TMW #210 | The search hydra

Welcome to The Martech Weekly, where every week I review some of the most interesting ideas, research, and latest news. I look to where the industry is going and what you should be paying attention to.

👋 Get TMW every Sunday

TMW is the fastest and easiest way to stay ahead of the Martech industry. Sign up to get TMW newsletters, along with a once-a-month complimentary Sunday Deep Dive. Learn more here.

We’ve come to the end of the month, and your free Sunday essay has arrived. Enjoy! And if you want to dig deeper into Martech, check out TMW PRO. 

The search hydra

Google’s market share dropped below 90% for three months in a row for the first time since 2015. Is it a sign of incoming change in the world of search?

Whenever I write about Google, I like to go back through the annals of history to see what people thought about the company back in the early, halcyon days of the internet. Here’s a quote from BusinessWeek from September 1999, questioning the underlying business model of Google Search:

“But at a time when other popular search sites such as Yahoo!, Excite, and Lycos have all morphed into diversified entertainment portals, is there really a future for a pure, advertising-supported search tool?”

Well, that business model turned out to be a money-printing machine and birthed one of the largest companies of all time. 

Of course I’m being very unfair on BusinessWeek, but in their defense, they weren’t the only ones doubting if Google had the staying power.

As we all now know, Google wasn’t just “an AltaVista killer”: It became a killer of almost every other search engine as it established what we can now officially call an illegal monopoly in the search game.

But judging by the media’s reaction last week, that monopoly is under serious threat! Why, you ask? Because Google’s market share in search has now fallen below 90% for the first time in almost 10 years! Ring the bells, for the Wicked Witch is dead!

The last time that Google experienced three months in a row sitting below 90% market share, it was chalked up to increased competition from a handful of search engines that dominate their respective local markets (Baidu in China; Yandex in Russia; Naver and Daum in South Korea).

Parking my sarcasm about the news media’s reaction to Google dropping market share to below 90% (for reference, the lowest in 16 years was 88.73%), the reason that journos are so excited about this relative blip is because there are valid arguments to be made that things might be different now. 

Not only are AI upstarts like ChatGPT and Perplexity reimagining search using LLMs, but social search via Facebook, Instagram, TikTok and co. is growing rapidly. Plus, product-specific searches tend to start on Amazon. Consumers are searching differently depending on what they are searching for.

The search market is starting to look a lot like the multi-headed hydra of Greek mythology. But can the US Department of Justice manage to cut off the head that it is Google’s illegal search monopoly? Do two AI search startups grow back in its place?

So, with that in mind, let’s explore three future scenarios for search and what they may mean for marketers:

1. Google’s dominance continues
2. AI answer engines take over
3. Search unbundles

Google’s dominance continues

Aaah, the stale taste of the status quo… 

As much as people yearn for a world where Google doesn’t control most of the information economy, there’s every chance that this is how things continue for a while longer. Let’s examine what will determine the probability of this future eventuating.

First and foremost: Regulation is an obvious existential threat.

Google was found guilty in the US in August 2024 of operating a monopoly in the general search and text advertising markets. I explained the presiding judge’s evidence of Google’s monopoly in TMW #187 | Martech Briefing:

In issuing his verdict, Judge Mehta cited a range of different manifestations of Google’s monopoly in search: Exclusionary contracts with phone manufacturers and web browsers that make Google the default; Search’s “remarkably durable” market share; the fact that even the largest Fortune 500 companies have no feasible alternative; and that Google has been able to raise prices on search ads without constraint.

The remedy phase is now underway, with the US Department of Justice (DOJ) seeking sweeping remedies, including a breakup that would force Google to divest its Chrome web browser and Android operating system, banning it from forming exclusive agreements with companies like Apple and Samsung, barring the company from favoring its Search over alternatives on its platforms and devices, and putting strict limits on the company’s M&A activity. 

To make matters worse, other countries are emboldened by the US’s stance on Google and have started to follow suit. Just last week, the UK’s Competition and Markets Authority (CMA) started its own investigation into Google’s dominance in search. It’s hard to see how it could possibly come up with a different answer to the US.

Regardless of all this regulatory action, there is every chance that the remedies proposed in the DOJ case against Google don’t make much of a difference. Platformer summed up the likelihood of each remedy as such:

• Certainty: Google will continue to benefit from its search monopoly through a lengthy appeals process.
• Very likely: Google will be banned from making deals with software and device developers for default placement of Search.
• Somewhat likely: Software and device developers will show choice screens to allow users to select their default search engine.
• Less likely: Forced divestitures of Chrome, Play Store, and Android.

It could require all these remedies in conjunction to really breathe new life into the search market, which seems unlikely to happen given the historical precedent around breakups, which have only happened three times in the last 114 years. The DOJ’s vision is that Google be forced to sell off Chrome and Android so that it can’t self-preference Search, and that a deep-pocketed competitor develops a search product to properly challenge it.

Even if all those things happen, there is an argument that search is a natural monopoly due to the network effects that constantly improve the most popular search engine’s results at a faster rate than competitors. So, although regulation is an existential threat to Google’s dominance, it would take something truly unprecedented for it to bring Search crashing down.

The other key challenge for Google comes from AI chatbots like ChatGPT and Perplexity. Both are putting a new spin on search that uses AI and a conversational interface to reimagine how consumers search for information on the internet. 

Although these AI “answer engines” have been attracting most of the buzz, Google still has massive advantages over these upstarts.

Firstly, Google likely holds the most data of any company in the world, with a massive web of interconnected consumer tech products. This means it has the most data for training LLMs, which don’t forget is a technology that Google researchers pioneered with their seminal paper on transformers.

Secondly, Google’s massive existing user base gives it an unfair head start. The company can wait to see which way the wind is blowing in terms of consumer take-up of AI search, and then quickly deploy that approach, as I explained in TMW 208 | Wednesday Martech Briefing:

“Google is rolling out AI chatbot features to compete with these upstarts, with reports claiming that the company will release an AI Mode for Search that closely resembles its Gemini AI chatbot… Rolling out AI Mode as a toggle within the core Search product is a clear sign of Google’s intent to use its 4.97B user base to continue to dominate in the age of AI search.”

Unless the DOJ can force through a wholesale breakup of Google, then it will still have all the advantages to continue its monopoly into the age of AI search.

Rise of AI answer engines

ChatGPT was the application that stoked the collective imagination into considering a new way to search the internet. Since its launch in November 2022, a slew of similar competitors and products have launched, all with their own ideas about how to improve search.

OpenAI clearly has its sight set on taking on Google, with the launch of its SearchGPT product and rumours that it is developing a browser to compete with Chrome. 

But how can an upstart like OpenAI take down Search when Google has so many advantages?

The answer can be found in Google’s history. Here’s a quote from TIME Magazine’s list of the top tech products of the year in 1999:

“With sites such as Yahoo, Infoseek and Excite constantly beefing themselves up into the online equivalent of mega-malls, it’s refreshing to find a search engine that does nothing but search. And search well. Google’s award-winning, commonsense approach nearly always seems to come up with exactly what you’re looking for.”

Google overtook Yahoo!, Infoseek, Excite, AltaVista, and every other search engine, by being a good product first and foremost. It was simple, reliable, and relevant.

Nowadays, the company struggles with its product quality. The relevance of search results has degraded as the internet has been flooded by SEO-optimized spam and AI-generated slop. Its initial launch of AI Overviews in Search was a classic own goal, as it suggested that users use nontoxic glue to thicken pizza sauce. Although not strictly related to Search, Google’s launch of its Gemini AI image generator was marred by inaccuracies such as depicting racially diverse Nazis.

AI answer engines have the benefit of starting with a blank slate to produce a better product, and there are some early indications that they are doing so...

Incentives reveal the real shifts in this space.

Google’s incentive is to maintain the status quo, which I explained in TMW #160 | Wednesday Martech Briefing:

“There is a rather large elephant in the room at Google headquarters and it's called self-cannibalization. GenAI is both an existential threat and a massive opportunity, but to fully seize the moment, Google must eat its golden goose, Search. Google’s current strategy in the face of this threat is to build AI-powered features around the fringes of its core search product rather than swallowing the golden goose whole.”

Pureplay AI search products don’t have the innovator’s dilemma to contend with, but they must find a way to commercialize their products to absorb the massive cost of training and querying their LLMs. With traditional search, each additional query costs nothing to Google and adds additional value to Google’s algorithm. This is not the same for LLMs where every additional query has an incremental marginal cost that needs to be covered.

Although many AI search tools started with a subscription model, all of the major players – OpenAI, Perplexity, Google – are now looking towards an ad-supported business model. How they will do that is another question altogether. Google Search’s ad model matches keywords directly to advertising opportunities; this is far more obfuscated and complex in a scenario where AI is churning out the response to queries.

Although this might sound like a reason that AI search engines won’t overtake Google Search (Sequoia describes it as “AI’s $600B Question”), it can also be seen as an opportunity… Whichever company – if any ever do – best works out how to enable advertising in AI search, they stand a good chance of achieving mass adoption.

But it is a narrow path. As Cory Doctorow argues, given the investment levels in LLMs required to make AI search a worthwhile venture, it requires a Google Search-sized opportunity. There may only be enough opportunity for one company to profit enough to survive. 

So, if AI answer engines can prove they are better products than Google Search and work out an ad-supported business model to scale and capture the market – perhaps with a helping hand from the US DOJ – then maybe they can be the future of search.

But given the narrow path and level of investment required to make this future a reality, would this be a case of jumping out of the fry pan and into the flames with another search monopoly? LLMs don’t really have a competitive moat except for exclusive training data deals so free cash flow becomes the determinant of which company can build and operate an AI search chatbot long enough to outlast its competitors. 

Ominously for the AI upstarts, Google’s total free cash flow over the last five financial years is $270B… straight out of the fire and into the furnace?

Search unbundles

When we talk about search, we tend to talk about Google Search and its market share against competitors like Bing and Yahoo!. But that is a narrow view to take when searching on the internet doesn’t just happen in search engines; there is a plethora of other platforms where users are searching for information and products.

Take social search for example. There are billions of users on social media sites searching for ideas, inspiration, information, products, recipes, new music – you name it. In fact, younger generations are increasingly starting their searches on social media rather than with a search engine.

Even within social media, the picture is fragmented by what the user is searching for and their age. Gen Z favors using TikTok for topics like hair, makeup, and gift ideas. The most popular search category on Instagram is fashion brands. Facebook users frequently search for news and current affairs, recipes, and meal ideas.

If we zero in on product searches, the picture is different again. The vast majority of product searches start on Amazon, although social media has been eating into Amazon’s lead over the last few years.

Part of this is simply a function of time spent on social media platforms, which is particularly pronounced for young people. Just look at the daily screen time breakdown for US teenagers: Browsing websites still gets 51 minutes per day, but that’s tiny compared to the combined 6 hours, 29 minutes they spend watching TV/videos, gaming, and scrolling social media.

This points to something that is not visible in the graph tracking Google’s dominant market share in search. Google may still have a 90% share of traditional search, but increasingly people are going to non-search engines to meet the same needs.

This is primarily a generational shift, which will over time see social media and ecommerce marketplaces like Amazon and Temu become the dominant first-port-of-call for social media natives like Gen Z and Gen Alpha when they are searching for new products or information. There’s a reason that social media companies are rolling out ecommerce capabilities (TikTok Shop, Facebook Marketplace, Instagram Shopping)… they see the opportunity to dominate the lucrative business of connecting internet users with products via ads.

The key point of difference with Google Search is that this future is distributed. There won’t be a single point of entry to search for information; instead, users will start in a different location depending on what they are searching for.

We have all subconsciously started doing this. If I’m buying cheap homewares or something I need quickly, I’ll go straight to Amazon. If I’m searching for a recipe, I’ll skip straight to Instagram. I always start with Perplexity when I’m researching for an essay.

Internet users are a lot more sophisticated now than they were when Google first started in the 90s. They know which site will get them to what they need the quickest, and which is the most reliable. That is why distributed search may be the future.

The marketer’s dilemma 

Ok Keanu, this is all great, but what does this mean for me as a marketer? 

I’m glad you asked!

If Google is destined to continue its monopoly, then the status quo is maintained for marketers. Search engine optimization (SEO) will be primarily geared towards Google’s algorithm and search engine marketing (SEM) budgets will be concentrated there too.

This means that the platform risk continues for marketers. One day you may have plenty of organic traffic, the next it might crater due to a core algorithm update. Look at the case of HouseFresh – a specialist air purifier review website – that went from being a thriving business to laying off most of its team within a few weeks after a Google algo update.

The big outstanding question is what happens if Google goes all-in on AI search.

If SEO and SEM isn’t as simple as matching keywords to sites and ads, how will marketers maintain predictable discoverability? 

How will they control the narrative around their brand and products if Google is obfuscating this via GenAI summaries? 

How will they comply with false advertising and product advertising laws when the AI hallucinates? 

How do they protect and maintain their brand when even Google isn’t sure what content the AI will spurt out next to your ad?

This is equally a challenge in the scenario where AI answer engines like ChatGPT and Perplexity become the dominant form of search. The other risk for marketers in this world is volatility.

For all the problems with Google Search’s dominance, there has at least been relative stability in terms of the cost of SEM (although this has increased steadily over the years). AI-native search is still looking for the right business model and given the costs these platforms need to recoup and cover moving forwards, there’s every chance that ad pricing bounces around as they find the right fit.

If I had to bet on one future, it would the final option: search unbundles. This outcome will make a lot more work for marketers.

Every piece of content will have to be optimized for a range of different platforms to ensure discoverability. But the optimization will look very different depending on whether the platform produces AI-generated results or not. Each of those platforms will then have to be monitored, reported on, and compared with each other. 

The advertising component is already quite fragmented across traditional search, social media, and where appropriate Amazon, so at least that part is already accounted for in terms of the effort marketers will have to expend.

The story of the internet is one of never-ending expansion of channels for consumers to find your brand. Google Search has been the central pillar of this search ecosystem for a quarter of a century, and it still dominates the traditional search market.

But search has gone way beyond traditions. Consumers have come a long way since 1999. In that way, despite Google’s dominance, search is already a hydra. If one head is cut off, two more spring up. Some of those heads will be AI search engines, some will be social media platforms, some e-commerce platforms. But the through line is that if search does unbundle, life gets harder for marketers as they manage more and more channels.

🔓 Continue reading full length deep dives every week with a free TMW PRO subscription

Get the full versions of upcoming Sunday Deep Dive and all of our newsletters every week with TMW PRO. It's the easiest and fastest way to make sure you're fully informed on how the marketing technology industry is evolving. 

For a limited time, you can earn 12 months of TMW PRO Advantage for free. Qualified brand side practitioners who complete a short survey and 1-hour interview with our knowledge partner Alium will receive 12 months of TMW PRO Advantage (a US$499 value) for free. CLICK HERE to get started. 

Here’s what some of our existing TMW PRO members have to say; 

“TWM is a finger on the pulse of what's happening, with the clarity of insight to see the important patterns underneath the noise” – Scott Brinker (VP, Hubspot).

“I recommend TMW to all leaders who are customer-focused and transforming companies via digital technologies” – Kazuki Ohta (CEO, Treasure Data).

“I've subscribed to many Marketing newsletters, but TMW is at the top of my list with the most valuable content for marketing professionals" – Lilly Lou (Marketing Director, Icertis).

TMW PRO is an invaluable resource that’ll help you to truly transform and future-proof your career in Martech. 

CLICK HERE to get 12 months of TMW PRO Advantage for free

CLICK HERE to purchase TMW PRO

Stay Curious,

Keanu Taylor

Make sense of marketing technology.

Sign up now to get TMW delivered to your inbox every Sunday evening plus an invite to the slack community.

Want to share something interesting or be featured in The Martech Weekly? Drop me a line at juan@themartechweekly.com.