By Matthew Bertram · October 28, 2026
Alphabet just lost roughly 10% of its stock value in after-hours trading following its third-quarter earnings announcement.
The market’s reaction is understandable. Google’s advertising growth is slowing, its AI infrastructure costs are enormous, and investors are questioning whether artificial intelligence will eventually undermine the business that made Google one of the most valuable companies in history.
But I think there’s another way to interpret what we’re watching.
Wall Street is measuring Google against a buying funnel that Google itself is working to replace.
And that distinction matters.
I’ve spent years working in search engine optimization, digital marketing, and AI visibility. I’ve watched Google evolve from a search engine into the central distribution system for much of the internet’s commercial activity.
What I see happening now is more significant than another algorithm update or a change in search rankings.
I believe we’re witnessing the beginning of the end of the traditional digital buying funnel.
And Google’s earnings call may have revealed just how far along that transition already is.
Two notes. The 10% is the October 28 after-hours reaction, not a closing price, and the outlets I checked put it between 8.7% and 10.3%. And I own Alphabet stock (GOOGL), so read this as the view of someone with a position.
On October 28, 2026, Alphabet reported quarterly revenue of approximately $127.6 billion, an increase of 25% year over year.
Google Cloud revenue increased an extraordinary 103%, reaching $30.7 billion.
CEO Sundar Pichai reported that Search query growth was at its highest year-over-year rate in five years. Gemini had reached one billion users.
Those aren’t the numbers of a company losing relevance.
Operating margin went from 23.7% to 36.2%. Revenue grew 103%.
Q3 2025Q3 2026
| Google Cloud ($M) | Q3 2025 | Q3 2026 |
|---|---|---|
| Revenue | 15,157 | 30,709 |
| Operating income | 3,594 | 11,121 |
| Operating margin (mine) | 23.7% | 36.2% |
Alphabet does not split the $51.4B of capital spending by segment, so this margin is evidence about Cloud, not proof the whole bill earns its return.
Yet Alphabet shares fell approximately 10% in after-hours trading.
Why?
Because the financial results revealed growing tension between Google’s existing business model and the infrastructure required to support its next one.
Google Search and related advertising revenue grew approximately 12%, down from 17% growth in the previous quarter.
Alphabet reported earnings of $2.12 per share, significantly below expectations. Importantly, a substantial non-operating investment loss contributed to the earnings miss, so it would be misleading to attribute the entire shortfall to weakness in Google’s operating business.
The same three quarters, three lines from Alphabet’s SEC filings. Each panel has its own scale.
Earlier quartersQ3 2026 (reported Oct 28)
| Line | Q3 2025 | Q2 2026 | Q3 2026 |
|---|---|---|---|
| Operating income ($M) | 31,228 | 40,770 | 40,576 |
| Gain (loss) on equity securities ($M) | 10,734 | 99,031 | (8,664) |
| Diluted EPS | $2.87 | $9.11 | $2.12 |
Figures as printed in the filing. The equity-securities line is pre-tax, and Alphabet does not print an EPS figure excluding it.
But investors had other legitimate concerns.
The company spent approximately $51.4 billion on capital expenditures during the quarter, with free cash flow falling to negative $9.1 billion.
Management also indicated that AI-related infrastructure spending would continue growing significantly in 2027.
Those are enormous commitments.
Still, what caught my attention wasn’t simply the financial performance.
It was management’s explanation of what is happening to Search.
During the earnings call, Google’s chief business officer, Philipp Schindler, acknowledged that the rapid growth of AI Mode contributed to slower Search advertising growth because monetization in AI Mode remains lower than in traditional search.
That statement deserves more attention than it’s getting.
Google isn’t necessarily losing the consumer.
Google is changing the consumer experience faster than it has figured out how to monetize it.
Those are very different problems.
Google Search and other revenue by quarter, $ billions. The step from the second quarter to the third added 4.4% last year and 0.1% this year.
Earlier quartersQ3 2026
| Quarter | Search and other ($M) | Change from prior quarter |
|---|---|---|
| Q1 2025 | 50,702 | n/a |
| Q2 2025 | 54,190 | +6.9% |
| Q3 2025 | 56,567 | +4.4% |
| Q4 2025 | 63,073 | +11.5% |
| Q1 2026 | 60,399 | −4.2% |
| Q2 2026 | 63,271 | +4.8% |
| Q3 2026 | 63,350 | +0.1% |
Year-over-year growth as printed: Q4 2025 +17%, Q1 2026 +19%, Q2 2026 +17%, Q3 2026 +12%. Quarter-over-quarter changes are my arithmetic from the printed dollar figures.
Pichai also explained that Google is prioritizing the quality of its AI search experience before fully developing monetization, comparing the transition to the company’s earlier shift toward mobile.
From a traditional financial perspective, that creates immediate pressure.
From a technology and distribution perspective, however, it raises a much bigger question:
What happens when Google no longer needs to monetize every step of the buying journey because it controls the moment the buying decision is made?
That’s where I think much of the current analysis falls short.
For decades, digital marketing has operated around a relatively predictable customer journey.
A consumer becomes aware of a problem, searches for information, compares alternatives, visits websites, evaluates products, and eventually makes a purchase.
Businesses have spent billions building marketing operations around those stages.
SEO captures information-seeking consumers.
Paid search captures commercial intent.
Content marketing supports product education.
Retargeting brings people back.
Conversion optimization helps turn website visits into customers.
The entire system assumes that consumers move between platforms, websites, and information sources before making decisions.
Artificial intelligence challenges that assumption.
Consider how the customer journey is changing.
Outlined steps are where the work happens: across the open web in the first journey, inside one system in the second.
A consumer may no longer need to visit five websites to compare products.
They may not need to read ten reviews.
They may not even need to interact with a traditional search results page.
They can simply describe what they want and ask an AI system to find the best option.
Eventually, they can authorize that system to complete the transaction.
The funnel doesn’t necessarily disappear completely. Consideration, trust, and evaluation still happen.
But increasingly, those activities happen inside the AI system rather than across dozens of websites.
The funnel is being compressed, and the intermediary is absorbing it.
That changes everything.
The compression is not the same at every step, so it helps to take the stages one at a time.
The compression is not the same at every step. This is my reading, not a measurement.
| Stage | Today | With an AI system in the middle |
|---|---|---|
| Search | SEO and content marketing answer the searcher and earn the click. | The person states the need in a sentence and the system answers it. The click may never happen. |
| Compare and visit websites | The shopper, tab by tab. Paid search buys the visit. | The system reads specs, prices, reviews and availability, then returns a short list. |
| Evaluate | Product pages and conversion optimization make the case. | The reasoning arrives with the recommendation, inside the system. |
| After the visit | Retargeting follows the visitor who left. | A decision made in the conversation leaves no abandoned visit to chase. |
| Purchase | The merchant’s own checkout. | Checkout inside the assistant. The merchant stays the seller of record. |
Retargeting is the cleanest example. It works by following a visitor who left a website. If the decision is made in a conversation, there may be no visit to follow.
What does not change is the work behind each stage. Someone still has to have the right product at the right price. Someone still has to be trusted. That work moves inside the system that makes the recommendation, and the businesses that built their marketing around the middle stages have the most to relearn.
I am not describing something that starts next year. Three pieces of evidence point the same way: the middle of the funnel is already being compressed.
Answers are replacing comparison. In July 2025 the Pew Research Center published a study built on the browsing of 900 U.S. adults and 68,879 Google searches. When an AI summary appeared, people clicked a traditional search result in 8% of visits. When it did not, they clicked in 15%. Clicks on a link inside the summary itself happened in 1% of visits. Google called the methodology flawed and says it has not observed significant drops in aggregate web traffic, and the data was gathered in March 2025, before AI Mode reached everyone in the U.S. I read it as direct evidence for what I am describing: fewer trips to the websites in the middle.
Share of Google visits that included a click, by what the results page showed. Browsing data from 900 U.S. adults.
Results page without an AI summaryResults page with an AI summary
| Share of visits | No AI summary | AI summary shown |
|---|---|---|
| Clicked a traditional result | 15% | 8% |
| Clicked a link inside the AI summary | not applicable | 1% |
18% of the 68,879 searches in Pew’s sample produced an AI summary. Browsing activity is from March 2025; search results were collected April 7 to 17, 2025.
The people who still click arrive further down the funnel. Adobe Analytics, which measures more than a trillion visits to U.S. retail sites, found that AI-referred visits converted at a rate 60% higher than other traffic in July and produced 53% more revenue per visit. AI referrals to U.S. retail sites were up 62% on the year, and it was the eleventh straight month that AI-referred traffic out-converted the rest. These are one vendor’s figures, and they count only visits that clicked through, so they cannot see a decision made without a click. My reading is that a shopper who arrives from an AI answer has already done the comparing.
AI-referred visits to U.S. retail sites, July 2026, compared with other traffic.
The checkout is moving into the assistant, and not only at Google. At I/O in May 2025, Google announced an agentic checkout in AI Mode that can buy an item with Google Pay when the price is right. In September 2025, OpenAI added Instant Checkout to ChatGPT, built on a protocol it developed with Stripe, with Etsy sellers first and Shopify merchants next. Google introduced the Universal Commerce Protocol in January 2026, announced Universal Cart in May, and on the October call Schindler said UCP is being rapidly adopted by retailers.
Dated announcements, in order.
OpenAIGoogle
AI Mode starts rolling out to everyone in the U.S. Google also announces an agentic checkout that can buy an item with Google Pay “when the price is right.”
Instant Checkout arrives in ChatGPT, built on a protocol OpenAI developed with Stripe. Etsy sellers first, Shopify merchants next.
Google unveils the Universal Commerce Protocol at the National Retail Federation conference (Pichai’s remarks).
Universal Cart is announced at I/O: one cart across Search, Gemini, YouTube and Gmail, with checkout through Google Pay or the merchant’s site.
On the earnings call, Philipp Schindler says UCP “is being rapidly adopted by retailers” and that Alphabet “launched Universal Cart this quarter.”
Two companies building the same layer within months of each other tells me the direction is not one company’s preference.
Here’s the development I believe investors should be examining more closely.
Google has introduced the Universal Commerce Protocol (UCP), an open standard designed to connect AI agents, merchants, payment systems, and commerce infrastructure.
Its purpose is straightforward.
Enable consumers to discover products and complete purchases directly within AI-powered experiences such as Google Search’s AI Mode and Gemini.
During the earnings call, Google discussed expanding agentic capabilities and the adoption of UCP by retailers.
Google also described Universal Cart, which helps facilitate shopping across retailers. Google announced it at I/O in May, and Schindler told investors it launched this quarter.
This is a meaningful change in the architecture of digital commerce.
Google is moving beyond helping consumers find merchants.
It’s developing the infrastructure that allows AI to help consumers select products and complete purchases.
To be precise, merchants can remain the seller of record and retain their customer relationships. Google doesn’t have to own the underlying transaction to control an increasingly valuable part of the experience.
It can control the interface through which purchase decisions are made.
And that interface is where commercial influence becomes concentrated.
The architecture Google describes for UCP checkout, drawn as three layers.
A caution on scale. Google’s Merchant Center help page describes UCP checkout as a feature for select merchants in the United States, Canada and Australia, and Alphabet’s earnings release does not mention AI Mode, UCP or agentic commerce. On the call, Schindler said UCP “enables direct checkout for hundreds of thousands of brands across Google.” That is management’s figure, and I could not check it independently.
Think about the implications.
Today, businesses compete to appear in search results.
Tomorrow, they may increasingly compete to be selected by an AI system acting on the customer’s behalf.
That’s not simply a new advertising placement.
It’s a different distribution model.
The immediate concern is that AI-powered search generates less advertising revenue than traditional Google Search.
That’s a legitimate issue.
But the assumption underlying much of the concern is that Google’s future economics must resemble its past economics.
I don’t believe that assumption is necessarily correct.
Historically, Google monetized commercial intent through advertising opportunities throughout the customer journey.
An advertiser paid to appear when someone searched for a product or service.
The user clicked.
The advertiser attempted to convert that visitor into a customer.
Google made money from connecting demand with potential suppliers.
But AI creates the possibility of a more direct connection between intent and transaction.
If Google knows what a consumer wants, can identify suitable products, and can facilitate checkout, it may eventually be able to monetize that interaction differently.
Potential models include advertising, sponsored recommendations, transaction-related services, merchant integrations, and other commerce infrastructure.
I’m not suggesting Google has already established a replacement revenue model at scale.
It hasn’t demonstrated that yet.
And there is no guarantee that transaction-oriented monetization will generate the same margins as traditional search advertising.
But I believe investors may be underestimating how much commercial value can shift toward the system that makes or influences the final decision.
The important metric may eventually be less about revenue per search and more about the economic value of purchase intent captured and converted.
That’s a fundamentally different way to evaluate Google’s position.
This is the paradox that makes the current situation interesting.
Google’s AI products may undermine parts of its traditional advertising model.
They may reduce website referrals.
They may eliminate intermediate searches.
They may compress the number of advertisements consumers encounter before purchasing.
And they may make some existing SEO and paid-media strategies less effective.
But that doesn’t automatically mean Google loses.
If Google becomes the primary interface through which consumers research, select, and purchase products, it may strengthen its position in the commercial ecosystem even while weakening its legacy revenue model.
In technology, protecting an existing business model can sometimes be more dangerous than disrupting it yourself.
Google appears willing to make that tradeoff.
Whether it can execute the transition profitably is another matter.
The infrastructure costs are real. Competition from ChatGPT and other AI platforms is real. Merchant adoption, payment integration, user trust, and regulatory scrutiny could all complicate the strategy.
But there’s a difference between a company experiencing disruption and a company attempting to control the architecture of that disruption.
I believe Google is doing the latter.
For marketers, the implications extend well beyond Google’s stock price.
I’ve argued for some time that optimizing exclusively for rankings and traffic is becoming an increasingly incomplete strategy.
Ranking still matters.
Websites still matter.
Content still matters.
But the competitive objective is changing.
Historically, companies optimized to be found.
In AI-mediated commerce, they increasingly need to be understood, trusted, and selected.
That means the information available about a company, its products, its expertise, and its credibility becomes critical.
AI systems need reliable information to evaluate alternatives.
For ecommerce businesses, that includes product specifications, pricing, availability, fulfillment, reviews, and structured product information.
For B2B organizations, it includes capabilities, certifications, service coverage, industry experience, authoritative documentation, and evidence of operational credibility.
For professional services, it includes expertise, reputation, regulatory context, and demonstrable experience.
What an AI system needs in order to evaluate a business, by type of business.
The challenge isn’t simply getting mentioned in an AI-generated answer.
It’s ensuring that the information an AI system uses to evaluate your business is accurate, consistent, current, and authoritative.
That’s the distinction I make between traditional SEO and broader AI visibility and Digital Information Governance® (DIG).
A business can rank well and still be misunderstood by AI systems.
It can generate traffic and still be excluded from an AI-generated shortlist.
And it can have an excellent website while the information AI systems rely upon elsewhere is inconsistent or outdated.
As more of the buying decision moves into AI interfaces, those weaknesses become commercial risks.
How stable is the shortlist? We measured it. Across 12,778 AI answers on five engines, ChatGPT reproduced only 13.3% of its own citations when asked the same question again (the study is on EWR Digital). Being on the shortlist is closer to a probability than a ranking position.
Visibility is no longer just about being discovered. It’s about being correctly represented at the moment a decision is made.
The same structural shift is likely to affect B2B purchasing.
Imagine a procurement executive asking an AI agent to identify qualified industrial suppliers that meet specific technical requirements, operate in certain jurisdictions, carry appropriate certifications, and can meet delivery deadlines.
Traditionally, that process might involve search engines, industry directories, websites, referrals, sales representatives, and multiple rounds of research.
AI can increasingly compress the initial evaluation into a single process.
The buyer may never visit most of the companies considered.
They may only interact with the two or three that the AI system recommends.
For some purchases, the AI agent may eventually help solicit quotes, validate vendor information, and initiate procurement workflows.
That changes the role of the company website.
It changes the role of search visibility.
And it changes the importance of authoritative, machine-readable business information.
In that environment, companies aren’t just competing for human attention.
They’re competing for inclusion in machine-mediated decisions.
For businesses that have spent years treating their online presence as a collection of marketing channels rather than a governed information system, the transition could be painful.
None of this means Alphabet’s selloff is unjustified.
The company has some difficult questions to answer.
Can AI Mode eventually monetize commercial activity as effectively as traditional Search?
Can Google generate sufficient returns on the hundreds of billions of dollars being committed to AI infrastructure?
Third quarter of 2025 against the third quarter of 2026, $ billions.
Q3 2025Q3 2026
| Line ($M) | Q3 2025 | Q3 2026 |
|---|---|---|
| Capital expenditures | 23,953 | 51,355 |
| Free cash flow | 24,461 | (9,065) |
| Repurchases of stock | 11,504 | 0 |
| Proceeds from debt, net | 6,624 | 28,996 |
| Proceeds from common stock, net | 0 | 12,371 |
| New debt and stock sold (sum, mine) | 6,624 | 41,367 |
Will merchants accept Google’s growing influence over customer discovery and purchase decisions?
And perhaps most importantly, will consumers use Google’s AI interfaces, or will independent AI assistants increasingly bypass Google entirely?
That last question may be the most consequential.
If consumers begin their purchase journey in ChatGPT or another independent assistant, Google’s historical advantage as the default starting point for commercial discovery weakens.
Google clearly recognizes this threat.
Its response is to integrate AI more deeply into Search, Gemini, advertising, commerce, and its broader product ecosystem.
The company has distribution, infrastructure, merchant relationships, and decades of consumer behavior data.
But those advantages do not guarantee dominance in the next model.
The outcome is far from settled.
I understand why Alphabet sold off.
Investors are being asked to tolerate massive capital expenditures, negative quarterly free cash flow, and uncertainty around the monetization of a rapidly changing consumer experience.
Those are not trivial concerns.
But I think there’s a risk in evaluating this transition exclusively through traditional advertising metrics.
Google may be moving from monetizing the process of searching to monetizing a more direct relationship with purchasing intent.
If that transition works, lower revenue per AI search today may look very different in retrospect.
If it fails, the current investment cycle could materially weaken the economics of one of the world’s most profitable businesses.
Both outcomes remain possible.
My argument isn’t that Alphabet is automatically a bargain because the stock fell.
It’s that the market may be evaluating Google’s next business model using measurements designed for its previous one.
And from the perspective of someone who has spent years working inside the digital marketing ecosystem, the larger change seems increasingly clear.
The customer journey is being reorganized around AI.
The traditional funnel is compressing.
The importance of traffic as an intermediate objective may decline.
And the systems that control product discovery, evaluation, recommendation, and transactions are positioned to capture an increasing share of commercial influence.
Google understands this.
The question is whether the market fully understands what Google is attempting to become.
The future of search isn’t just about who answers the question. It’s about who influences the decision, and who facilitates the transaction.
That’s the shift I believe we’re watching.
And it could be much more important than a single quarter’s earnings miss.
Matthew Bertram is a digital strategy advisor, owner of EWR Digital, and host of the Best SEO Podcast. His work focuses on search, AI visibility, Digital Information Governance® (DIG), and the evolving relationship between information, technology, and business decisions.
This is commentary on public filings, a conference call and published research, written the evening of 28 October 2026. It is not investment advice and not a recommendation to buy or sell any security. I own Alphabet stock (GOOGL). Figures are US dollars from Alphabet’s SEC filings unless a source is named. Where I calculated something (margins, quarter-over-quarter changes, the new-capital total), the arithmetic is mine and is shown in each chart’s data table.
How buyers get shortlisted by machines: what generative engine optimization is and the CEO guide to AI SEO. The governance argument behind this work is in the Digital Information Governance® (DIG) research paper.