Google and the Innovator's Dilemma

For the first time in decades, Google’s grip on search looks vulnerable. AI-native tools are reshaping how consumers search, and according to new details from Google’s antitrust lawsuit, they appear to be moving faster than Google can keep up.

The Innovator’s Dilemma and the Search Disruption

This kind of disruption is reminiscent of Professor Clayton Christensen’s 1997 book The Innovator’s Dilemma, a staple in startup reading lists that feels especially relevant as Google, the undisputed incumbent in search, begins to see its search volume decline.

Christensen argues that market leaders are often so focused on serving their most profitable customers and refining existing products, what he calls sustaining innovation, that they ignore new, lower-margin alternatives. Over time, those disruptive technologies improve, gain traction, and eventually displace the incumbents.

For consumers and early adopters of novel technology, this means new and exciting tools to improve how we work, play, and learn. For incumbents like Google, it means we may be watching Christensen’s theory unfold in real time. The innovator’s dilemma.

Cracks in the Foundation

Google has enjoyed a search monopoly for the better part of two decades. It is, and has been for quite some time, the default search provider. So much so that it has quite literally become a verb. To “Google” is the original grabbing a Kleenex or ordering an Uber.

Google’s search business has come under increasing pressure with the proliferation of LLMs, and even more so with the rapid development of AI agents. But fuel was added to the fire when a rather interesting detail came to light amid Google's ongoing antitrust lawsuit. Google has been paying Apple $20 billion a year to remain the default search engine on Apple devices. As part of the DOJ’s investigation, Apple’s Senior Vice President of Services, Eddy Cue, testified that Google searches through Safari, the default browser on iPhones, have declined over the past two months. He noted that this hasn’t happened in more than 20 years. Cue attributed the drop to a growing shift toward generative AI tools like ChatGPT and Perplexity, with more users turning to these services over traditional search engines. Bloomberg published an article about the testimony at 11 a.m., and by noon, Google had lost $100 billion in market cap.

Search generates over $200 billion annually and accounts for nearly 60 percent of Google’s revenue. Unlike cloud services, another large part of the business, or AI queries like those run by ChatGPT or Perplexity, traditional search remains an extremely high-margin business.

A Strategic Crossroad

So what is Google to do? Stay the course and preserve its high-revenue, high-margin, cash cow business? Or respond to newer, lower-margin technologies serving smaller or emerging markets? It is a textbook case of the Innovator’s Dilemma.

Google is in an interesting position: flush with cash, completely unlevered, and arguably possessing some of the best models. Gemini 2.5 Pro outperforms peers across several benchmarks. But this isn’t about performance or raw compute. This is, we believe, a distribution problem. Google already has the platform and the capital to compete in the AI arms race. The company is ramping up its investment in cloud and AI infrastructure, with plans to allocate $75 billion toward expanding capacity in 2025. The real question is, is search a race to the bottom? And is it even a race they want to win?

Frequent users will have noticed that Google has already added a box at the top of search results for certain queries with an “AI Overview.” Google is well aware of the paradigm shift taking place in search and is now trying to walk the same line all organizations face when redefining their distribution strategy. What channels should they leverage, and at what scale? They know the traditional search business carries meaningfully higher margins than AI queries, and they know some consumers aren’t looking for an AI-generated answer to when their local pizza place closes. It’s the same fundamental problem every organization faces: how to serve customers through the most efficient channel while meeting them where they actually prefer to engage. The challenge is doing so without getting ahead of user sentiment, without falling behind the curve, and without cannibalizing the old, still massively profitable business.

Three Paths Forward

1. Go All In on Gemini for Search

One school of thought is that Google should immediately push Gemini results to the top of every search query. Supporters of this approach argue that the market has spoken. People prefer AI-powered search over traditional search. If Google waits any longer, the problem will get worse, and what remains of its competitive moat in search may disappear. This approach would get the wildly powerful, and often best-in-class, Gemini models directly into consumers’ hands, but it would cannibalize Google’s existing search business. It is a bold strategy that would fundamentally change how Google’s search business operates.

2. Keep Search Intact, Separate Gemini

Another option is to keep traditional Google Search as is and offer Gemini as a standalone app, similar to how OpenAI offers ChatGPT. This separation could allow Google to maintain its search business profitability while competing with generative AI-powered search applications. Search would continue to serve users looking for links, not direct answers, and remain a high-margin business. Meanwhile, the Gemini app, which is currently far less popular than OpenAI’s, could compete head-to-head with ChatGPT and Perplexity. While this seems like a sensible option, it becomes more complicated when you consider the internal dynamics and operating models. Different executive teams with their own P&L and incentives pushing towards different outcomes. While Google has a history of managing large-scale transformations, this scenario would be a cultural test. It would require the company to make hard decisions about how to prioritize competing business units with overlapping capabilities.

3. Reposition Around Other Business

At the other end of the spectrum, Google could maintain the status quo with search and shift its focus toward other strong businesses already within its portfolio. Waymo and YouTube are excellent examples. Waymo being AI-native and has the potential to be a generational company at the forefront of human mobility. There is a case to be made that the knowledge worker economy, and the broader search market, is turning into a race to the bottom. Google, after two decades of dominance, may choose to step back from that race. If the cost of compute remains high and AI queries do not benefit from a dramatic Moore’s Law like cost curve, it may not be a market Google wants to prioritize.

While trying to remain impartial across all three scenarios, it is admittedly hard to imagine Google committing $75 billion in capex to AI and infrastructure without a serious intention to defend its position in search. It is hard to see a world where Google simply walks away. But if the company gradually shifts its focus toward other category-defining businesses, like cloud, YouTube, or Waymo, it may unlock success outside the increasingly competitive low-margin world of AI-powered search. Take the recent acquisition of Wiz (covered in a previous Ventures Edge article here) as one clear signal that Google remains bullish on the future of its cloud business, as an example.

Google’s Real Advantage

The X factor in this is Google’s overwhelming data advantage. Should they decide to compete in AI search, the ability to integrate with Gmail, Docs, Calendar, YouTube watch history, and other products creates a uniquely powerful moat. With access to this ecosystem, Google can orchestrate the user experience in a way that has the potential to be a step-change improvement over what is currently possible with Perplexity or ChatGPT. For instance, imagine a world where you are flying out to meet with a founder for a pitch.

You receive a flight confirmation in Gmail, save the founder’s pitch deck in Drive, and add the meeting location to Calendar. You’ve been reviewing industry trends on YouTube and reading articles on Chrome about the company’s latest product launch. The founder also sent a follow-up form through Google Forms, which is automatically stored in your Drive.

On Monday morning, you ask Gemini:

“What do I need to prep for the pitch this morning?”

Gemini responds:

“You’re flying into Toronto at 10:10 a.m. and meeting with the company at their office at 2:00 p.m. The address is already in Google Calendar and synced with Maps. Their product launch last week was covered in two articles you opened in Chrome. You also saved their supplementary deck in Drive and watched a YouTube breakdown of their industry trends. Would you like me to summarize the deck, draft a few talking points in Google Docs, and send a confirmation email through Gmail?”

More than a search result or AI-generated response, it’s a personalized, contextual decision engine. This kind of experience is only possible with Google's integrated data ecosystem. Gmail, Calendar, YouTube, and Search working together in a way that siloed apps without the data assets that google has currently can’t compete with.

What Comes Next

Search may change, and Google may no longer win, but as consumers, operators, and investors, it is a tremendously exciting time to be part of the innovation ecosystem and witness a textbook case of the Innovator’s Dilemma play out in real time. Google is facing the exact challenge Christensen described, balancing the protection of a massively profitable core business with the urgency to embrace disruptive technologies that may one day replace it. Whether Google leads the next era of search, evolves around it, or shifts focus to other already established businesses, we are watching a generational company navigate a generational shift. The outcome is uncertain, but the implications are significant for consumers, operators, and investors alike.

As we wrap up this piece, there was a moment of reflection. The first place we went when researching this article was the browser-based ChatGPT, not Google (which we individually pay a premium subscription for). Is this a sign of the times? We shall see.