Alphabet, Google’s parent company, reported its results for the third quarter of 2026 (Q3, July to September) on 28 October, after the US market closed. Revenue growth was 25%, and revenue reached $127.6 billion [1]. That made it the 13th quarter in a row with growth above 10%. Google Cloud revenue doubled.

Even so, the stock fell 12.6% the next day. It closed at $283.82, against $324.89 a day earlier [2]. The value of all Alphabet shares fell by about $505 billion in a single day. The market did not punish Cloud, but Search, which brings in almost half of revenue. Search revenue growth was 12%, after 16.8% in Q2. Among the reasons, management also admitted that the new way of searching with artificial intelligence (AI Mode) currently brings in less advertising money than classic search. At the same time, Google spent more on equipment than it generated in cash, and it is issuing new shares.

For a long-term investor, a drop of this size is a signal to check whether anything has changed in the reasons for owning the company. The analysis therefore looks for the answer to one question: is the drop a sign that artificial intelligence is eating into Google Search, or just a transitional period in which Google invests heavily and is still waiting for the payoff?

A number in square brackets, such as [1], shows a chart when you rest the mouse on it; a click or tap opens a larger one. All charts are also collected at the end of the article.

What Alphabet said

First, what analysts expected before the results and what Alphabet delivered in Q3 2026:

Q3 2026Analysts’ expectationsResult
Revenue$128.22bn$127.58bn
Search revenue$64.9bn$63.35bn
YouTube ads revenue$11.4bn$11.21bn
Earnings per share excluding investments$3.02$2.67
Free cash flow−$2.7bn−$9.07bn

Earnings per share is net profit after all costs and taxes, divided across all shares. Under accounting rules it was $2.12. It includes the fall in the value of Google’s investment in SpaceX, which reduced it by $0.55. Analysts expected earnings excluding such effects: at $2.67, that figure was also below the expected $3.02.

What revenue is made of and how fast each part is growing:

RevenueQ3 2025Q3 2026Growth
Search$56.57bn$63.35bn+12.0%
YouTube ads$10.26bn$11.21bn+9.2%
Subscriptions, platforms, devices$12.87bn$14.29bn+11.0%
Ads on other sites (Google Network)$7.35bn$7.46bn+1.4%
Google Cloud$15.16bn$30.71bn+103%
Total revenue$102.35bn$127.58bn+24.7%

Cloud contributed 62% of all revenue growth [3]. Without it, revenue growth for what Alphabet groups under the name Google Services (Search, YouTube, subscriptions, devices) was 10.6%.

Alphabet does not give a numerical revenue forecast for Q4 2026. Chief financial officer Anat Ashkenazi said growth would be a few points lower because of a high base last year, and that currencies would weigh somewhat more, mainly on Search and YouTube. She also flagged pressure on margins, because Google is starting to rent computing power from other providers. Capital expenditure will stay at $195 to $205 billion this year [4], and in 2027 it will, in her words, “increase significantly”.

Why the stock fell

There are four reasons. The first is the most important, because it concerns the business most investors own Alphabet for.

1. Search slowed, and it shows in the number of ad clicks. Search revenue is earned when a user clicks an ad. So it can be split into two parts: how many paid clicks there were, and how much an advertiser paid per click.

The Search slowdown is in the number of paid clicks Growth on the same quarter a year earlier; currencies change the dollar price of a click, not the numberPaid clicksCost per click 0%5%10%15%20%7%7%Q3 2025Search growth 14.5%13%5%Q1 2026Search growth 19.1%13%3%Q2 2026Search growth 16.8%8%4%Q3 2026Search growth 12.0%

All growth rates in the chart compare a quarter with the same quarter a year earlier. Growth in paid clicks was the same in Q1 and Q2 2026, at 13%. In Q3 it fell to 8%. Growth in cost per click was 4% in Q3, one percentage point higher than in Q2. So the slowdown is mainly in the number of clicks.

Management gave four reasons. The first was currencies, which chief business officer Philipp Schindler called the largest. The second was the EU removing a tax exemption for small parcels, which led some Asian online retailers to advertise less. The third was the football World Cup. There were more matches in June than in July, so media and entertainment companies advertised more in June, that is, in Q2. Q2 was therefore unusually strong, and in Q3 their advertising returned to normal. The fourth reason is that the new way of searching with artificial intelligence (AI Mode) currently brings in less money.

These explanations can be partly checked. Currencies change the dollar price of a click, not the number of clicks. For total revenue, the currency effect was +$0.74 billion in Q2 and −$0.70 billion in Q3, which explains part of the slowdown. But because it is mainly growth in the number of clicks that fell, currencies are not the whole story. Alphabet does not say how much each of the other reasons contributed.

At the same time, chief executive Sundar Pichai says the number of searches is growing at its fastest rate in five years. People are using Google more than before, while revenue is growing more slowly than in previous quarters. Earnings per search are therefore growing more slowly than before. That is the heart of the concern.

2. Free cash flow is negative. Free cash flow is the cash left after a company subtracts capital expenditure from its operating cash flow. In Q3 Google spent $51.4 billion on equipment, more than twice as much as in Q3 2025 [4]. It generated $42.3 billion from operations. Free cash flow was −$9.1 billion, against analysts’ expectations of about −$2.7 billion [5]. Part of the gap is a one-off: in July Google paid a $5.2 billion European Commission fine dating from 2018. Inventory, mainly TPU chips it will sell to customers, rose by $3.2 billion in Q3.

3. Earnings per share missed expectations even excluding SpaceX. The gap between $2.67 and $3.02 came from operations, not from investments.

4. The outlook: more investment and lower margins. Investment will rise significantly in 2027, renting outside computing power will weigh on margins, and currencies and a high base will hold back Search growth in Q4 as well.

Cloud is not among the reasons for the drop. The question is whether Search is in permanent decline or in transition. The answer depends on how quickly Google makes ads in AI-powered search as profitable as they are in classic search.

Where growth comes from

Alphabet has been through a full cycle in recent years [1]. Revenue growth was 9.8% in 2022 and 8.7% in 2023. Then it picked up: it was 13.9% in 2024 and 15.1% in 2025. In Q3 2026 it reached 25%. The acceleration did not come from advertising, but from Cloud.

In Q3 2024 Cloud brought in 12.9% of revenue, in Q3 2026 already 24.1% [3]. Google Cloud sells computing power and artificial intelligence services to businesses, and since this year also systems built on its own TPU chips (a chip Google develops for artificial intelligence as an alternative to Nvidia’s). According to the company, sales of these systems were a “meaningful contributor” to growth, but it does not report them separately.

Customers have committed to far more than Cloud now earns in a year. The backlog (the value of signed contracts not yet recognised as revenue) stands at $560.8 billion in Cloud. A year earlier the total backlog was $157.7 billion, but since this year Alphabet also counts contracts shorter than one year, so the comparison is not exactly like for like. Just over 55% of the backlog is expected to become revenue within two years. That is about $154 billion a year on average. Cloud currently brings in about $123 billion a year (Q3 revenue multiplied by four).

The link with Anthropic. One of the large buyers of computing power is Anthropic, the developer of the Claude model. In its filing to go public (S-1, 26 October 2026) it states that it has committed to spend at least about $114.5 billion with Google between April 2026 and July 2033, regardless of actual usage. For additional TPU computing power between 2028 and 2031 it estimates another $125 to $150 billion. Google invested $10 billion in Anthropic in May 2026 and can invest up to $20 billion more by the end of 2030. Its stake is capped at 14.99%. Anthropic’s commitment is about one fifth of the size of the Cloud backlog. Google does not disclose how much of this commitment is already in the backlog. For Google, Anthropic is therefore a Cloud customer, a chip customer and an investment at the same time, and a competitor in models.

Why profit is growing and what holds it back

Operating income is the profit from the business itself, before the company pays interest and taxes. In Q3 2026 it reached $40.6 billion [6]. Growth on Q3 2025 was 29.9%. The operating margin was 31.8%: of every $100 of revenue, $31.80 remained. The breakdown shows where it came from:

Operating incomeQ3 2025Q3 2026Growth
Google Services$33.53bn$39.16bn+16.8%
Google Services excluding the Q3 2025 fine$37.03bn$39.16bn+5.8%
Google Cloud$3.59bn$11.12bn3.1 times
Shared AI research and other−$4.47bn−$7.66bn+71% in costs
Waymo and other bets (Other Bets)−$1.43bn−$2.05bn+44% in losses
Total$31.23bn$40.58bn+29.9%

Google Services profit in Q3 2025 was reduced by a one-off European Commission fine of $3.5 billion. Excluding it, operating income growth in the core business was only 5.8%, while its revenue growth was 10.6%. The Google Services operating margin was 40.7%, against 42.5% in Q3 2025 excluding the fine. According to the company, profit was also held back by higher payments to partners that build Google Search into their devices and browsers; these payments grow with revenue.

Profit growth is now carried by Cloud [7]. Its operating margin rose from 23.7% to 36.2%. On the other side are costs that Alphabet does not allocate to its businesses: mainly shared artificial intelligence research, the development of the Gemini models. In Q3 they reached $7.7 billion, and their growth was 71%.

The pressure point is depreciation: the cost of equipment does not appear in the income statement when it is bought, but is spread over the years it is used. In Q3 it reached $8.4 billion, and its growth was 49%. Because Google is investing about $200 billion this year, this cost will keep rising for years, regardless of how fast revenue grows.

How much cash remains

In 2025 Google was left with $73.3 billion of free cash flow. Over the last four quarters (Q4 2025 to Q3 2026), only $19.7 billion [5]. In Q3 2026 capital expenditure reached 40% of revenue, against 15% in Q3 2024 [4]. According to the chief financial officer, about 65% of the investment goes to servers and network equipment, the rest to data centres.

Because operating cash flow is not enough, Alphabet has changed how it finances itself. In 2025 it spent $45.7 billion on share buybacks, this year nothing [8]. In a buyback, a company buys its own shares on the stock market and cancels them, so every remaining shareholder owns a larger part of the company. This year Alphabet is doing the opposite. From Q1 to Q3 2026 it raised money in three ways:

  • New common shares for $42.9 billion. In June it sold shares in a public offering for $20.5 billion, and at the same time sold $10.0 billion more to a Berkshire Hathaway company. In Q3 it began selling shares gradually on the stock market at the market price (an ATM, “at the market”, programme), raising $12.4 billion. According to the company, the money from the ATM programme is mainly intended for taxes that arise on employee share awards. Because there are more shares, every existing shareholder owns a slightly smaller part of the company and of its profit.
  • Preferred shares for $19.1 billion (in June). A preferred share is something between a bond and a common share. The holder receives a fixed dividend, here 6.25% a year, and except in rare cases has no voting rights. Around 15 May 2029 these shares will automatically convert into common shares, so they too will reduce existing shareholders’ stake.
  • Debt of $85.2 billion, while it repaid $8.9 billion over the same period. Debt does not dilute shareholders, but it has to be repaid with interest, which reduces profit.

As a result, there are 1.7% more shares than at the end of 2025 [8]. The dividend on common shares remains $0.22 per share per quarter. For equipment, energy and content, Google has signed purchase commitments worth $924.4 billion. Of that, it will have to pay $213.4 billion over the next year.

One-off items distort the figures:

One-off itemAmountWhat it affects
European Commission fine (Q3 2025)charge of $3.5bnGoogle Services profit in Q3 2025
Payment of the 2018 fine (Q3 2026)$5.2bnfree cash flow in Q3 2026
Fall in the value of the SpaceX investment (Q3 2026)$8.7bn, $0.55 per sharenet profit
Gains on investments (Q1 and Q2 2026)other income of $135.7bnnet profit

Because of the investments in SpaceX and Anthropic, net profit is a poor measure of the business this year [9]. Operating income is a better guide.

How much does capital earn? Return on invested capital (ROIC) shows how much after-tax profit every dollar invested in the business by owners and lenders generates. Over the last four quarters it was about 32% (Engineering Investor calculation). Google is now investing more than it earns after tax: investment amounts to 108% of after-tax operating income. At 22% revenue growth, that means new investment is earning about 20%, less than the average on existing capital. That is still more than the required return, but less than before.

Management and direction

Management has shown its direction clearly through its actions. First, it is investing as much as it can, because, according to the chief financial officer, demand for computing power exceeds supply. Second, it now also sells TPU chips to outside customers and to other companies’ data centres, including through a partnership with Oracle and Crux AI, a joint venture with Blackstone. Third, it has replaced share buybacks with issues of shares and debt. Fourth, it is extending ads in AI-powered search to more users and, for the first time, outside the US.

Through these actions management is saying it believes the investment will earn more than the new capital costs. Whether it is right will be shown by Cloud and by what Search earns. This is where opinions divide.

The bear case: artificial intelligence takes money from Search, and investment earns less

The bear case is the view of those who expect harder times. At Alphabet it has two parts.

The first is Search. People ask more and more of their questions to chatbots and new forms of search, where they get an answer without clicking an ad. Management itself admits that AI Mode currently brings in less money. Growth in paid clicks fell from 13% to 8%, and that is not caused by currencies. Search brings in almost half of revenue and has so far been the most profitable part of the business. If earnings per search fall permanently, Cloud cannot replace all of it.

The second part is investment. The case runs like this:

  1. Investment is growing faster than revenue: $160 billion over the last four quarters, and more in 2027 [4].
  2. New investment earns less than existing capital, about 20% instead of 32%.
  3. Depreciation will hold down the margin for years, even if revenue growth falls.
  4. Commitments of $924 billion remain even if demand for artificial intelligence cools.

The result would be a company with a lower margin, less cash for shareholders and more shares.

The bull case: more searches, ads still to come, Cloud growing

The bulls do not deny that Search has slowed. They argue that it is a transition, not a decline.

If people were leaving Google, the number of searches would fall. According to Pichai, however, searches are growing at their fastest rate in five years, including those with buying intent. Management says most searches in AI Mode are additional: people are not using them instead of classic search. So people are not leaving. The problem is how much Google earns per search, and that is solvable.

Ads in the new search are only now arriving. Google is extending them in AI Overviews and AI Mode to more users and to foreign markets. According to Schindler, people in AI Mode click on advertised products about as often as on non-advertised ones. Ads in classic search are refined; in the new search they are only being introduced.

Cloud is not a bet, but a business. Revenue doubled, the margin is 36%, and the backlog is $561 billion. The Gemini app has a billion users, and Google’s models process almost 32 billion tokens (units of text) per minute, more than four times as many as a year ago.

Google has every layer of artificial intelligence. Its own chips, models, cloud and access to billions of users through Search, Android, YouTube and Chrome. That gives it an advantage over companies that have to rent computing power.

What analysts think

Most analysts remain on the bull side, but they cut their price targets after the results. A price target is the price an analyst expects in about a year.

Analysts (S&P Global, 29 Oct 2026)Value
Number of analysts58; 53 recommend buy, 5 hold
Average price target$408.63, 44% above the price of $283.82
Range of price targetsfrom $325 to $500
Expected revenue growth2026: +23.9%, 2027: +27.0%

After the results, price targets were cut by, among others:

  • J.P. Morgan, from $425 to $375,
  • Susquehanna, from $460 to $400,
  • Wedbush, from $445 to $387,
  • Truist, from $420 to $410,
  • D.A. Davidson, from $350 to $300 (rating hold).

The average target also includes targets set before the results.

What the price already includes

After the drop, the stock is 29.5% below its peak of 13 May 2026 and 9.3% lower than at the end of 2025 [2]. At $283.82 all shares are worth about $3,489 billion. Of that, about $257 billion is cash and investments, including SpaceX shares and the stake in Anthropic, minus debt.

What the price expects is shown by a reverse DCF. Normally one would forecast the future and calculate a value from it. A reverse DCF goes the other way: it takes today’s price and calculates what revenue growth that price already includes. The model takes into account that growth costs money: part of the profit has to go into investment. The required return is the return an investor wants each year. For comparison: from 1928 to 2025 the US S&P 500 index returned about 10% a year on average, including dividends.

Required returnPrice before results ($324.89)Price after the drop ($283.82)
8%9.4%7.6%
9%12.0%10.1%
10%14.5%12.6%

The table shows the revenue growth per year over the next ten years that the price includes, if the operating margin stays at 33.3%. After ten years, growth gradually falls to 2.5% over five years.

After the drop, the price expects growth close to the past Revenue growth per year; for the price: next ten years at a 10% required return and the same margin 0%10%20%30%Growth 2021 to 2025, per year11.8%Last four quarters22.2%Analysts for 202623.9%Analysts for 202727.0%Price before results ($324.89)14.5%Price after the drop ($283.82)12.6% Reverse DCF: Engineering Investor calculation (ROIC 31.6%, after ten years a five-year fade to 2.5%).

What this means. Suppose you want as much from the stock as the whole market has returned on average, about 10% a year. Then Alphabet’s revenue growth over the next ten years has to average 12.6% a year. That is only a little more than from 2021 to 2025, when it was 11.8% a year [1]. Before the drop, the price required 14.5%.

Analysts expect revenue growth of 23.9% this year and 27.0% next year. That is only two years out of ten, but both are well above 12.6%. After the drop, the price therefore expects roughly as much growth from Alphabet as it has achieved so far, not the acceleration analysts forecast. If revenue growth is higher than 12.6% a year, the stock’s annual return will be higher than 10%. If it is lower, the return will be lower.

This assumes new investment earns as much as existing capital (32%). If it earns only about 20%, growth has to be higher, about 15% a year.

What an investor gets at the price after the drop if one of three scenarios plays out over the next ten years (an illustration, not a forecast):

Scenario for ten yearsRevenue growth per yearOperating marginAnnual return
Cautious8%30%7.7%
Central12%33%9.7%
Favourable16%36%11.9%

The cautious scenario has growth below 2022 and 2023, the central one around the average of the last five years, and the favourable one above growth in 2025. Even in the favourable scenario the annual return is only a little above the market’s long-term return. The central scenario gives about as much as the market, the cautious one less. For a return clearly above the market, growth or the margin would have to be higher than in the favourable scenario.

This creates an asymmetry, meaning a difference between how much you can lose and how much you can gain. On the downside: if Search growth keeps slowing and depreciation pushes the margin down, the return is about 8% a year (the cautious scenario). On the upside: Cloud and chips grow faster than the central scenario assumes, and ads in the new search catch up with classic ones; then the return is about 12% a year. The two outcomes are roughly equally far from the market’s return. But on the downside the scenarios do not capture everything: a lasting decline in Search would give less than the cautious scenario.

What to watch to know who is right

All five points are different views of the same question: is artificial intelligence taking money from Search or only reshaping it, and does the investment pay off?

  1. Paid clicks and cost per click (quarterly Form 10-Q). If click growth returns above 10%, management was right that the causes were temporary.
  2. What ads in AI Mode earn. Management currently describes this only in words. The first figure on revenue or earnings per search will matter.
  3. Cloud backlog and margin. Renting outside computing power will weigh on the margin. It needs to stay well above 30%.
  4. The 2027 capital expenditure outlook and free cash flow. The outlook comes in February 2027 with the results for Q4 2026.
  5. The number of shares. Whether share sales on the stock market and the conversion of preferred shares dilute existing shareholders faster than profit grows.

The next milestone is the results for Q4 2026, which Alphabet will publish in early February 2027.

A conclusion everyone has to draw for themselves

Q3 2026 showed two sides of Alphabet. Cloud has gone from a side business to the company’s engine: revenue doubled, profit tripled, and customers have committed $561 billion. Search, however, clearly stalled for the first time: growth in paid clicks fell, management admits AI Mode earns less, and the core business’s operating income excluding last year’s fine was only 5.8% higher. At the same time, Google is investing so much that it has no cash left over, and it is issuing new shares.

Anyone who thinks the Search slowdown is temporary today gets a company with every layer of artificial intelligence under one roof, at a price that assumes roughly its past growth. Analysts expect considerably faster growth over the next two years. If growth over ten years is higher than 12.6% a year, the stock’s return will be higher than the market’s long-term average. Anyone who thinks it is the beginning of a decline sees investment that earns less and less, $924 billion of commitments, and the most profitable part of the company in a transition with no guaranteed end.

Who is right will not be clear within a year. 2027 will give the first signs: whether growth in paid clicks returns above 10%, and how much ads in AI Mode earn. Because investment in 2027 will, according to management, be even higher, low free cash flow that year will not yet prove that it does not pay off. The answer will come from Search and Cloud revenue and profit over several years, once today’s investment is at work.

Key takeaways

  • Revenue in Q3 2026: $127.6bn, growth of 25%; the 13th quarter in a row with growth above 10%.
  • The stock fell 12.6% the day after the results; the value of all shares fell by about $505bn.
  • Search revenue growth was 12%, against 16.8% in Q2. The slowdown is in the number of paid clicks (growth of 13% in Q2, 8% in Q3), not in cost per click.
  • Management gives four reasons: currencies (the largest), the end of an EU tax exemption, the World Cup, which shifted advertising into Q2, and lower earnings from AI Mode.
  • Google Cloud: revenue of $30.7bn (growth of 103%), operating income 3.1 times higher, a 36.2% margin, a contract backlog of $560.8bn.
  • The core Google Services business: excluding the Q3 2025 fine, operating income growth was only 5.8%.
  • Capital expenditure: $51.4bn in Q3 (40% of revenue); in 2027 it will “increase significantly”.
  • Free cash flow in Q3: −$9.1bn, including a $5.2bn European Commission fine payment.
  • No share buybacks: Alphabet is financing itself with new shares ($42.9bn), preferred shares ($19.1bn) and debt ($85.2bn); there are 1.7% more shares than at the end of 2025.
  • Reverse DCF: at the price after the drop and a 10% required return, the price includes about 12.6% revenue growth a year over ten years; from 2021 to 2025 it was 11.8%.
  • Ten-year scenarios give annual returns of 7.7% to 11.9%; over the long term the S&P 500 has returned about 10% a year.
  • What to watch: growth in paid clicks, earnings from AI Mode, the Cloud margin, investment in 2027 and the number of shares.

Key metric charts

The charts that the numbers in the text point to. Quarterly metrics for the last five years are from Alphabet’s Forms 10-Q and 10-K.

Revenue by quarter Last five years; top: revenue, $bn; bottom: revenue growth 03570105140Q4 2021: $75.3bnQ1 2022: $68.0bnQ2 2022: $69.7bnQ3 2022: $69.1bnQ4 2022: $76.0bnQ1 2023: $69.8bnQ2 2023: $74.6bnQ3 2023: $76.7bnQ4 2023: $86.3bnQ1 2024: $80.5bnQ2 2024: $84.7bnQ3 2024: $88.3bnQ4 2024: $96.5bnQ1 2025: $90.2bnQ2 2025: $96.4bnQ3 2025: $102.3bnQ4 2025: $113.8bnQ1 2026: $109.9bnQ2 2026: $119.8bnQ3 2026: $127.6bn75.3127.6Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3202120222023202420252026Revenue growth on the same quarter a year earlier0%10%20%30%Q4 2021: 32.4%Q1 2022: 23.0%Q2 2022: 12.6%Q3 2022: 6.1%Q4 2022: 1.0%Q1 2023: 2.6%Q2 2023: 7.1%Q3 2023: 11.0%Q4 2023: 13.5%Q1 2024: 15.4%Q2 2024: 13.6%Q3 2024: 15.1%Q4 2024: 11.8%Q1 2025: 12.0%Q2 2025: 13.8%Q3 2025: 15.9%Q4 2025: 18.0%Q1 2026: 21.8%Q2 2026: 24.2%Q3 2026: 24.7%32.4%1.0%24.7%Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3202120222023202420252026
Alphabet: share price over the last 12 months Daily closing prices of Class A shares (GOOGL), USD 250300350400peak $402.62before results $324.89$283.8229 Oct 2026−12.6%in one dayNov 2025Jan 2026Apr 2026Jul 2026Oct 2026
Cloud carries a growing share of revenue Quarterly revenue, $bn; above each bar, Cloud as a share of total revenueGoogle ServicesGoogle Cloud 0357010514013%Q3 2412%Q4 2414%Q1 2514%Q2 2515%Q3 2516%Q4 2518%Q1 2621%Q2 2624%Q3 26 Other Bets and hedging (together under 1% of revenue) are not shown.
Capital expenditure by quarter Last five years; top: capital expenditure, $bn; bottom: share of revenue 015304560Q4 2021: $6.4bnQ1 2022: $9.8bnQ2 2022: $6.8bnQ3 2022: $7.3bnQ4 2022: $7.6bnQ1 2023: $6.3bnQ2 2023: $6.9bnQ3 2023: $8.1bnQ4 2023: $11.0bnQ1 2024: $12.0bnQ2 2024: $13.2bnQ3 2024: $13.1bnQ4 2024: $14.3bnQ1 2025: $17.2bnQ2 2025: $22.4bnQ3 2025: $24.0bnQ4 2025: $27.9bnQ1 2026: $35.7bnQ2 2026: $44.9bnQ3 2026: $51.4bn6.451.4Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3202120222023202420252026Capital expenditure as a share of revenue0%15%30%45%Q4 2021: 8.5%Q1 2022: 14.4%Q2 2022: 9.8%Q3 2022: 10.5%Q4 2022: 10.0%Q1 2023: 9.0%Q2 2023: 9.2%Q3 2023: 10.5%Q4 2023: 12.8%Q1 2024: 14.9%Q2 2024: 15.6%Q3 2024: 14.8%Q4 2024: 14.8%Q1 2025: 19.1%Q2 2025: 23.3%Q3 2025: 23.4%Q4 2025: 24.5%Q1 2026: 32.5%Q2 2026: 37.5%Q3 2026: 40.3%8.5%40.3%Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3202120222023202420252026
Free cash flow by quarter Operating cash flow minus capital expenditure, $bn; last five years -1501530Q4 2021: $18.6bnQ1 2022: $15.3bnQ2 2022: $12.6bnQ3 2022: $16.1bnQ4 2022: $16.0bnQ1 2023: $17.2bnQ2 2023: $21.8bnQ3 2023: $22.6bnQ4 2023: $7.9bnQ1 2024: $16.8bnQ2 2024: $13.5bnQ3 2024: $17.6bnQ4 2024: $24.8bnQ1 2025: $19.0bnQ2 2025: $5.3bnQ3 2025: $24.5bnQ4 2025: $24.6bnQ1 2026: $10.1bnQ2 2026: −$5.9bnQ3 2026: −$9.1bn18.624.8−5.9−9.1Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3202120222023202420252026 Q3 2026 includes a $5.2bn European Commission fine payment (Form 10-Q).
Operating income by quarter Last five years; top: operating income, $bn; bottom: operating margin 0153045Q4 2021: $21.9bnQ1 2022: $20.1bnQ2 2022: $19.5bnQ3 2022: $17.1bnQ4 2022: $18.2bnQ1 2023: $17.4bnQ2 2023: $21.8bnQ3 2023: $21.3bnQ4 2023: $23.7bnQ1 2024: $25.5bnQ2 2024: $27.4bnQ3 2024: $28.5bnQ4 2024: $31.0bnQ1 2025: $30.6bnQ2 2025: $31.3bnQ3 2025: $31.2bnQ4 2025: $35.9bnQ1 2026: $39.7bnQ2 2026: $40.8bnQ3 2026: $40.6bn21.940.6Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3202120222023202420252026Operating margin: how much of each $100 of revenue is left as operating income20%30%40%Q4 2021: 29.1%Q1 2022: 29.5%Q2 2022: 27.9%Q3 2022: 24.8%Q4 2022: 23.9%Q1 2023: 25.0%Q2 2023: 29.3%Q3 2023: 27.8%Q4 2023: 27.5%Q1 2024: 31.6%Q2 2024: 32.4%Q3 2024: 32.3%Q4 2024: 32.1%Q1 2025: 33.9%Q2 2025: 32.4%Q3 2025: 30.5%Q4 2025: 31.6%Q1 2026: 36.1%Q2 2026: 34.0%Q3 2026: 31.8%29.1%23.9%36.1%31.8%Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3202120222023202420252026 Q3 2025 includes a $3.5bn European Commission fine.
Cloud: revenue doubled, operating income tripled Google Cloud by quarter, $bnRevenueOperating income 0816243211.41.9Q3 2412.02.1Q4 2412.32.2Q1 2513.62.8Q2 2515.23.6Q3 2517.75.3Q4 2520.06.6Q1 2624.88.8Q2 2630.711.1Q3 26
Buybacks have stopped and the share count is rising again Last five years; top: cash spent on share buybacks, $bn; bottom: number of shares 05101520Q4 2021: $13.5bnQ1 2022: $13.3bnQ2 2022: $15.2bnQ3 2022: $15.4bnQ4 2022: $15.4bnQ1 2023: $14.6bnQ2 2023: $15.0bnQ3 2023: $15.8bnQ4 2023: $16.2bnQ1 2024: $15.7bnQ2 2024: $15.7bnQ3 2024: $15.3bnQ4 2024: $15.6bnQ1 2025: $15.1bnQ2 2025: $13.2bnQ3 2025: $11.5bnQ4 2025: $5.9bnQ1 2026: $0.0bnQ2 2026: $0.0bnQ3 2026: $0.0bn13.516.20Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3202120222023202420252026Shares outstanding at quarter end, billions12.012.513.013.5Q4 2021: 13.24bn sharesQ1 2022: 13.18bn sharesQ2 2022: 13.08bn sharesQ3 2022: 12.97bn sharesQ4 2022: 12.85bn sharesQ1 2023: 12.72bn sharesQ2 2023: 12.63bn sharesQ3 2023: 12.54bn sharesQ4 2023: 12.46bn sharesQ1 2024: 12.38bn sharesQ2 2024: 12.32bn sharesQ3 2024: 12.26bn sharesQ4 2024: 12.21bn sharesQ1 2025: 12.15bn sharesQ2 2025: 12.10bn sharesQ3 2025: 12.08bn sharesQ4 2025: 12.09bn sharesQ1 2026: 12.12bn sharesQ2 2026: 12.23bn sharesQ3 2026: 12.29bn shares13.2412.0812.29Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3202120222023202420252026 Class A, B and C shares; counts before the 2022 20-for-1 split are restated.
Net income swings with investments, operating income does not Last five years, by quarter, $bnOperating incomeNet income 0306090120Q4 2021: $21.9bnQ1 2022: $20.1bnQ2 2022: $19.5bnQ3 2022: $17.1bnQ4 2022: $18.2bnQ1 2023: $17.4bnQ2 2023: $21.8bnQ3 2023: $21.3bnQ4 2023: $23.7bnQ1 2024: $25.5bnQ2 2024: $27.4bnQ3 2024: $28.5bnQ4 2024: $31.0bnQ1 2025: $30.6bnQ2 2025: $31.3bnQ3 2025: $31.2bnQ4 2025: $35.9bnQ1 2026: $39.7bnQ2 2026: $40.8bnQ3 2026: $40.6bnQ4 2021: $20.6bnQ1 2022: $16.4bnQ2 2022: $16.0bnQ3 2022: $13.9bnQ4 2022: $13.6bnQ1 2023: $15.1bnQ2 2023: $18.4bnQ3 2023: $19.7bnQ4 2023: $20.7bnQ1 2024: $23.7bnQ2 2024: $23.6bnQ3 2024: $26.3bnQ4 2024: $26.5bnQ1 2025: $34.5bnQ2 2025: $28.2bnQ3 2025: $35.0bnQ4 2025: $34.5bnQ1 2026: $62.6bnQ2 2026: $112.2bnQ3 2026: $26.5bn112.240.626.5Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3202120222023202420252026 Q1 and Q2 2026: gains on investments ($135.7bn of other income); Q3 2026: a fall in the value of the SpaceX stake.

Notes on the data: Changes in the tables are calculated from unrounded amounts. Cloud growth is 102.6%, which the company rounds to 103%. Analysts’ expectations come from different consensus providers (LSEG, StreetAccount); other providers cite slightly different values. Earnings per share excluding investments ($2.67) is reported earnings per share ($2.12) plus the effect of the loss on equity securities ($0.55) as stated in the earnings release. Paid clicks and cost per click are from Form 10-Q filings; for Q4 2025 Alphabet reports only full-year figures. The reasons for the Search slowdown are summarised from Philipp Schindler’s answer on the earnings call; in his words, the shift of advertising from July to June is part of the World Cup reason. The currency effect is from the revenue-by-geography tables in the Q2 and Q3 2026 releases and applies to total revenue; Alphabet does not report it for Search. Cloud’s share of revenue growth is the increase in Cloud revenue divided by the increase in total revenue. Cloud revenue per year (about $123bn) is Q3 2026 revenue multiplied by four; the average from the backlog (about $154bn a year) is 55% of $560.8bn divided by two. Free cash flow is operating cash flow minus capital expenditure, as defined by Alphabet. The quarterly metrics in the charts are from XBRL data; Form 10-K reports only the full year, so Q4 is the year minus the first three quarters. Values from Q3 2024 to Q3 2026 match the earnings releases. Share counts before the 2022 20-for-1 split are restated. Issues of shares, preferred shares and debt are from the cash flow statement for Q1 to Q3 2026, with details (the public offering, the sale to Berkshire Hathaway, the ATM programme, the terms of the preferred shares) from the notes in Form 10-Q; the increase in the share count compares 12,088 million (31 Dec 2025) with 12,294 million (30 Sep 2026). ROIC is operating income over the last four quarters ($157.0bn) after a 19% tax rate (Q3 2026 excluding the investment loss), divided by invested capital at 30 Sep 2026 (equity plus debt, minus cash, marketable securities and non-marketable investments; $402bn). The return on new investment is revenue growth over the last four quarters (22.2%) divided by investment as a share of after-tax profit (108%), where investment is after-tax operating income minus free cash flow after stock-based pay. Reverse DCF: revenue over the last four quarters $471.1bn, ten years of the growth being solved for, then five years of decline to 2.5%, investment for growth via ROIC, market value from 12,294 million shares at the GOOGL price, non-operating assets $257bn (cash and marketable securities 265.1, minus debt 125.1, plus non-marketable investments 135.2, minus preferred shares 18.0). The first version of this article (30 Oct 2026) solved for growth over five years; this one looks at ten. Share classes A, B and C trade at slightly different prices; the Class A price is used. The long-term S&P 500 return (10.0% a year) is the average of annual returns including dividends from 1928 to 2025, compounded year by year (Engineering Investor calculation from Damodaran’s data); the simple average is 11.9%. Anthropic’s commitment as a share of the backlog is 114.5 divided by $560.8bn; Google does not disclose whether the commitment is fully in the backlog. Return scenarios are an illustration of the calculation over ten years, not a forecast.

Educational content, not financial advice. The author does not urge anyone to buy or sell any security. The data are taken from the sources listed above as of 29 Oct 2026 and may change.