Beneath the noise: Google’s multi-engine growth story

In a year marked by rising anxiety over the disruptive potential of AI, Google has found itself at the centre of investor debate. Concerns around competition, changing user behaviour, and regulatory pressure have weighed on the stock, despite continued growth in revenues, product innovation, and AI capability. In this note, we take a step back from the noise and examine the underlying economics of Google’s business. Our view is simple: Google remains one of the most dominant platform businesses in the world. With a leading position in AI, an expanding cloud business, and a still-durable advertising engine, we believe the market is underestimating the company’s long-term earnings power, and mispricing its stock accordingly.

Why the Market Is Worried – and Why We Disagree

The market has recently become fearful that, with the advent of AI chatbots, competition will kill Google’s advertising business. This line of thought has driven investors to dump the stock with maximum pressure arriving when an Apple representative, during an antitrust court case, said that it saw declining Google searches on iOS for the first time in April.

However, we believe the fear of declining ad revenues is misplaced. There are a few reasons why.

• Resilient Growth Despite AI Adoption: Over the last 2-3 years when AI chatbot adoption has exploded from essentially no users to some estimates of 1 billion users worldwide, Google’s revenues have grown from $283 billion to $350 billion. That’s an 11% annual growth rate, even with peak hype around AI. If anything, we believe the most aggressive phase of chatbot adoption is behind us. It stands to reason that slower adoption from here is unlikely to reverse or even meaningfully impair revenue growth.
• Enduring AI Search Advantages: Google’s ability to pair its own search index with LLMs offers a unique edge in delivering timely, relevant answers – something standalone models often struggle with. For instance, LLMs are notoriously poor at understanding time-sensitive queries, such as sports scores, news events, or product availability. Google’s real-time data advantage is not easily replicable.
• User Habits and Platform Stickiness: Google already has a natural advantage in being the place where people go to ask their questions already. Microsoft has found this out the hard way as it pushes AI tools into its software ecosystem, reportedly scaling back its data centre buildout in response to lukewarm adoption. Google, by contrast, is integrating AI features into products that users already rely on daily, from Search and Gmail to Maps and Docs, making the transition seamless and sticky.
• Contextual AI Requires Deep Data: The next frontier in AI utility lies in its ability to understand user context. To be truly useful, AI will need access to workplace and personal data to tailor responses. Realistically, only two companies have this depth of structured, user-level data at scale: Google and Microsoft. That gives Google an embedded advantage as contextual AI evolves.
• Product Integration Unlocks Differentiation: Google’s unmatched ability to bundle services strengthens its competitive moat. For example, it’s the only company with a fully geolocated index of websites. Combining this with tools like Gmail and Maps allows Google to provide more relevant, comprehensive answers than any standalone AI agent. Consumers tend to converge around the platform they perceive to offer the best utility, and Google is positioning itself to win that battle.

AI Isn’t a Threat, It’s an Opportunity

Google’s leadership in AI enables two key revenue opportunities:

1. Charging directly for AI tools, which supports the continued growth of its cloud business, and
2. Bundling those tools into a cohesive, high-utility interface that reinforces platform stickiness.

Importantly, user attention, and advertiser budgets, will likely continue to centre around Google Search for the reasons outlined above. In “AI mode,” advertisers can still reach audiences much as before. Similar concerns about monetisation surfaced in the early days of Search, yet Google proved it could integrate advertising seamlessly into its core product. As long as the platform remains the natural destination for finding information, it will continue to find ways to monetise user intent effectively.

Meanwhile, competitors are discovering just how difficult it is to sustain progress in large language model development. OpenAI has repeatedly delayed the release of GPT-5, while Meta’s Behemoth LLM has faced performance issues and missed timelines. Apple appears to have deprioritised in-house AI development, and Microsoft’s model access is contingent on its partnership with OpenAI, a deal currently set to expire in 2030 absent renegotiation. Amazon, for its part, has struggled to gain meaningful traction.

In contrast, Google has already released some of the most capable LLMs to market and continues to iterate quickly. It also benefits from a more robust and integrated business model than newer challengers. For example, companies like Perplexity lack a meaningful revenue stream and are heavily reliant on ongoing venture capital funding. OpenAI’s ChatGPT charges a subscription, but its long-term user economics remain unproven. If Google is able to deliver a superior and more affordable product, while also capturing the largest share of user attention, these aggregators may struggle to maintain the financial backing needed to scale.

The Infrastructure Edge Few Understand

One of Google’s most underappreciated advantages in AI is its infrastructure. Advanced AI development is constrained by access to compute power, which is currently being rationed by GPU manufacturers. Unlike most competitors, Google has been designing and refining its own custom processors (TPUs) for decades. These chips give the company greater control over its AI development cycle, enabling faster experimentation and cheaper inference at scale. In the trial-and-error world of model development, compute power is critical – and Google has it in spades.

If Google were to spin out its chip division, it could arguably command a valuation comparable to NVIDIA’s, given its technical depth and strategic importance. Even if the advertising business were to slow, Google’s rapidly growing cloud division, now powered by proprietary infrastructure and increasingly integrated with its AI stack, offers a credible growth offset. Recent quarters have shown accelerating momentum in cloud revenue, further reinforcing the company's multi-engine growth model.

Advertising’s Core Role in Business – Then and Now

To understand why user attention still matters so much, it’s worth stepping back in time. In the era of broadcast and print media, media moguls like Rupert Murdoch held enormous power due to the scarcity of audience attention. Whether through local newspapers or TV stations, media was a game of concentration: limited viewer time meant that only a few dominant outlets thrived in each city. These outlets, with their larger audiences, commanded premium ad rates, reinvested in content, and further widened the gap with smaller competitors. As a result, the industry tended to consolidate into the hands of a few highly profitable enterprises.

Times have obviously changed. What was true of these media companies in decades past is now true of digital companies today. News is now read online and people, especially of the younger generation, connect via social media. Google and Meta, in particular, worked out how to advertise directly to a specific customer. A targeted approach has proved more effective than advertising at large on TV or newspaper. Advertisers flocked to the new approach.

Advertising will always remain a core function of business. Even the best business or product in the world is worthless unless it reaches the minds of potential customers. National advertising, especially in the US, created a lasting competitive edge by allowing businesses to scale beyond their local markets. With centralised operations and broader reach, national firms reduced their cost per unit relative to smaller competitors. As advertising costs rose, larger companies could afford premium placements during major sporting events or prime-time news that smaller rivals simply couldn’t match.

This dynamic continues in the digital world. As Google’s search engine scaled, its algorithm became more effective. Users want fast, relevant answers, and Google’s underlying data gives it an unmatched ability to deliver them. Its model learns what is most useful by observing aggregate user behaviour: when millions of people consistently click on certain answers, Google uses that feedback loop to prioritise results for the next user. By being able to combine its product suite of Maps, Search, shopping and news together into an integrated product ecosystem, we believe Google still offers the best AI solution for what customers will want overall.

A Platform Built to Compound

Critically, Google is more than just a search engine – it’s a platform business. When it builds something new, it can roll it out quickly to billions of users. From driverless cars and video sharing to TV streaming, cloud services, and productivity tools, Google has established leading positions across multiple categories. Its platform structure cuts out intermediaries and enables high-margin scalability. Whether or not AI becomes as transformative as many expect, Google is already positioned to monetise the trend through its infrastructure, distribution, and innovation culture of coming up with new ways to grow and generate more earnings per share.

We believe Google will maintain its market share with time. The company’s advantages in resourcing and providing the customer with a holistic approach to AI will allow it to continue growing earnings per share at a fast rate. Google has, historically, never traded at current valuations. It has never sustained a forward P/E multiple as low as 14x forward earnings. Its TPU chip business alone could justify today’s market capitalisation. Absent a recession, which we no longer forecast, we see Google’s stock in the low $150s as a compelling long-term compounding opportunity.