Why Google Ads Shows Ads Outside Your Service Area

A plumber in Tampa gets a call from Ohio. A cleaning company targeting three ZIP codes sees clicks from a state it has never worked in. The owner assumes Google is broken, or that the agency fat-fingered a setting.

Neither. The account is doing exactly what Google ships it to do. There is one radio button behind almost all of it, and it is set to the leaky option by default.

Why are your ads showing to people outside your service area?

Because Google Ads targets your chosen location on two signals, not one: where a person actually is, and where a person seems interested in. The default setting, called 'Presence or interest', includes both. Someone in Ohio who searched about Tampa is eligible to see your Tampa ad.

This is not a bug or an obscure edge case. Google's advanced location options documentation describes 'Presence or interest' as the default and recommended option, and calls it broad geo targeting.

So the money leaves your account in a way that looks legitimate on every report you normally check. Impressions look healthy. Clicks look healthy. Only the phone tells you something is wrong — and by then you have paid for the click.

What does 'Presence or interest' actually target?

It targets two separate populations that Google merges into one setting. Per Google's documentation on targeting geographic locations, physical location is inferred from IP address, GPS, Wi-Fi, Bluetooth and Google's cell-tower database. Location of interest is inferred from something else entirely.

Google lists the interest signals explicitly. Any one of them can put a stranger into your local campaign:

• Terms used in searches that indicate a location — the obvious one. Someone types your city into the query.
• Past searches that indicated a location of interest — they searched your city last week. Not today.
• A person's past physical locations — they visited once. They are home now.
• The content and context of a website where an ad is displayed — a Display placement about your city. Google itself warns that 'the mention of a location on a page doesn't always indicate an interest in that location.'
• Searches on Google Maps or Google Maps for Mobile — panning around a map counts.
• A custom location set in Google search settings — the searcher told Google to pretend it is somewhere else.

Google is blunt that interest is not bounded by geography. Its documentation states that location of interest 'isn't restricted to a person's country, or the Google search domain the person is searching on' — the example given is someone in Paris searching for a Los Angeles taxi on google.fr and being matched to Los Angeles.

Read that as a service-business owner. If a French search can match Los Angeles, an Ohio search can match Tampa. There is no protective radius around interest.

One more default worth knowing: Google's 'related areas' behavior. When you target a city, Google may also show your ad in nearby areas that are too small or too data-poor to target directly. Google's own example is targeting Portland and having ads served to people in the Portland suburbs. For a business that charges a trip fee by ZIP code, 'nearby' is not a rounding error.

Where is the location setting hidden in the Google Ads interface?

It is inside the campaign's Locations panel, collapsed under location options — not on the main settings screen where you picked your city. You have to expand it to see that a choice was ever made on your behalf.

There are now exactly two options for who you target, and Google has removed the third:

That removal is documented at the API level, and it matters. Google's advanced location options page states that positive_geo_target_type can no longer be set to SEARCH_INTEREST for Search, Display or Shopping campaigns, and that the default value is PRESENCE_OR_INTEREST. When asked whether advertisers can opt out of the migration, Google's answer in its own FAQ is one word: 'No.'

So you cannot escape the default by refusing the change. You can only override it, per campaign, by selecting 'Presence'. Google warns that most campaigns see fewer impressions when they switch away from the default. For a local service business, that is the point — those impressions were never yours.

How do you prove the leak in the geographic report?

Open the Locations report and switch from the 'Targeted locations' view to the 'Matched locations' view. Targeted locations shows what you asked for. Per Google's documentation on measuring geographic performance, matched locations shows 'the locations where your ads appeared' — and those 'could be users' physical locations or locations of interest.' The gap between the two views is your leak.

This distinction is the whole diagnostic, and most people never see it because the default view flatters you. The 'Targeted locations' view can only ever show the cities you chose. It is a mirror, not a report.

The Google Ads API makes the split explicit. Its geographic_view resource carries a location_type field whose values are AREA_OF_INTEREST and LOCATION_OF_PRESENCE — two different things Google tracks separately and reports together. If you segment your geographic data by that field, the interest-matched spend stops hiding inside your city totals.

Two traps when you read the numbers. First, the 'Most specific location' column will not settle the argument for you: Google says it 'may either be your potential customer's physical location or a location that they've shown interest in.' It does not label which. Second, expect an 'Unspecified' bucket — Google attributes it to cases where it cannot resolve the IP address and the query named no recognizable location, or where the search happened on Maps across a large area.

Google also notes that metrics below country level may not add up to campaign totals, because some impressions are only ever matched at the country level. Do not spend an afternoon reconciling that to the penny. You are looking for a pattern, not an audit trail.

If your matched-locations view is showing spend in states you cannot legally or physically serve, you have found it. That single report is one of the first things we pull in a free Google Ads audit, because it costs nothing to check and it is wrong constantly.

Should you target by radius, city, or ZIP code?

For a van-based service business, radius targeting is usually right, and Google requires a minimum radius of 1 km — it will not let you target smaller. City targeting is right when your service area follows municipal lines. ZIP targeting is right only when you already know your profitable ZIPs from closed revenue, not from clicks.

The honest verdict: start with radius. It maps to the physical constraint that actually governs your business — how far you will send a truck — and it does not require data you do not have yet. Move to ZIP-level targeting only after you can rank ZIPs by closed jobs, not by clicks or by leads.

Google is candid about the limits here. Its documentation states that location targeting is based on signals including user settings, devices and behavior, and that '100% accuracy is not guaranteed in every situation.' Anyone selling you perfect geo-fencing on Google Search is selling you something Google does not offer. We cover what proximity targeting can and cannot do in our guide to geofencing ads for service businesses.

Do excluded locations really block anyone?

Excluded locations block people who are physically in them — and nothing else. Google's advanced location options documentation is explicit: negative_geo_target_type can no longer be set to PRESENCE_OR_INTEREST, and the default is PRESENCE. Exclusions are presence-only, by force.

This kills the most common advice on the forums. 'Just exclude the other states' cannot remove interest-matched traffic, because exclusions do not operate on interest at all. You would also be playing whack-a-mole against every geography on earth, forever, one exclusion at a time.

Exclusions are a scalpel, not a fence. Use them for a specific area inside your target that you do not want — a neighboring county you refuse to drive to, a city where your close rate is terrible. Do not use them as a substitute for fixing the targeting setting. Fix the setting; then exclude.

The same logic applies to your keyword layer. Location settings decide who is eligible; negative keywords decide which searches you refuse to pay for. They are two different valves, and a leaky account usually has both open.

How much wasted spend does the default setting cost you?

Nobody can tell you a universal number, and any agency that quotes you one is guessing. The honest answer is that your own matched-locations report already contains it: out-of-area clicks multiplied by your average CPC, multiplied out to a month, is your leak. Pull the last 90 days and do that arithmetic before you renew anything.

What we can do is examine the only number Google publishes to justify the default. Google's own tip states that advertisers who switched from 'Presence' to 'Presence or interest' in the Travel, Real Estate, and Education verticals see 5% more conversions on Search campaigns, sourced to Google internal data from a window in 2022.

Now look at which verticals those are. Travel, real estate and education are three categories where wanting a place you are not currently standing in is the entire business model. Booking a hotel in a city you are not in is the product. That 5% is real for them.

A plumber is the exact inverse. Nobody in Ohio is a qualified lead for a burst pipe in Tampa, no matter how interested in Tampa they are. Google's recommended default is optimized for a business model that is the opposite of yours — and Google does not publish a figure for local service businesses, because the answer would not flatter the default.

So the decision is not close. If you dispatch a vehicle, switch to 'Presence'. Accept the impression drop. Then judge the change on booked jobs over the next 30 days, not on impressions, clicks, or CTR — the three metrics that will all get worse while your business gets better.

What should you do in the next twenty minutes?

Four checks, and none of them require a login you do not already have. This is the sequence we would run on any local account before touching bids, budgets or copy.

• Open location options on every campaign. If it says 'Presence or interest' and you dispatch a vehicle, change it to 'Presence'.
• Switch the Locations report to matched locations. Look for spend in places you do not serve. That column is the evidence.
• Check your radius against drive time, not miles. A 25-mile radius across a metro at 5pm is not a 25-mile service area.
• Date-stamp the change. Compare booked jobs for 30 days after, against 30 days before. Not clicks.

A leaking geo setting is one of the cheapest problems in paid search to fix and one of the most expensive to ignore, because it never announces itself. It just quietly bills you.

If you would rather have someone check it for you, our paid ads service starts with the geographic report, the search terms report, and the conversion setup — before anyone touches your budget. No lock-in, month to month, and you keep the account. Get my free audit.