Google Search and the Innovator's Dilemma

Google was able to transform its core product, and avoid a strategic threat, during the shift to Mobile; now it's trying to do the same with AI.

From Stratechery, following Google’s recent earnings announcement:

… the biggest tell of all that Google very much sees AI as a threat to their core business is that that is the one area the company really is desperately trying to figure out their AI product, and moving fast. The race is between incorporating their own AI into Search before someone else’s AI disrupts Search, and it’s one of the most compelling races I’ve ever watched.

I agree that it’s fascinating to watch how Google is racing to transform its core business, and maintain its dominance in the AI era; that is why I think it’s a good opportunity to revisit the tale of how Google Search was successfully transformed in order to avoid similar threats posed by the shift to mobile. It’s an interesting story, and one I was able to experience while I worked for Google Search.

Peak Google?

[,,,] These eclipses are obvious in retrospect, but the truth is few if any could have predicted them before they occurred. PCs were thought to be a tremendous boon for IBM, and they did profit greatly until Compaq copied their BIOS leaving Microsoft all the leverage; similarly, Microsoft looked set to conquer mobile (which was why Google bought Android in the first place) before a resurgent Apple introduced the iPhone. In both cases it turned out that the incumbents’ prior success resulted in misdirected incentives: IBM focused on selling and servicing PCs, instead of building a platform, while Microsoft focused on extending Windows to mobile instead of the user experience. If you’ll forgive a war analogy, both companies won the battle but lost the war.

And so, if one wishes to predict who might follow in this illustrious but ultimately tarnished path, it might be useful to look for similar characteristics: the company should be dominant in its field, and the company should seem to have an advantage in a far larger adjacent field, but that advantage, on closer inspection, should prove to be just as much a hindrance as a help.

The clear candidate is Google.

It may sound like an argument written today about the challenges posed by AI, but Ben Thompson actually wrote these words over a decade ago, in an October 2014 article titled Peak Google; even though the prediction did not hold up – as Google quadrupled its ad revenue over the recent decade, and still commands the largest share of the digital advertising market in 2025 – the analysis was spot-on.

The smartphone indeed posed a major threat for Search: the act of typing queries into a text box and clicking on links is a very desktop-suited experience, optimized for a large-screen-with-mouse–and-keyboard setting; on mobile, however, native applications offered a superior experience. Google Search was the best tool for many tasks on desktop – such as checking the weather forecast or looking up stock quotes – but on mobile, native apps may make more sense for the same use cases. Advertisers' budgets would follow, making mobile app advertising eclipse the revenue Google is making on search advertising.

Search was about to face the same destiny as Microsoft Windows; a once-dominant platform of the PC era, eclipsed by the rise of mobile. As Ben Thompson explained it:

(source)

Brand advertising is worth a lot more than search advertising; if it moves to the Internet, Google’s share of digital advertising would be dwarfed

This is the primary basis of my thesis that Google may very well be in a similar situation to early-eighties IBM or early-oughts Microsoft: a hugely profitable company bestride the tech industry that at the moment seems infallible, but that history will show to have peaked in dominance and relevancy.

Thompson wasn’t wrong! The threat was real, it’s just that Google was able to evade the destinies of early-eighties IBM or early-oughts Microsoft by completely transforming its Search product.

Knowledge Panels: "mini-apps" within the Google Search results page

Google effectively built a huge collection of information boxes. Essentially, they were mini-apps of sorts, embedded within the search results page. They offer the kind of experience you would expect from a native mobile app, and so many users never had to download standalone apps for weather or stock quotes. They’re effectively included within the Google Search App¹. That’s how a large portion of the native ads section of Ben Thompson’s chart above, ended up as Search Ads². These information boxes were named “Knowledge Panels”, and back in October of 2014 – when Ben Thompson published Peak Google – that’s the project I was working on at Google.

Disruptive Technologies and Incentives

“Luckily for me, I don’t have to worry about the money.”

That was a surprising statement coming from a VP of Engineering. It’s been over a decade since, so I feel comfortable sharing this story: during an all-hands meeting, the VP overseeing the development of Knowledge Panels invoked a sense of urgency – by explaining how old Google Search was losing relevance in mobile – and highlighted how critical the knowledge-panels initiatives we were working on were for the future of the company. She set pretty ambitious goals in terms of the share of Google Search traffic she expected to be taken over by these new experiences.

During Q&A, one engineer asked whether this might negatively impact Google’s revenue; we were essentially working on directly displaying an answer for whatever the user was asking, thus bypassing their need to click on links. If they don’t click – Google wouldn’t make any money from sponsored links. I remember thinking, “yeah, that might be a big problem!” Curious to hear the executive’s answer, she just dismissed the concern. We don’t need to worry about the money.

That was very strange; some of the folks in the meeting may have admired her dedication to serving users’ needs above all. But. All of us held Google stock grants. And, as a Vice President in the company, she must have had more shares than anyone else in the room. “shouldn’t she care about the money the company was making?” I remember thinking to myself.

It wasn’t until I read The Innovator’s Dilemma that I was able to figure out the mystery: Professor Christensen explained:

Established firms that successfully built a strong market position in a disruptive technology were those that spun off from the mainstream company an independent, autonomously operated organization ... [they] all succeeded because they created organizations whose survival was predicated upon successful commercialization of the disruptive technology: These firms embedded a dedicated organization squarely within the emerging value network.

There is a great Jeff Bezos story that illustrates this point, told by Dan Rose on the Invest like the Best podcast:

Steve Kessel was asked by Jeff Bezos in 2004 to start up this new division. And Steve, at the time was running the entire media business at Amazon. He was running the books, music, and video business, which was the largest business by revenue, but even more importantly, the books business alone was the vast majority of Amazon's profits at the time. And Jeff had seen the iPod come out and decimate our physical music business and had the recognition that the same thing was going to happen to books.

And if that was going to happen, we better be the ones to do it, not someone else. He said to Steve one day, “Steve, I need you to come over and run this digital business and get this digital book platform started so that we don't get iPoded out of books”. And Steve said, "Great, I'll take one of my best people. We'll put them on it, and we'll get a team going, and it will be great”. And Jeff said, "No, you don't understand. I want you to do it”.

And Steve said, "But perfect, I'm excited. I'm fired up. Let's go build this. I'm going to put this person who I think is the best executive in Oregon, and we're going to have him go build a team”, and Jeff goes, "No, Steve, let me make this clear. As of today, you're fired from your job. Your new job is to kill your old business. I want you to put the physical books business out of business by building a digital product that's so good that people don't buy physical books anymore. If you run both, you'll never be motivated to do that”.

The mentality Bezos had around the Kindle and the physical books business, is how I interpret the not-needing-to-worry-about-money statement: bonus targets for the VP in charge of Knowledge Cards weren’t tied to the company’s ad revenue; rather, her incentives were defined based on the impact and success of knowledge cards. I don’t have any evidence to back this up – as one of many software engineers working on this project, I wasn’t really exposed to any executive compensation packages – but it seems to make sense. Just like the Bezos Kindle story, my guess is that she was told that her job was to kill the ten-blue-links in Google Search results page.

That, alongside initiatives like Voice Search, Web Answers, and PageSpeed, kept Google Search relevant in the age of mobile. According to a 2022 study by Semrush, the monetization rate is far lower on mobile – 0.02%, compared to 1.8% on desktop³ – but there are just so many more queries users make through a device that’s always in their pocket.

source: semrush.com

It’s worth noting that only 43.12% of mobile searches lead to the user even leaving the Google Search results page to an external website (either through an organic or a paid link). In the remaining 56.88% of cases, Google works hard to satisfy the user’s need, but doesn’t make any money. As MBI Deep Dives pointed out on his Alphabet Deep Dive:

Of course, Google doesn't make money when users don't click, but Google correctly identified it is far more important to keep your query share and monetize the search habit over time.

Similar to how Costco is losing money on the $1.50 hotdogs, but gaining loyalty – as customers leave the store happy, come back more often, and renew their memberships – Google was fine with essentially losing money on roughly 99% of queries, since it kept customers coming back. And it was able to make so much money out of the tiny fraction that was being monetized.

That’s how Google Search escaped the Innovator’s Dilemma paralysis during the shift to mobile – as management was smart enough to set the right incentives for teams transforming the core search product – and it may be running the same playbook today.

AI Search and Cannibalization

Could history repeat itself?

That is probably the key question keeping GOOG bulls and bears busy. Google’s stock was depressed throughout 2014 and the first half of 2015, when the company reported that mobile search traffic was a key driver of ad revenue growth. The stock is somewhat depressed again today, due to concerns around AI, and Google’s current strategy is very much reminiscent of what it did with mobile: Google is – again – transforming its core product. The current equivalent of Mobile’s Knowledge Panels is Generative Search Experience (GSE) – things like AI Overviews, AI Mode, Google Lens or Circle to Search – which aim to keep Google Search relevant in the age of AI.

I no longer work at Google, and have no idea if engineers working on GSE are again told not to worry about the money. One interesting data point, though, is that Elizabeth Reid, a VP that was leading Google’s GSE (“Search AI”) efforts, was promoted last year to manage all of Google Search. Reading between the lines, it seems as though – just like with Mobile – Google is trying to go all-in on the new paradigm, without being paralyzed by the Innovator’s Dilemma.

In a recent interview with the Financial Times, Reid was asked specifically about the risk of cannibalizing existing search revenue:

Interviewer: Is there a risk that you end up cannibalising your own product? Generative search is expensive, and this is changing the whole ad revenue model.

Elizabeth Reid: There are a lot of opportunities for ads. We show them both above and below in AI Overviews, but also within. Ads are relevant whenever users are going to make a choice that has some commercial aspect.

When a query is predominantly commercial intent — like we think you want to buy something — then we might often show ads. But sometimes we think you probably don’t want to [see] ads, and so we don’t want to give everyone ads. But some people might want to buy something. If [you search] “how to clean a stain out of the couch” and the first thing we show is a bunch of ads, you’re like, “Whoa, I just wanted some advice.”

But if we’re giving you ideas and then we say, “if you’re having trouble you might want to consider a stain-remover product”, and then we give you some ads for stain-remover products, it feels natural and in context. And so, there are new opportunities.

With AI, Google is putting out the metaphorical money-losing $1.50-hotdog stand, once again, hoping for a payback in terms of user loyalty; the assumption is probably that if customers keep coming back (big if) – rather than switching to an upstart “store” like ChatGPT across the street – the Ads team would figure out a way to monetize some fraction of their queries.

The same exact mindset had worked out phenomenally well for Google back in the mobile days; it remains to be seen, however, whether it can lead to a successful outcome, once again, with AI. In either case, it’s going to be fascinating to watch.

---

Disclosure: Nothing here is investment advice. I own shares of Alphabet, and may sell at any time without notice.

1. Or the Google Search website, for mobile users who are searching through their browser.

2. It’s not the only reason: they’re another fascinating nuance around how the internet completely changed the nature of brands (which Ben Thompson named “the disruption of everything”). I’ll leave it for a different article.

3. The numbers might not be accurate, but the gap is so wide, that even if they’re just directionally correct: the point still stands.