PPCleaner

Google Ads Accidental Clicks on Mobile and What to Cut

Published November 8, 202612 min read
A business owner and marketer reviewing an advertising decision in their workplace.
The short answer

Most accidental clicks on mobile come from placements, not search terms: app inventory, in-feed units and the non-search parts of Performance Max. Negative keywords can't reach those clicks. The tools that can are placement exclusions, content exclusions, device adjustments and campaign structure.

On Search, a real fat-thumb click is rarer than it looks. Most one-second bounces from search ads come from the query: someone looking for another brand's login page, a phone number or a quick answer. To tell the two apart, compare the same search term across devices. If it fails everywhere, block the term. If it only fails on mobile, the problem is the click or the page, and blocking the term would throw away good traffic.

Accidental clicks and irrelevant queries are two different leaks

An accidental click is a tap the person never meant to make. An irrelevant query is a deliberate click from someone who wanted something you don't sell. Both show up as a click, a bounce and no conversion, so most accounts treat them as one problem. They aren't, and the fix for one does nothing for the other.

Here is the difference in practice. Someone searches "plumber salary", sees your ad for emergency plumbing and taps it on purpose, hoping for salary data. That's query waste. A negative keyword on "salary" stops it. Someone else is playing a puzzle game, an ad slides in at the bottom of the screen just as their thumb comes down, and the browser opens your page. There's no search term behind that click. No negative keyword in the world will stop it.

You need to know which leak you have before you act on it. If you treat accidental clicks as query waste, you end up blocking terms that convert on desktop. If you treat query waste as accidental, you lower mobile bids and keep paying for "plumber salary" at a lower price. We see both mistakes often in the accounts we audit.

Where accidental mobile clicks actually come from

Most accidental clicks come from inventory where the ad sits next to something the user is touching: games, utility apps, news feeds and video players. Search results pages produce fewer of them, because the person is already reading results and choosing one.

The usual sources are:

  • Mobile games and free utility apps. Ads appear between levels or next to buttons the player taps constantly. A close button that's small or slow to appear invites mistaken taps.
  • In-feed and in-article units. The ad moves into place as the page scrolls. The user taps where text used to be.
  • Video and Display inventory reached through Performance Max. You didn't pick these placements, but your budget can reach them through Performance Max.

Display, Video and App campaigns report no search terms at all. Performance Max reports search terms for its search inventory only. The clicks it buys on other inventory leave nothing in the search terms data for you to block. That is one reason the cost in your search terms view rarely adds up to the campaign cost, which we cover in Search Terms Report Missing Data and Where the Cost Went.

Clicks on Search that look accidental

On mobile Search, the ad often fills most of the first screen. Someone who wants the organic result below it taps the first thing they see. Autocomplete adds to this: the user picks a suggestion that's close to what they wanted, gets a results page they didn't intend, and taps through before noticing. These clicks are real, but they aren't random. They cluster around certain kinds of query, so you can find them in the search terms report.

Why the advice that already ranks is incomplete

Most of what ranks for this topic is written for publishers or by Google. It explains that Google penalises layouts built to cause accidental clicks and that some of those clicks get filtered as invalid. That's true, but it doesn't help you much as the advertiser.

Three gaps matter:

  1. Google's protections target deceptive layouts, not normal ones. A game with a policy-compliant ad unit still gets mistaken taps. Those clicks are valid by Google's rules, and you pay for them.
  2. Invalid-click filtering is a black box. You see an invalid clicks column and occasional credits. You can't see which placements or terms were filtered, or why. Google Ads Invalid Clicks Refund: What You Actually Get explains what those credits do and don't cover.
  3. None of it covers intent mistakes on Search. Someone who taps your ad looking for a competitor's login page made a deliberate click. No invalid-click system will refund it. Only your own exclusions will stop it.

The common advice is "exclude mobile apps and check your invalid clicks." That's a reasonable start for Display. It's incomplete because it doesn't tell you how much of your mobile bounce problem is placements and how much is queries. In Search-heavy accounts, queries are often the larger share.

How to tell an accidental click from a bad query

Compare the same search term on mobile and desktop. A bad query fails on every device. An accidental click or a mobile page problem fails only on mobile. This one comparison settles most cases.

In the Google Ads interface, open the search terms report and segment by device. Sort by mobile cost. For each expensive term, put mobile and desktop results side by side.

Here's an illustrative example from a plumbing account with $1,500 of search spend in one month:

Search termDeviceClicksCostConversions
emergency plumberDesktop40$3206
emergency plumberMobile90$7203
plumber salaryDesktop25$1000
plumber salaryMobile60$2400
chase plumbing loginMobile30$1200

Total: 245 clicks, $1,500, 9 conversions.

Read it row by row:

  • "emergency plumber" converts on desktop at $320 / 6 ≈ $53 per conversion. On mobile it's $720 / 3 = $240 per conversion. The term is fine. Something about the mobile click or the mobile page is broken. Blocking the term would remove your best query.
  • "plumber salary" fails on both devices. That's $340 of query waste, and the negative keyword is obvious.
  • "chase plumbing login" only appears on mobile and never converts. Someone is looking for a specific company's customer portal. It's a deliberate click from the wrong person: query waste that only looks accidental. Block it.

So in this example, $460 is query waste you can remove with negatives. The other $720 of mobile spend on "emergency plumber" needs a different fix.

Add engagement data

Conversions alone are slow to show a pattern on low-volume terms. Engagement shows it faster. In GA4, a session isn't counted as engaged by default unless it lasts at least 10 seconds, records a conversion event, or includes two or more page views. A term or placement whose mobile sessions are almost never engaged is your strongest sign of one-second bounces.

Two cautions. First, joining GA4 sessions back to individual search terms is never complete. Expect partial coverage and treat the result as a direction, not a precise figure. Second, a slow mobile page produces the same zero-engagement pattern as an accidental tap. If the page takes several seconds to load, many people leave before GA4 records a session at all. Before you blame the click, open the landing page on a real phone over mobile data. Google Ads Low Quality Traffic: Bad Clicks or Bad Pages? goes through that check in detail.

Search terms that correlate with one-second bounces

On mobile Search, the terms that most often produce instant bounces are navigational, answer-seeking or mistyped. The person wanted to go somewhere specific or get a quick fact, and your page was in the way. In the accounts we audit, these patterns come up again and again.

  • "[other brand] login", "[other brand] account", "[other brand] app"
  • "[other brand] customer service number", "[other brand] phone"
  • "track my order", "[carrier] tracking"
  • Bare brand names of large companies that share a word with your keywords: an accounting firm bidding on "chase" variants, a pool company appearing for "aqua" brand names.

These people want a specific destination. When your ad appears first on a small screen, some of them tap it before they've read it. The exit is immediate. These terms are usually safe to block as phrase match, because the modifier ("login", "tracking") carries the meaning. See Phrase Match Negative Keyword Examples You Can Copy Safely for how to write them so they don't catch your own brand.

Quick-answer queries

  • "how much does a plumber charge per hour"
  • "is [product] worth it"
  • "[service] hours", "open now"

These are trickier. "Open now" from a mobile user can be a very strong buying signal for a locksmith or a pharmacy. For a B2B software company it's almost always noise. Decide by business, not by pattern. The same is true of cost questions: "how much does a plumber charge" can come from a buyer comparing prices.

Thumb typos and close variants

Mobile typing produces misspellings, and close variants match your keywords to many of them. Most are harmless: "plumer near me" is the same intent as "plumber near me". The ones that hurt are typos that land on a different word altogether: a "roofer" keyword matched to "rover", or "boot repair" matched to "boat repair". These look like accidental clicks in the data but come from matching. Close Variants in Google Ads and Why Exact Isn't Exact explains why exact match lets them through.

Very short queries

One-word and two-word queries on broad match often bounce on mobile, because the user hadn't finished their thought. They tapped a suggestion or an ad before the search was really formed. Don't block short queries as a category. Block the specific ones that fail across devices.

Placements and formats that correlate with one-second bounces

Where there's no search term, the bounce shows up against a placement, a format or a channel. You have to cut it with the tools that work at that level. In practice this means Display, Video, app inventory and the non-search side of Performance Max.

What to look for:

  • App placements with high click-through rates and near-zero engagement. A click-through rate far above your other placements, with no engagement, is the classic accidental-click signature. People don't find your ad that compelling inside a puzzle game.
  • Clicks at odd hours from one placement. A steady stream of clicks at 3 a.m. from a single app is either kids playing with a phone or something worse.
  • Mobile-only conversion gaps in campaigns with no search terms. If a Display campaign converts on desktop and never on mobile, look at the placement report before you touch the creative.

For Display and Video, the placement report shows where your ads ran, and you can exclude placements and app categories. For Performance Max, your visibility is narrower. The campaign shows search-side terms (and only partly — see Performance Max Search Term Insights Are Not Queries). The rest of its inventory reports far less detail. Account-level placement exclusions apply to Performance Max, but you're excluding with less information than you have in a Display campaign.

Be honest with yourself about the limit here. You can't run a device-by-placement-by-term analysis on Performance Max the way you can on Search. If accidental mobile clicks on non-search inventory are a big share of your spend, the real fix may be structural: move the search budget you care about into Search campaigns, where you can see and control it.

What to do about each, and what it costs

Each kind of waste has its own fix, and every fix has a cost. Here they are in rough order of how safe they are.

Query waste: negative keywords

For terms that fail on every device, such as "plumber salary", "[brand] login" and "[carrier] tracking", add negative keywords. Use phrase match for modifiers and exact match for single bad queries you don't want to generalise. Broad negatives behave unpredictably across match types, so use them rarely and on purpose.

Where they go matters:

  • Search and Standard Shopping accept negative keywords and shared negative keyword lists.
  • Performance Max accepts up to 10,000 campaign-level negative keywords. It does not accept shared negative keyword lists, so a list you maintain for Search won't protect your Performance Max campaigns unless you add the terms to each campaign.

The cost: every negative can block a future good query. Before blocking a term, check that it hasn't converted, and check it across devices. If it works on desktop, it's not a bad term. Add Negative Keywords From Search Terms Without Guessing covers that check.

Mobile-only failure on good terms: fix the page first

When a term converts on desktop and fails on mobile, start with the landing page. Check load time on mobile data, whether the phone number can be tapped, whether the form fits the screen, and whether a cookie banner or popup covers the first screen. A page that takes too long to load will look exactly like an accidental click in your data.

The cost is development time. It's still cheaper than the alternative, which is paying full price for mobile clicks that can't convert.

Mobile-only failure that the page doesn't explain: device adjustments

If the page is fine and mobile still fails, lower your mobile bids where you can. With Smart Bidding, device signals are already part of the bid, and manual device adjustments are mostly ignored. The exception is −100%, which excludes the device in Search campaigns.

The cost here is real. Mobile is often where the cheapest good clicks are, especially for local services. Cutting mobile to save on accidental clicks can lose you the "emergency plumber" call at 11 p.m. Our article on Google Ads Wasted Spend by Device: Where It Hides shows how to size that trade-off before you make it.

Placement waste: exclusions and structure

For Display and Video, exclude the placements and app categories that show the accidental-click signature. For Performance Max, use account-level placement exclusions and content exclusions where they're available, and accept that you'll see less than you'd like.

The cost: exclusions narrow your reach, and on Display, app inventory is often cheap. Exclude based on engagement and conversion data, not on a dislike of apps in general.

Why bother with small amounts

A few dollars a day of accidental mobile clicks looks harmless. The bigger cost is what those clicks teach your bidding. Smart Bidding learns from every click. A pool of mobile clicks that never convert tells it that a segment is worth less, or teaches it the wrong thing about which signals predict a sale. Smart Bidding Bad Data: How Junk Queries Skew Your Bids explains how that distortion builds up.

Where PPCleaner helps and where it doesn't

PPCleaner works on the query side of this problem. It doesn't touch the placement side. Every night it reads the search terms your account paid for, from Search, Shopping and Performance Max, and adds negative keywords for terms that were never going to convert: the "[brand] login", "[carrier] tracking" and "plumber salary" type of query that makes up much of mobile bounce waste on Search.

It never blocks a term that converted in the last 90 days. That matters here, because "emergency plumber" with three mobile conversions stays untouched even though its mobile cost per conversion looks poor. That's a page or device question, not a query question. Every block can be undone in one click for 90 days. PPCleaner doesn't change your budgets, bids, ads or keywords, so it won't adjust device bids for you either.

What it can't do:

  • It can't see accidental clicks with no search term. Display, Video and App campaigns report no search terms. Neither does Performance Max inventory outside search. Those clicks never reach PPCleaner, and placement exclusions remain your job.
  • It can't fix a slow mobile page. If good terms fail only on mobile, the answer is on your site.
  • It can't recover clicks Google already counted as valid. Mistaken taps on compliant layouts are billed, and no negative keyword changes that after the fact.

A sensible split: let negatives handle the queries every night, and set aside time each month for the placement report, the device comparison and a mobile page check. The first job is repetitive and runs on volume. The second needs judgment that no tool can make for you.

mobileaccidental clickssearch termsplacementswasted spend

Frequently asked

Does Google refund accidental clicks on mobile?

Only if Google classifies them as invalid. Its systems target deceptive layouts and suspicious patterns, and you can see the totals in the invalid clicks column. A mistaken tap on a normal, policy-compliant ad unit counts as a valid click, and you pay for it.

Can negative keywords stop accidental clicks?

Only accidental-looking clicks that have a search term behind them, such as navigational queries for other brands or mistyped queries matched by close variants. Clicks from app placements, Display, Video and the non-search inventory of Performance Max have no search term, so negative keywords can't reach them. Use placement and content exclusions for those.

Should I just lower mobile bids to cut accidental clicks?

Usually not as a first step. First check whether the failing terms also fail on desktop. If they do, the fix is a negative keyword. Next, check the mobile landing page. With Smart Bidding, manual device adjustments other than -100% are mostly ignored, and cutting mobile often loses cheap, high-intent clicks such as late-night emergency service calls.

How do I know a one-second bounce was accidental and not a slow page?

From the click data alone, you often can't. Both look like zero engagement. Load the landing page on a real phone over mobile data. If it's slow or the first screen is covered by a popup, fix that first. If the page is fast and the bounces cluster in particular app placements with unusually high click-through rates, the clicks are likely accidental.

Does PPCleaner block search terms that only perform badly on mobile?

It never blocks a term that converted in the last 90 days, so a term that converts on desktop but struggles on mobile stays live. That's the right outcome, because the problem is the device or the page, not the query. PPCleaner also doesn't change device bids or placements.

About the author

Alain Abulafya

Founder of PPCleaner · Google Ads consultant & media buyer

Alain has worked in digital marketing for more than 15 years, over 10 of them managing and optimizing Google Ads campaigns. He has worked with more than 100 advertisers in e-commerce and lead generation, and he is a Google Partner. He built PPCleaner to automate the search term reviews he used to do by hand for his clients.

Stop the leak

Clean your search terms. Save time and money.

Every night, PPCleaner reviews the searches your campaigns actually paid for, groups them by intent, and excludes the ones that never convert. No spend threshold to wait for, nothing touched but negatives — and one click undoes any of it.

Start 7 days free

Search, Shopping and Performance Max · Cancel in one click

Related articles

Google Ads Accidental Clicks on Mobile: Find and Cut Them — PPCleaner