Google sold 14% fewer network ads and still grew. How to read a falling volume number without fooling yourself
TAlphabet's Google Network ad impressions fell 14% year on year in Q3 2026. The number is spreading as a warning sign. It is real. It is also not new, and on its own it tells you very little.
This post is about the method more than the stock. When a volume metric drops, most readers stop at the headline. Here is how to take it apart in four steps, using Google's own filings.
Step 1: Find out what the number actually covers
The -14% is not Google's ad business. It is Google Network only: ads Google places on other publishers' sites and apps through AdSense, AdMob and Ad Manager.
In Q3 2026 Network brought in $7,455M of $82,012M in total Google ad revenue. That is 9.1% of the ad business. Math is mine.
So before reacting, ask the boring question first. What slice of the company does this metric measure, and how big is that slice?
Step 2: Put the volume next to its price
The same table in the 10-Q reports cost per impression. In Q3 2026 it rose 14%.
Fewer ads at higher prices. Network revenue was $7,455M against $7,354M a year earlier, roughly flat. A volume number read without its price is half a sentence.
Step 3: Chain three years, not one
One quarter can be noise. Three years is a pattern. Network impressions fell 14% in Q3 2024, 5% in Q3 2025 and 14% in Q3 2026.
Impressions have been negative in every quarter reported since Q1 2024. Chained, that is about 30% fewer ads sold than in Q3 2023. Math is mine, using rounded yearly rates, so treat it as approximate.
Price per impression rose every year too, and chains to about +39% over the same three years. Math is mine.
Revenue is $7,455M now against $7,669M in Q3 2023. That is a 2.8% decline.
So the honest summary is not "demand is collapsing". It is: Google has spent three years holding Network revenue roughly flat on a shrinking number of ads at rising prices.
Step 4: Read the words the company uses for the drag
Google's general explanation is generic. Impressions move with advertiser spending and "ongoing product and policy changes". Price moves with device, geographic, product and property mix, plus currency.
The useful sentences sit in the Google Network section of each 10-Q. In every filing since Q2 2025, Google names a decline in AdSense revenue as the drag, offset by AdMob. In Q3 2026 Network grew primarily on AdMob, partly offset by AdSense.
That points to two businesses under one number. AdMob is ads inside apps. AdSense is mostly ads on publishers' own websites. My read: apps are growing and the open web is shrinking. A blended -14% hides which side is moving.
Who carries the cost
For Google itself, this barely matters. Search & other revenue was $63,350M in Q3 2026, up 12.0%.
YouTube ads were $11,207M, about 1.5 times Network.
In Q3 2023 YouTube ads and Network were roughly level, at $7,952M and $7,669M. Three years later the order is clear.
Google also gains from the shift. The 10-Qs say its traffic acquisition cost rate fell partly because revenue mix moved from Network to Search. Network pays a bigger share to partners. Each dollar that moves to Search keeps more for Google.
For publishers it is a different story. Network fell from 12.9% of Google ad revenue in Q3 2023 to 9.1% now. If apps are taking more of a flat total, the web side is likely getting less. That is inference, not a disclosed number.
The case against my read
The strongest counter-case is serious. "Product and policy changes" can mean Google is cutting low-quality inventory on purpose. Fewer, better ads at higher prices would look exactly like this. Then it is a choice, not a loss.
Google also does not split AdSense and AdMob in dollars. My app-versus-web split rests on the filing's wording, not on reported figures. I hold it with moderate confidence, no more.
What would change my mind: a 10-Q that names AdSense as a source of growth again, as the Q3 2024 and Q1 2025 filings did.
A dated test
AppLovin reports Wed 4 Nov after the US close, about 05:00 SGT Thu 5 Nov. It runs an in-app ad platform. If in-app demand is the healthy half, its ad revenue growth should run far ahead of Google Network's.
If in-app ad growth slows sharply too, my split is wrong, and the -14% is closer to a broad demand signal than I think. I will say so either way.
Why this matters beyond Google
Delete the company name and the method still works. Any time a volume metric falls (units, users, shipments, impressions), run the same four steps:
1. Size the slice. What share of revenue does the metric cover?
2. Pair it with price. Volume times price is the only thing that pays.
3. Chain three years of both. One quarter is a headline, three years is a trend.
4. Read the named drag. Filings often say which product is weak in plain words, even when the dollars are not split.
A flat line made of two opposite trends is a different story from a flat line. Most of the risk, and most of the mispricing, hides in that gap.
After reading this, you can take any falling volume headline, test whether it is a demand signal or a mix shift, and name the one filing sentence that would change the answer.
Question for you: when a company stops splitting out a business you care about, do you treat the wording in the filing as evidence, or do you wait for the numbers?
Not advice. I may hold positions in the names mentioned.
Stop guessing. Start calculating.
Live t
o fight another day. 🤙
#options #tradingpsychology #riskmanagement #optionstrading
Disclaimer: Investing carries risk. This is not financial advice. The above content should not be regarded as an offer, recommendation, or solicitation on acquiring or disposing of any financial products, any associated discussions, comments, or posts by author or other users should not be considered as such either. It is solely for general information purpose only, which does not consider your own investment objectives, financial situations or needs. TTM assumes no responsibility or warranty for the accuracy and completeness of the information, investors should do their own research and may seek professional advice before investing.
