

WHERE TO START THIS WEEK

Two jobs, and the first is almost instant.

Open one campaign and look at the location options. If it is on presence or interest and your customers have to be near you to buy, change it. That is the single highest return ten seconds in this whole article.

Then open the geographic report. Ninety days, by where the user actually was, sorted by cost. Read the top thirty rows and mark anything that cannot become a customer.

Exclude those areas rather than bidding them down, because on an automated strategy a bid-down is a comfort rather than a control.

One caveat before you measure any of it. If your conversion tracking is unreliable, you will not be able to tell whether enquiries held steady after the change, which is the only number that decides whether you did the right thing.

The map in your account is not a description of your business. It is an instruction, and by default it is a generous one.

Did You Enjoy This Blog Post?

I hope you enjoyed this blog post, and thank you so much for being here. We also upload videos to our YouTube channel every weekday. Please subscribe so you are one of the first to be notified.

If you enjoyed this blog, you may also like: