Google Ads Wasted Spend by Device: Where It Hides

The search terms report tells you what people searched. It doesn't tell you what they searched on. One term can make money on desktop and burn money on mobile. Added up across devices, it looks average, and you never touch it. That's device-level waste, and you won't see it until you split the numbers by device.

What device-level waste actually is

Device-level waste is money spent on one device for clicks that don't convert there, even when the same keywords and terms do convert on other devices. It isn't the same as an irrelevant search term. An irrelevant term like "plumber salary" wastes money on every device, and a negative keyword fixes it. Device waste is a term like "boiler replacement cost". It might bring quote requests from desktop and almost nothing from phones.

The search terms report adds up all devices. Say "boiler replacement cost" cost $400 last month and brought 4 leads. That's $100 per lead, which might be fine. But if desktop spent $150 for 4 leads and mobile spent $250 for none, you have a $250 problem that the total hides.

Three things make device waste different from search term waste:

• The term is fine. Blocking it would lose the desktop leads too.
• Negative keywords can't fix it. A negative applies on every device. Google Ads has no setting for "block this term on mobile only".
• The cause is often outside the search. A slow mobile page, a form that's hard to fill in on a phone, or a phone call you don't track can all look like mobile waste.

So keep two questions apart. Is this term worth paying for at all? That's a search terms question. Is this term worth paying for on this device? That's a device question, and it needs different tools.

How to find wasted spend in the device report

Start at campaign level, segmented by device. Compare cost, conversions and cost per conversion for mobile, desktop and tablet over at least 30 days, or 90 days for small accounts. In the Google Ads interface, open Campaigns, click Segment, then Device. Every row splits into one line per device.

Here's an example account spending $3,000 a month on one Search campaign:

At campaign level the account looks fine: 27 conversions at about $111 each. Split by device, it looks very different. Mobile takes 60% of the budget and converts at more than twice desktop's cost. Tablet spent $150 and got nothing.

Look for these three patterns

One device with zero conversions on real spend. In this account that's tablet. $150 with nothing to show for it is worth a look. But $150 is also a small sample. If your average cost per conversion is $111, a device that spent $150 without converting hasn't proven much yet. Decide on a threshold before you look. A common rule is to give a device at least two to three times your target cost per conversion before you call it a loser.

One device converting at a much higher cost. Mobile at $150 against desktop at $70 is a big gap. A gap of 10–20% is usually noise or normal behaviour. A gap of two times or more, on a lot of spend, is worth digging into.

A device whose share of spend keeps growing. Compare this quarter with last quarter. If mobile went from half the spend to two-thirds and conversions didn't follow, something changed: a new broad match keyword, a bidding strategy change, or a landing page update that broke on phones.

Then go one level down: search terms by device

Campaign totals tell you where the gap is. Search terms by device tell you why. In the interface, open the search terms report and apply the same Device segment. Sort by cost and look at your 20 to 30 most expensive terms.

You'll usually find a few kinds of term:

• Terms that waste on every device. "how to fix a leaking tap", "plumbing apprenticeship", "plumber salary". These are ordinary search term problems. Add a negative and move on. Job-related terms are a common case, covered in job seekers clicking Google Ads.
• Terms that convert on one device only. "boiler replacement cost" converts on desktop and not on mobile. Don't add a negative.
• Terms that convert best on mobile. "emergency plumber", "plumber near me", "plumber open now". These are urgent searches, usually from a phone and often ending in a call. They tend to be the reason mobile is worth paying for at all.

That third group matters. If you cut mobile because its average looks bad, you also cut the urgent searches that bring your best leads. For service businesses this is a common mistake, and local business wasted spend goes into it in more detail.

Check your tracking before you blame a device

Before you cut any device, make sure you can see its conversions. Mobile is the device most likely to look worse than it is, because it converts in ways that are easy to miss. In the accounts we audit, a "bad mobile" diagnosis turns out to be a tracking gap more often than people expect.

Go through these before you change anything:

Phone calls. Someone who taps "call" on a mobile ad or on your mobile site has converted. If you don't track calls from ads and calls from the website as conversions, mobile gets the cost and none of the credit. Go back to the example. Say 8 untracked calls came from mobile last month. Mobile then has 20 conversions, not 12, and its cost per conversion falls from $150 to $90 ($1,800 ÷ 20). That's still higher than desktop's $70, but now it's a fine-tuning question rather than a reason to switch mobile off.

Cross-device journeys. People often find you on a phone at lunch and buy on a laptop that evening. Google Ads counts some of these as cross-device conversions, but only when it can link the two devices, and that doesn't always work. Expect mobile to be under-credited for research-heavy purchases like furniture, software or B2B services.

Forms that break on phones. Fill in your own form on a phone, on mobile data, without autofill. If a field is hidden behind the keyboard or the submit button sits off screen, mobile traffic is fine and your page is the problem. A negative keyword can't fix that, and neither can a bid change.

Conversion actions that only fire on desktop. A tag in a desktop-only page template, a thank-you page that mobile users skip, or a chat widget that doesn't load on small screens can all make one device look dead.

If you work through these and still find a large gap, the gap is real. If you're still unsure where the money goes, Google Ads spending money with no conversions gives the order to check things in. To put a number on the waste without inflating it, see how to calculate wasted ad spend honestly.

Device waste patterns we see most often

Most device waste follows a few patterns, and each needs a different fix. Here they are, roughly in the order we come across them.

Mobile research clicks on high-consideration terms

Searches like "kitchen extension cost", "best CRM for small business" or "solar panels worth it" are research. On a phone they're often casual browsing: someone on the sofa, or killing time on a train. The same person may buy later on desktop, but often nothing happens. These terms tend to show a big mobile cost-per-conversion gap. The fix is usually a lower mobile bid or a mobile landing page built to collect a contact rather than close a sale. A negative keyword is the wrong fix.

Accidental and low-intent mobile clicks

Small screens bring more accidental taps. This matters most on Display, Video and placements inside apps, where nobody types a search. Those campaigns report no search terms at all. Search terms only exist for Search, Shopping and Performance Max, so app and site waste is invisible to any search terms review. For Performance Max, the controls you have are covered in Performance Max placement exclusions, honestly.

Tablets: small spend, often no return

Tablet traffic is usually a small share of spend. It's often a shared family device, used for browsing more than buying. It's also easy to ignore because the amounts are small. In the example, $150 a month is $1,800 a year. Check it once a quarter with a long date range so you have enough data. If tablet still hasn't converted after spending a few times your target cost per conversion, lowering its bid or excluding it is reasonable.

Desktop waste on "quick answer" searches

It isn't always mobile. Desktop sometimes takes most of the spend on informational terms, like people at work searching "what is a heat pump" or "how does payroll software work". Those terms usually waste money on every device. Check the device split, and if no device converts, it's a negative keyword problem, not a device problem.

Broad match spreading differently by device

Broad match keywords can bring in different queries on different devices, because Google uses signals like location and context to decide what matches. If one device's spend jumped after you added broad match keywords, check the search terms for that device first. Are broad match keywords worth it in 2026? covers when that trade-off is worth it, and why Google Ads shows irrelevant search terms explains how the matching happens.

How to fix device waste, and what each fix costs you

Pick the fix to match the cause. Bid adjustments for a cost gap, exclusion for a device that never converts, a separate campaign for terms that behave very differently, page work for a broken experience, and negatives only for terms that waste on every device. Each one has a cost, so here they are with the downsides.

Device bid adjustments

On Search campaigns with manual bidding, you can raise or lower bids by device, from -100% up to +900%. In the example, with tracking fixed, mobile costs $90 per conversion and desktop $70. A mobile adjustment of around -20% is a sensible first step. Check again after a month.

The cost: Smart Bidding strategies like Target CPA, Target ROAS and Maximize Conversions mostly ignore device bid adjustments. A -100% adjustment still works as an exclusion. Anything in between has little or no effect, because the algorithm already sets bids per auction using device as one of its signals. If you use Smart Bidding and still see a big device gap, the bidding either has too little data, or you're feeding it conversions that don't reflect real value.

Excluding a device with -100%

Setting a device to -100% stops ads showing on it. It's a blunt tool, and it's right for some cases, like tablets that have never converted or a B2B campaign whose buyers only order from office computers.

The cost: you lose every conversion on that device, including the ones you couldn't see. If mobile tracking is incomplete, excluding mobile can cut real leads you never knew about. Only exclude after you've checked tracking.

A separate campaign for one device

When a group of terms behaves very differently on mobile, some advertisers run a mobile-only campaign next to a desktop-and-tablet campaign. They use -100% adjustments to split the traffic. That gives each device its own budget, its own ads, its own landing page and its own negatives.

The cost: twice the campaigns to manage, and each one gets less data. That hurts Smart Bidding most. It's worth doing for large spend and clear differences, not for a $70 against $90 gap.

Fixing the mobile page

If the gap comes from a slow page, a long form or a hidden button, fix the page. This is often the cheapest fix of all, and it's the only one that recovers the lost conversions instead of just spending less.

The cost: it isn't something you do in the Google Ads interface. It needs whoever runs the website, and that's often the hardest part.

Negative keywords, only for terms that waste everywhere

Negatives are the right fix for terms that waste money on every device, and only for those. Before you block a term because of a bad mobile number, check its desktop row. If desktop converts, a negative throws away the good with the bad. If you've already blocked something too broadly, how to undo negative keywords shows how to reverse it.

What about Performance Max and Shopping?

Performance Max gives you much less device control than Search, so device waste there is harder to see and harder to fix directly. Shopping sits in between. Search terms exist for both, but how much you can do about devices varies.

Standard Shopping accepts negative keywords and shared negative keyword lists, like Search. Device bid adjustments are also available when you bid manually. The same rules apply: segment by device, check tracking, then choose between bids and negatives. Shopping traffic on phones is often browsing, so expect a gap on expensive products, and judge it with your return on ad spend, not clicks.

Performance Max is built to decide placements, devices and audiences itself, and the device controls it offers are limited compared to Search. Check what your account shows today instead of assuming. Device segments in reporting are also less complete than in Search. What you can control is the queries. A Performance Max campaign accepts up to 10,000 campaign-level negative keywords. It doesn't accept shared negative keyword lists, so each negative is added to the campaign itself. Search terms for Performance Max appear in the same campaign_search_term_view as Search and Shopping, with the reporting gaps described in what the PMax search terms report actually shows.

In practice, for Performance Max: if one device looks wasteful, look for the terms that waste money everywhere and block those. That's the control you have. For the other places PMax spends without showing it, see where Performance Max wasted spend hides.

Where PPCleaner helps, and where it doesn't

PPCleaner handles one half of this: search terms that were never going to convert, on any device. It doesn't touch device bids. That's deliberate, and it means the device half is still yours to manage.

Every night PPCleaner reads the search terms your account paid for across Search, Shopping and Performance Max. It works out which ones were never going to convert and adds them as negative keywords. Terms like "plumber salary", "how to fix a leaking tap" and "free invoice template" get blocked whichever device they came from. That removes the background waste, which makes the device split easier to read. When the remaining gap between mobile and desktop isn't padded out with junk terms, you can tell whether it comes from the device or from the queries.

What it won't do:

• It won't set device bid adjustments. PPCleaner never touches budgets, bids, ads or keywords. If mobile needs -20%, you set it.
• It won't block a term that converted. A term that converted in the last 90 days is never blocked, even if it converted on only one device. That's the safe choice. It also means a term that converts on desktop and wastes money on mobile stays live, and the mobile side is yours to handle with bids, a page fix or a split campaign.
• It can't see Display, Video or App waste. Those campaigns report no search terms, so there's nothing to read.

If a block turns out to be wrong, you can undo it in one click for 90 days.

A device check to add to your routine

Once a month, or once a quarter for small accounts:

1. Segment campaigns by device over the last 90 days. Note cost per conversion for each device.
2. For any device with a gap of two times or more, or no conversions after spending a few times your target cost per conversion, check call tracking, cross-device conversions and the mobile form.
3. Segment your top 30 search terms by cost, split by device. Mark each term as "wastes everywhere", "wastes on one device" or "converts".
4. Negatives for the first group. Bids, page work or a split campaign for the second. Leave the third alone.

It takes about 20 minutes, and it fits into a wider review like the Google Ads audit checklist you can run in an hour.