The State of UK Google Ads Spend: Q2 2026

Key takeaways

• Over £1.1 million of live UK Google Ads spend analysed in a single quarter, our biggest sample yet, on top of 3,000+ audits since 2017.
• At least 56% of active accounts have a conversion-tracking fault serious enough to distort the numbers they optimise on. Second quarter running.
• Wasted search spend rose by almost a quarter in three months: the median account now puts 24% of its search spend into keywords that do no real work, up from 20% in Q1.
• The median account’s ads appear in fewer than one in five eligible searches: ad rank takes 54% of eligible impressions, roughly three times what capped budgets take.
• The CPC panic is overdone: like for like, clicks cost just 4% more than a year ago, and conversion-rate gains of 7% more than covered it.
• Around a third of UK Google Ads spend goes to Performance Max, whose reported return still cannot be fully verified.
• The median UK ecommerce account reports a return of only about 1.4 times spend, and that is as much a tracking problem as a performance one.

Every quarter we take an anonymised look across the live UK Google Ads accounts in our audit programme. We are not looking for reasons to spend more. We are looking for money that is already going to waste. This is the second report in the series, and the first that can say how things are moving, not just where they stand. You can read the Q1 2026 edition for the baseline.

The short version: UK advertisers held their budgets broadly flat this quarter and got more for the money, with conversion rates up and cost per action down. The loud story doing the rounds, that click prices are running away, is not what the clean data shows. The quiet story is worse. The share of search spend doing no real work rose by almost a quarter in three months, and most accounts still cannot fully trust the numbers they steer by.

Source: PPC Geeks, Q2 2026 (April to June). Figures are floors or medians. See the methodology below.

UK Google Ads spend: the headline findings

For Q2 2026 (April to June), across the live UK accounts we analysed:

• At least 56% of active accounts have a conversion-tracking fault serious enough to distort the numbers they optimise on. Unchanged from Q1, and still the floor.
• Wasted search spend rose by almost a quarter in three months. The median account now puts 24% of its search spend into keywords that do no real work, up from 20% last quarter.
• The median account’s ads show in fewer than one in five eligible searches. Ad rank takes 54% of eligible impressions, roughly three times what capped budgets take (18%).
• Like for like, cost per click rose about 4% year on year. Conversion rates rose 7%, so efficiency improved.
• Around a third of UK Google Ads spend goes to Performance Max, and its reported return still grades its own homework.
• The median ecommerce account reports a return of only about 1.4 times spend.
• We analysed over £1.1 million of live UK spend in a single quarter, our biggest sample yet, on top of the 3,000 or so audits we have run since 2017.

Finding 1: the tracking problem is not going away

Last quarter we reported that at least 56% of active accounts had a conversion-tracking fault serious enough to distort the numbers they optimise on. A sceptic could call one quarter a blip. It is now two quarters, same method, same result: 56% again. More than half of UK accounts are steering by instruments that are wrong, and most of them are letting Google’s automated bidding optimise on that same wrong data.

As before, that figure is the cautious floor. It counts only the faults we can prove from the outside. Consent settings, tag errors and untracked revenue would only push it higher.

Finding 2: wasted search spend rose by almost a quarter

Here is the finding that should sting. Across the accounts we could measure properly, the median account put around 24% of its search spend into keywords that converted nothing or cost more than three times its own average. Last quarter that figure was 20%. That is a rise of almost a quarter in a single quarter, and it was not driven by a few bad apples: nearly half of the measurable accounts, 22 of 49, now run above 30%, up from around a third in Q1.

The sharpest edge is the money that bought nothing at all. Spend on keywords that converted absolutely nothing jumped from 11% to 16% of the keyword spend we measured, a rise of more than two fifths in three months. Put plainly: around £1 in every £6 of measured search keyword spend bought no conversions whatsoever, up from roughly £1 in 9 last quarter.

The same care applies as before: some of that spend defends a brand or assists a sale that last-click reporting never credits, so we would not call all of it pure waste. Our sample also grew this quarter, from 43 measurable accounts to 49, so this is the state of the market we can see, not a like-for-like panel. But the direction matters. In a quarter when overall results improved, the share of search budget doing nothing grew. The strong keywords carried more of the load, and the dead weight got heavier. That is not a market problem. That is a housekeeping problem, and it is growth you have already paid for.

Everyone wants to blame rising click prices. In the accounts we can compare properly, clicks got 4% dearer while conversion rates rose 7%. The market is not taking your money. The wasted quarter of your own budget is.

Dan Trotter, Head of PPC, PPC Geeks

Finding 3: your budget is not the reason your ads don’t show

In Q1 we reported that the median account loses around 18% of its available search clicks to capped budgets. That is still true, and 39 of the 62 accounts with meaningful spend still lose more than one click in ten that way. But this quarter we looked at the whole picture of where eligible impressions go, and the budget story turns out to be the small half.

The median account loses around 54% of its eligible search impressions to ad rank. Google ran the auction, considered the ad, and left it out, because other advertisers’ combination of bid and ad quality beat it. Put the two together and the median account’s ads actually appeared in fewer than one in five of the searches they were eligible for. This is not a rare failure mode: 44 of the 53 measurable accounts lose more than 30% of their eligible impressions to rank, and the figure was effectively the same in Q1, so it is a pattern, not a blip. It is also the number that most often stuns people in our audits, because nothing inside the account screams about it.

Two honest notes. “Eligible” counts only the auctions Google considered the account for, and broad targeting widens that pool, so a low share partly reflects wide match settings rather than weak ads alone. And losing to rank means outranked or priced out, a mix of bid and quality, which is precisely why it is fixable: tighter targeting, better ads and smarter bids all move it. Rank takes roughly three times more of your eligible impressions than capped budgets do. If you only ever look at the “limited by budget” flag, you are watching the wrong leak.

Put Findings 2 and 3 together and the conclusion writes itself. Most accounts are wasting roughly a quarter of their search spend while simultaneously being outranked out of half their eligible impressions. The win is still moving the money you already spend, not adding to it.

Finding 4: the truth about rising CPCs

“Google Ads is getting more expensive” is the industry’s favourite headline, and on the lazy numbers it looks true. Blend our whole book together and the average cost per click rose about 18% year on year. But blended figures mix in new accounts and shifts between channels, so they measure the mix, not the market.

Compare the same accounts like for like, Q2 2026 against Q2 2025, and the picture changes completely. Cost per click rose about 4%. Conversion rates rose about 7%, and cost per action actually fell around 3%. The same advertisers spent a broadly flat amount and got more out of it. If your costs jumped far more than 4% this year, the honest first question is not “what is Google charging?” but “what changed in my account?”

Finding 5: Performance Max still grades its own homework

Around a third of all UK Google Ads spend in our sample went to Performance Max this quarter, 34 of the 78 accounts. Its share actually slipped a little, from 34% to 32%, the first dip we have measured, though one quarter does not make a retreat. On reported numbers it remains the star. Like for like it showed a return on ad spend of about 4.8, against about 3.0 for standard Search.

The asterisk from Q1 has not shrunk. Brand and Shopping cannibalisation flatter that reported return, the biggest single category of its activity is the one Google labels as unknown, and the platform cannot prove its conversions are incremental. It also rests on the same conversion tracking that is distorted in over half of accounts. A third of the nation’s Google Ads budget still flows into a channel that marks much of its own homework.

Finding 6: the ecommerce reality check

New this quarter: we looked at what ecommerce accounts actually report getting back. Across the 51 accounts tracking revenue, the median reported return was only about 1.4 times spend, and half sat below that. Set against the glossy 4-times-plus benchmarks the industry likes to quote, that looks bleak.

Read it carefully, though. Reported return is a floor, not a verdict. It ignores margin, repeat purchases and anything that happens offline, and it rests on conversion tracking that Finding 1 says is broken in most accounts. Untracked revenue makes a good account look unprofitable. If your reported return sits near that median, check the measurement before you condemn the marketing.

Finding 7: how much of your budget just buys your own name?

Also new this quarter: across the accounts where brand terms could be cleanly separated from generic search, the median account put around 9% of its search spend into its own brand name. Those tend to be the cheapest clicks in the account. In our Q1 analysis they cost a median 29% less per click than non-brand keywords.

Cheap is exactly why they need defending. When a competitor bids on your name, it is those cheap clicks that get more expensive, and most accounts only notice after the damage is in the numbers. If you have never separated brand from non-brand in your reporting, you do not actually know what your generic search costs you.

What this means for your Google Ads spend

The order of operations from Q1 survives contact with a second quarter of data, with one sharpened edge.

1. Fix the tracking first. Two quarters running, most accounts are optimising on distorted numbers. Everything else you do compounds this error until it is fixed.
2. Reclaim the wasted quarter. The dead share of search spend grew to 24%. Move that budget onto the keywords that demonstrably sell, and onto the clicks you are currently losing to capped budgets.
3. Ignore the CPC panic. Like for like, clicks cost 4% more and convert 7% better. Rising costs are rarely the real problem in a UK account. The wasted quarter is.
4. Watch Performance Max with clear eyes, and know what share of your budget is simply buying your own brand name back.

How this compares with the wider market

Our findings are first-party, pulled from real accounts rather than a survey, and the wider market data still points the same way. The IPA Bellwether report earlier this year found UK marketing budgets revised up to their highest level in nearly two years, with online channels among the biggest gainers; our own sample shows those budgets holding rather than retreating. Our waste figure also still sits inside independent estimates, which commonly land in the 15 to 40% range. If anything, our 24% median suggests the middle of that range is the honest place to plan from.

About this data

These figures come from the live UK Google Ads accounts in PPC Geeks’ audit programme, analysed anonymously and in aggregate for Q2 2026 (April to June). That is 78 enabled accounts and 62 with meaningful spend, depending on the measure, totalling just over £1.1 million in the quarter. Year-on-year comparisons are like for like across the 53 accounts present in both years. The ecommerce figure covers the 51 accounts tracking revenue; the brand-share figure covers the 16 accounts where brand terms could be cleanly classified. We name no client, and no one can identify a single account. Several of our numbers are deliberately cautious floors, and where the account data alone cannot reveal a fault, we have said so and given the conservative figure.

PPC Geeks is a UK PPC agency working across Google, Microsoft, Meta, Amazon and TikTok. If you would like an honest look at where your own account stands, tracking included, our free Google Ads audit does exactly that. No pressure to spend more. Just the truth about the money you already spend.

Frequently asked questions

Why is my search impression share so low?

Usually rank, not budget. Across the UK accounts PPC Geeks analysed in Q2 2026, the median account lost 54% of its eligible search impressions to ad rank (a mix of bid and ad quality) and only 18% to capped budgets. Check the ‘lost to rank’ and ‘lost to budget’ impression-share columns separately: they need opposite fixes.

Are Google Ads CPCs rising in the UK?

Slower than the headlines suggest. Comparing the same UK accounts like for like, cost per click rose about 4% year on year in Q2 2026, and conversion rates rose about 7%, so efficiency actually improved. Blended figures that mix in new accounts and channel shift exaggerate the rise.

How much Google Ads search spend is typically wasted?

Across the live UK accounts PPC Geeks analysed in Q2 2026, the median account put around 24% of its search spend into keywords that converted nothing or cost more than three times its own average, up from 20% in Q1. Independent estimates commonly land in the 15 to 40% range.

What is a normal ROAS for ecommerce on Google Ads?

Lower than most benchmarks admit. Half of the UK ecommerce accounts we analysed reported a return below 1.4 times spend. Treat a low reported figure as a prompt to check your conversion tracking first: untracked revenue makes good accounts look unprofitable, and most accounts under-track.

Is Performance Max’s reported ROAS reliable?

Treat it sceptically. Brand and Shopping cannibalisation flatter its reported return, Google withholds most of the query data, and the platform cannot prove the conversions are incremental. Like for like it reported 4.8 against Search’s 3.0 in our Q2 2026 sample.