Not where you areNot where you are

Your brand search is clear from the office in Manchester: your own ad, then your shop. A colleague in London sends a screenshot of the same query, and above your ad sits one you don't recognise, “Yourbrand codes · Up to 30% off”.

Both screenshots can be accurate. A results page is built for one searcher, in one place, on one device, at one moment, and every advertiser chooses where its ads can show. Your office can be the hole in the map.

This lesson is about where. Out of hours is about when, and in one screenshot the two look alike.

How a campaign leaves a place out

Location targeting is the Google Ads setting that says where a campaign's ads can show. It takes a country, an area within one (a region, a city, a postcode district) or a radius around a point. The smallest radius is 1 km, and Google warns that very small targets may serve only now and then.

In the UK the postcode targets are districts: the outward code, the first half of a postcode, such as NR29 or M1. Google's UK location list (the edition of 6 October 2026) holds 2,705 postcode districts, 2,961 cities and 352 neighbourhoods.

Any of them can be excluded too. A campaign can target the United Kingdom and exclude Manchester, or only M1, the district around your head office. A radius can't be excluded, only targeted. A campaign holds up to 10,000 locations and 500 radius targets, exclusions included.

Exclusions work on where Google thinks the searcher is. They're Presence only: they keep the ad from people likely to be in the area (“Presence or interest”, the default, is for targeting alone). Google places a searcher mainly by IP address, plus GPS, Wi-Fi, Bluetooth and phone masts, and doesn't guarantee accuracy.

For a smaller hole there's the IP exclusion: internet addresses a campaign won't show ads to, up to 500 per campaign. In Google Ads an entry is one address or a range with an asterisk over the last block (203.0.113.*); the Google Ads API also takes CIDR blocks (203.0.113.0/24) and IPv6, each block one entry. Performance Max, video, Hotel, App and Smart display campaigns can't take campaign-level lists, but account-level lists cover every campaign type.

Five hundred is plenty for an office's public address and for a checking service on a small, fixed pool. It wouldn't hold one large cloud provider's published IPv4 ranges, of which there are more than 1,700, and Google Ads doesn't appear to offer a ready-made list of data-centre addresses. Checks from many changing UK addresses can't be listed out at all.

Then there's device. A −100% bid adjustment switches off computers, mobile or tablets, and unlike a schedule adjustment it still applies under Smart Bidding. A campaign can run on phones alone while every office check happens on a computer.

Why your own checks can mislead you

A check from your desk has three blind spots. The map: if Google places your office inside an excluded district, or your address is on an exclusion list, the page you see is true for you alone. The device: a phone-only campaign never appears on the office computer.

And repetition. Google says that if you keep searching for your own ad without clicking it, you may stop seeing it, because it stops showing ads it thinks you aren't interested in. Each search by hand also adds an impression. Your checking can change the page for you alone, so a refresh is not an independent second opinion.

Google's answer is Ad Preview and Diagnosis, in Google Ads. It gives “the exact same results as a Google search without accumulating any impressions”, for a location, language and device you choose, so it can show other advertisers' ads too. Google notes that when you target an area you're not in, your ad shows in Ad Preview but not in your own searches. It has limits: ads ranking on page two or later may be missing, it may not reflect all campaign activity, and it has no setting for the hour.

What it costs

Your office's clear page changes neither bill. Another bidder in your brand auctions elsewhere raises your price there: actual cost per click is the least needed to beat the ad ranked below yours, and bidders further down affect it too. And a shopper you'd already paid to bring in can come back through an affiliate's ad, with commission paid on the same order.

Neither cost shows from Manchester.

Where it shows in your reports

• Auction Insights on your brand campaign, segmented by device (it covers Search, Shopping and Performance Max, with display domains as its rows). A domain on mobile only, while your office checks on a computer, deserves a closer look.
• Your own absolute top impression share (how often yours was the first ad above the results, out of the times it could have been), by device, this month against last. A drop on one device can vanish in the all-devices figure.
• GA4: google / cpc sessions whose Google Ads campaign is “(not set)”, which is how Google says traffic from Ads accounts not linked to your property shows. Break them down by city and device.
• Your network: the publisher's clicks and sales by region and by device, where your network can split them.

The ordinary explanations

Most differences between two places have a plain cause. Start with these.

• Your own settings. If your office's address is on your own campaigns' IP exclusions, or your office sits outside your own targeting, your own ad won't show at your desk either.
• Google's estimate. Location comes mainly from the IP address, and a phone on mobile data can be placed by its carrier's address, some way from where it actually is.
• Budget and schedule. A campaign with a small budget can show on one search and not the next, and a scheduled one shows at some hours only.
• A territory. An approved partner, a reseller or another team at your company may run in some regions only, for a delivery area or a regional offer.
• Terms that never said. If your programme terms don't mention territories or devices, a publisher had nothing to follow.
• A mistake. An exclusion added for another reason, a copied campaign, an old one still running.

Not seeing an ad from your office proves nothing by itself. Seeing it from London proves only that it showed there, then.

What to do

1. Record every search you compare: the query, the time, the device, the location, the ad as it read, and the advertiser Google names in “About this advertiser”. A difference in schedule can pass for a difference in place.
2. Check your own campaigns' location targeting and IP exclusions, so you know what your own desk should see.
3. In Ad Preview and Diagnosis, run your brand name and “yourbrand discount code” for several UK cities, on computer, mobile and tablet. Use the diagnosis to check your own coverage in each.
4. Split Auction Insights and your absolute top impression share by device, month on month, and read GA4's “(not set)” google / cpc sessions by city and device.
5. Search the Ads Transparency Center for your brand name with the region set to the United Kingdom. Targeting doesn't hide an ad from it, and each ad shows the last date it ran.
6. Take what you found to your network as a question, with your records attached, and write any permitted territories into your programme terms.

The answers are often dull: a delivery area, a phone-only campaign, a budget gone by lunchtime. Getting them in writing is the point. What you can do covers the terms.

Seeing it in Godotoro

Godotoro searches google.co.uk from UK addresses in desktop browsers, away from your desk, and records every ad as it showed, without clicking any. An empty results page is rechecked through another proxy before it counts as empty.

Its address pane, “Ads showing your address”, doesn't depend on where anyone searches from: it lists the advertisers Google's Ads Transparency Center shows for UK ads using your web address, whether Google verified them, how many ads, and when each was first listed. Godotoro searches on desktop only, so for phones, use Ad Preview and the device split. It describes what was seen; what that means under your terms is yours to decide. To see your own, check your brand.