Google Paid Clicks Rose 8% in Q3 2026 While CPC Rose 4%: What the 10-Q Shows

Alphabet's Q3 2026 10-Q, filed on 28 October, says Google paid clicks grew 8% year over year and cost per click grew 4%. The fear going into the quarter was that AI Mode would make advertisers pay more for fewer clicks. Eight quarters of the same filing say the opposite. Advertisers are paying slightly more per click for many more clicks, and the price line has been the weaker of the two all year.

What Alphabet's Q3 2026 10-Q reports on paid clicks and CPC

Short answer: Google paid clicks rose 8% and cost per click 4% year over year. That is from Alphabet's 10-Q for the quarter ended 30 September 2026. Over nine months, clicks rose 11% and cost per click 4%. Advertisers are not paying more for fewer clicks. They are paying slightly more for more clicks, and growth in both has slowed.

Alphabet publishes two monetization metrics for Google Search and other properties in every quarterly filing. Paid clicks count clicks on ads across Google search, Gmail, Google Maps and Google Play. Cost per click is click-driven revenue divided by those clicks.

The filing does not publish the level of either metric. It publishes only the percentage change against the same period a year earlier. That makes the series useful for direction and awkward for anything else, which is why most coverage quotes one quarter and moves on.

The exact Q3 2026 table

The Monetization Metrics table shows paid clicks up 8% for the quarter and 11% for nine months. Cost per click rose 4% on both measures. On the Google Network side, impressions fell 14% and cost per impression rose 14%.

The sentence Alphabet places under that table has not changed in two years. It attributes click changes to advertiser spending, product and policy changes, and "fluctuations in search queries resulting from changes in user adoption and usage, primarily on mobile devices." It names no AI product. Neither does the MD&A paragraph explaining Search revenue growth.

Google Search and other revenue was $63,350 million in Q3 2026, against $56,567 million a year earlier. That is 12.0% growth. Multiply 1.08 by 1.04 and you get 12.3%, close enough to confirm the two metrics carry the revenue line.

The quarter on quarter view is the weaker one. Revenue was $63,271 million in Q2 2026, so Q3 added $79 million. A year earlier the same step added $2,377 million. That flat sequential quarter is the number worth explaining, and the click metrics explain only part of it.

Eight quarters of Google paid clicks and cost per click

A single quarter tells you very little about Google search revenue under AI. The useful view is the sequence, filed on the same definitions, across the period when AI Overviews ads scaled and AI Mode launched.

The table below covers eight quarters, Q4 2024 to Q3 2026. Two of the eight are blank on purpose. Alphabet reports quarterly click metrics only in its 10-Qs, and its 10-K gives only the full-year change. No Q4 click figure exists in any Alphabet filing.

Click and price changes are as filed in each 10-Q. Q4 revenue is derived by subtracting the three 10-Q quarters from the 10-K annual figure: $224,532M less $161,459M for 2025, and $198,084M less $144,050M for 2024. Revenue growth is our calculation from the filed dollar figures.

The Q4 hole, and a rough way to fill it

The FY2025 10-K reports paid clicks up 6% and cost per click up 7% for the full year. If you weight the four quarters equally, the clicks figure implies roughly 11% in Q4 2025. The working is 4 times 6, less 2, 4 and 7. The same method puts Q4 cost per click near 6%.

Treat both as plus or minus 2 points. Quarters are not equal in size, and Q4 is the largest advertising quarter of the year. The check that makes the estimate usable is revenue. Derived Q4 2025 Search revenue grew 16.7%, and 1.11 times 1.06 gives 17.7%, so the estimate sits in the right range.

Google search revenue under AI is a volume story, not a price story

The popular reading of AI search goes like this. AI answers absorb the easy queries, fewer ads get clicked, and Google raises the price of the clicks that remain. That reading predicts falling paid clicks and rising cost per click. The filings show rising clicks and a cost per click that has slowed.

I think the popular reading is simply wrong for Google's paid inventory, at least through Q3 2026. It may be right for organic traffic to publishers. Those are different markets and they should stop being discussed as one.

How much of Search growth came from clicks

You can split each quarter's growth into a click part and a price part. Take the natural log of 1 plus each rate and divide the click log by the sum. It is a standard decomposition, and the result is our calculation, not Alphabet's.

In Q1 2025, clicks supplied about 23% of Search growth and price the other 77%. By Q2 2026 clicks supplied about 80%. In Q3 2026 it was about 68%, still volume-led even as both lines slowed. The shape of that shift matters more than any single quarter.

One test separates a real decomposition from a coincidence. If clicks and price really drive the line, their product should land near the reported revenue growth. It does, within a point, in every disclosed quarter.

Gaps come from rounding, since Alphabet publishes whole percentages, and from Search and other revenue that is not click-driven. None exceeds one point, which is why we treat the click and price split as reliable for direction.

Where AI Mode sits on the paid clicks timeline

The brief for this analysis was to mark the AI Mode dates on the chart and see whether the lines bend at them. Two dates are firm enough to use. A third, the one most people mean, is not.

May 2025: AI Mode reaches US users and ads go into test

Google rolled AI Mode out to US users in May 2025. On 21 May 2025 TechCrunch reported that US users would see Search and Shopping ads in AI Mode as part of a test, on desktop and mobile. The same piece said ads in AI Overviews would expand to desktop in the US.

That lands in Q2 2025. Paid click growth went from 4% that quarter to 7% the next, then into double digits. Cost per click went from 8% to 7%, then lower. The timing is consistent with AI surfaces adding paid clicks. It does not prove they did, because query growth rose across Google at the same time.

Q3 2026: the Search button change is partial, not a clean break

The idea that AI Mode "took over the search button" in Q3 rests on reporting that is real but narrow. TIME reported on 29 September 2026 that Google had placed an AI Mode button on the Search homepage "for many users." It gave no rollout date and no country list.

So Q3 2026 is not the first full quarter of an AI Mode default. At most it is a quarter in which some users saw the change for some weeks. Anyone who tells you the Q3 numbers measure the button swap is overreading a partial exposure. Q4 2026 is the first quarter that could, and Alphabet will not report its click data until the 10-K.

Why Google search revenue growth slowed in Q3 2026

Search and other growth fell from 16.8% in Q2 to 12.0% in Q3. Paid click growth fell from 13% to 8% over the same step. Alphabet's own explanation names three causes, and only one of them is AI.

What management said on the call

On the 28 October 2026 call, chief business officer Philipp Schindler called foreign exchange "the single largest driver of Search deceleration relative to Q2". He also cited retail advertisers in APAC hit by the EU ending its de minimis exemption, and media spend normalising after a strong Q2 driven by the FIFA World Cup.

Then came the AI line. Schindler said "the strong growth in AI Mode queries had an impact because of the lower monetization we were seeing in AI Mode". In plain terms, AI Mode queries grew faster than Google can currently earn from them. Chief executive Sundar Pichai said overall query growth was "the highest it's been in the last 5 years."

Chief financial officer Anat Ashkenazi added the warning for Q4. Google will "continue to lap the search performance acceleration that began in the third quarter last year, which will pressure revenue growth in Q4". The tougher comparison is part of what the 8% click figure is absorbing.

Cost per click is measured in US dollars. When the dollar strengthens, the same local-currency bid converts into fewer dollars per click. Alphabet lists exchange rates among the drivers of cost per click in every filing.

That means part of the low 2026 cost per click reading is a currency effect, not an auction effect. Ashkenazi described FX as a slight year over year headwind with a larger sequential impact. A US-only advertiser should read the 3% to 5% price growth as a floor, not a measurement of their own market.

What the 10-Q cannot tell you about AI Mode advertising

This is where the argument above is weakest, and it is weaker than the clean chart suggests.

The metric does not separate AI surfaces

Paid clicks blend classic search ads, Shopping ads, AI Overviews ads, AI Mode ads, Gmail, Maps and Play. Alphabet discloses no split. A rise in total clicks could hide falling clicks on classic results offset by growth elsewhere. The filing cannot rule that out, and neither can we.

Alphabet also states that it periodically refines how it counts paid clicks. Nothing in the filings flags a change in the period charted. Still, a definitional change inside a blended metric is exactly the sort of thing a percentage series would hide.

The best counter-argument

The strongest reading against this post is simple. If AI Mode monetizes less per query, as Schindler said, then cost per click growth of 3% to 5% may already be the early sign of price dilution. On that view the click boom is advertisers buying cheaper AI inventory, and the price line will keep sliding as AI Mode grows.

That view fits the data as well as mine does. The 10-Q gives a direction, not a mechanism. What would settle it is a quarter where AI Mode exposure jumps and cost per click growth turns negative while clicks hold. We have not seen that quarter yet.

What this means for your paid search budget

If you run B2B paid search, the Google-wide numbers are a backdrop, not a forecast for your account. Your cost per click depends on a few dozen competitors bidding on a few hundred terms. Alphabet's figure averages every auction on earth, in dollars, across consumer and business queries.

Three things do carry over. Click supply on Google is still growing. Price inflation across the system has slowed. And new ad placements inside AI answers are coming whether you opted in or not. Schindler announced on the same call that ads in AI Mode are expanding to international markets, alongside multi-turn ads in AI Mode in the US.

Watch your own click mix, not your average CPC

Your average cost per click can stay flat while the clicks underneath it change shape. Check how many of your conversions now come from placements reported under AI Overviews or AI Mode, and whether those clicks convert at the rate of classic results. That split is the one Alphabet will not publish for you.

The organic picture is a separate problem. Field evidence that AI Mode cut external clicks sharply in a controlled test is about publisher traffic. The paid search click-through fall measured on AI Overview queries is about one surface. Neither contradicts the 10-Q, because the 10-Q counts all paid clicks everywhere.

If you are deciding how much automation to hand to Google as these placements grow, the trade-offs are in the breakdown of controls inside AI Max for Search. For a comparison of where else AI answer engines now sell ads, see the side by side of AI search ad products.

Frequently asked questions

Where to start

Pull the last eight quarters of your own Google Ads data and put clicks and cost per click side by side, exactly as in the first table above. If your account shows the reverse of Google's pattern, rising prices on flat volume, the problem is your auctions and not AI Mode.

Then put a date in your calendar for early February 2027. The FY2026 10-K will give the full-year click change, and with this post's three quarters you can back out the first Q4 that sat after the Search button change. That is the number the Q3 filing could not give you.

Related reading