- 16 hours ago
This is a unit-economics breakdown: ad-funded Chrome vs AI-native search like Perplexity.
3 years ago, Perplexity was hyped as the browser that will destroy Chrome. Fast forward to 2025, Google Search still dominates with 8.5 billion queries per day, while Perplexity processes a fraction of that volume.
In this episode of AI Hype vs Reality, I break down the economics and scale capacity for AI-native vs traditional search.
Chapters:
00:00 The Google replacement hype
00:26 Economics of AI-native and Ad-funded search
03:30 How much Perplexity needs to grow to match Chrome
03:53 Why Chrome dominates the market
04:30 Perplexity's weaknesses
05:19 How ChatGPT helped Perplexity
06:13 Perplexity’s Revolution
06:46 Why Perplexity didn't beat Google
09:05 Google's Antitrust Ruling
11:38 Bidding Wars over Chrome
13:33 Conclusion
3 years ago, Perplexity was hyped as the browser that will destroy Chrome. Fast forward to 2025, Google Search still dominates with 8.5 billion queries per day, while Perplexity processes a fraction of that volume.
In this episode of AI Hype vs Reality, I break down the economics and scale capacity for AI-native vs traditional search.
Chapters:
00:00 The Google replacement hype
00:26 Economics of AI-native and Ad-funded search
03:30 How much Perplexity needs to grow to match Chrome
03:53 Why Chrome dominates the market
04:30 Perplexity's weaknesses
05:19 How ChatGPT helped Perplexity
06:13 Perplexity’s Revolution
06:46 Why Perplexity didn't beat Google
09:05 Google's Antitrust Ruling
11:38 Bidding Wars over Chrome
13:33 Conclusion
Category
🤖
TechTranscript
00:00There were a lot of people who called Perplexity the Google killer three years ago. Fast forward
00:04to now, Google search is doing just fine. The core problem that Perplexity faces is that they built
00:10a Ferrari for a world that needs Corollus. The math, the engine economics, and even Google's
00:15antitrust ruling all point to the same thing. Perplexity still cannot beat Google search.
00:21Let's break down why. Episode 7 of my series AI hype versus reality. Let's go.
00:25Just two years ago, Perplexity AI processed 42 million queries per month on average. And by
00:32October of the following year, the number grew to 400 million, a thousand percent growth in under a
00:37year. A very solid number. Solid, but microscopic. Compared to Google's 8.5 billion queries per day.
00:46These numbers for a startup that managed to make headlines as a potential Google rival,
00:50something nobody has done in decades, is phenomenal. But there is a fundamental difference,
00:55an enormous scale gap between the economics of AI native search and an ad funded search.
01:02This isn't even apples and oranges. This is apples and an entire orchard. What do we need to know about
01:08AI native search and an ad funded search? Let's talk about Google. Google's advertising revenue reached
01:15237 billion dollars, which is a 6% YOY increase, which is the second smallest growth rate in two
01:25decades. Once again, almost 240 billion dollars is their second smallest growth rate in two decades.
01:34In 2025, this ad revenue went up to almost 300 billion dollars. Ads are a gigantic revenue stream for
01:41Google. It makes up 77% of Alphabet's total revenue and translates to approximately 700 million dollars
01:49earned daily from advertising alone. Combine that with YouTube ads and Google Network and you'll realize
01:57that search ads aren't a side business for Google. They are the business. Now let's talk about perplexity.
02:03Perplexity doesn't have any of this infrastructure. Perplexity's annual projected revenue for this year
02:09is 147 million. Compare that to Google's daily revenue of 780 from advertising alone. Perplexity's
02:16active user base is estimated at 25 million users, generating 780 million monthly queries as of today.
02:24It is estimated that the revenue that perplexity makes per each user is $4.50 annually.
02:31Google's estimated revenue per user per year is $80 globally. It is higher in North America,
02:37but on average is $80. Now let's talk about the mechanism of search. At Google, it's keyword
02:42matching. At Perplexity, it's neural information retrieval plus LLM. Perplexity's conversational
02:48interface requires significantly higher computational costs than Google's traditional keyword matching.
02:54Because for each response, Perplexity's AI search combines LLM with real-time synthesis from
03:00multiple sources, whereas Google simply retrieves the link. Now let's talk about the unit economics of
03:06an AI search versus ad-based search. If we factor in Perplexity's infrastructure costs for processing
03:12natural language queries and generating cited responses and pair it up with their clean and
03:17beautiful UX, there is little to no room for advertising. In contrast, Google's seamless ad placement
03:24within search results generates a ton of money for them automatically. Perplexity is currently valued
03:29at $18 billion, and that's a double of their valuation since December 2024, and it does show
03:36investor confidence in long-term disruption potential. But current Perplexity's revenue
03:40would have to grow 2,500 times to match Google's advertising revenue alone. That is the cost of scale
03:48for an AI-first search versus traditional search. Chrome's dominance gives Google such a massive
03:53distribution advantage that it dictates the search behavior that we as users adopt, and it makes it
03:58brutally difficult for competitors like Perplexity to win those users. Chrome's dominance is more than
04:04just a market share. It allows Google to integrate search directly into the browsing experience. And the
04:10browsing experience is such that the address bar doubles as the search interface. And when you open a new page,
04:16you get all of the Google services at once. They deliberately integrate Chrome with the rest of
04:21Google's ecosystem. So much so that even superior AI search must overcome ingrained user habits.
04:27And as you know, old habits die hard. Perplexity, on the other hand, is well aware of its weak
04:32modes. Perplexity does not have its own proprietary database of web pages like traditional search engines.
04:38Only about half of its users pay for the pro plan, and 50% of the paid users either come
04:44from trials,
04:44promotions, referrals, or bundles, meaning Perplexity doesn't get the full revenue out of those subs.
04:50That fragility explains why Perplexity team keeps experimenting. They introduced this feature called
04:55Pages, they just launched their own browser extension, and they even bid to buy Google Chrome.
05:00Keep watching till the end to find out what happened. Now, a few words on Perplexity's browser. On one hand,
05:06you may argue that Chrome's dominance is so entrenched that nobody can seriously erode it. But Perplexity's got
05:13quite a bit of an advantage. A fairy godmother, if I may. The one and only ChatGPT. And you guys
05:19know why.
05:20ChatGPT was the product that changed the search behavior. A shift that nobody was able to pull off
05:26in decades. The thing is that traditional search began as a way to return lists of sites. Over time,
05:33they layered on extensions, widgets, services, shopping, booking, knowledge panels, and all of them were
05:39designed to cut down the user's path from query to solution. But ChatGPT has removed the need for the
05:45user path entirely. Large language models began producing answers right away. And that opened our
05:51eyes on a completely different search behavior. People realized that they no longer have to look
05:56for the information or for the answer. They could get instant answers without clicking at all. But
06:01LLMs come with three unavoidable flaws. They hallucinate, their stored knowledge capacity is
06:07limited, and they're not trained on the most up-to-date information. So what Perplexity and
06:12products like Perplexity did is that they married the relevance of the information on Google to LLM
06:18behavior. What they did is they built a conversational search with beautiful UX. By the way, they were one of
06:24the first AI apps to demonstrate how LLMs do reasoning, which planted that credibility in our eyes because it
06:30looked like Perplexity was thinking. And all of that was paired with very recent news and the most
06:36up-to-date information. Now, Perplexity does hallucinate every now and then, but the UX of the app
06:40almost made it seem like it doesn't. So why didn't the Google killer moment actually happen? It didn't
06:47happen because not all searches are made with the same purpose. What we love doing in this channel is
06:52playing product management 101. So let's do it again. Why do people search in the first place? There are two
06:58primary buckets of search queries. The first one is to get direct answers. What's tomorrow's weather? Who won
07:05the game? What's the meaning of this word? For these questions, Google is an obvious choice because it's
07:10simply faster. Now, if you plug any of those questions into Perplexity, it'll start fetching data, comparing
07:15sources, giving you pros and cons, summarizing trade-offs, etc. Whereas Google just serves the result. And the second bucket
07:22is when people are working on a task or on a question that has caused a relationship when
07:27they're doing research. In other words, when the question isn't a yes or no. In this case,
07:32a user wants a completely opposite behavior. They don't want generic summarization or a one-word answer.
07:38If anything, they get irritated when a complex question gets answered with one word. And the problem
07:43is that Perplexity covers only one of these use cases, the second one. It also doesn't support any of
07:49the service requests. It can't make a reservation. It can't make a call. It can't book a service.
07:54Sequoia was blunt about this. The Gen.ai agent market is effectively a services market. And if you want to
08:00beat
08:00Google, you can't stop at information. You must provide a full service of a search engine, quick answers
08:06and research, and services. Perplexity's native browser removes a reliance on someone else's platform and gives them
08:13a competitive edge that it lacks. Because users are already logged in, the experience is personalized.
08:18And most importantly, the browser becomes the single front door for every type of query, from a basic question
08:23to complex tasks. That is something Google is boxed out of. Chrome is structured around the search bar,
08:29which is structured around the search ads. Building an AI-first browser that simply provides answers would
08:36cannibalize the very revenue stream that Chrome is designed to protect. Perplexity Comet browser just came out recently.
08:42And if you've had a chance to test it out, please do let me know what you guys think about
08:45it in the comments.
08:46I'm really looking forward to seeing perplexity's end of year results after this launch. But
08:50nevertheless, on pure economics, perplexity doesn't stand a chance against Google. But business isn't
08:56just about economics. It's also about politics and law. And who doesn't love a good old plot twist?
09:02Let's talk about what's hot off the press. Google's antitrust ruling.
09:06The US Department of Justice has been pursuing antitrust lawsuits against Google. As a quick side note,
09:12antitrust laws are meant to protect competition and commerce. In the US, it is not illegal to be big
09:17and powerful. There is absolutely nothing wrong with gaining monopoly position from superior products
09:22or services. However, what is illegal is when a monopoly takes predatory steps to stop its rivals.
09:29The antitrust lawsuit against Google is focused on two main areas. The first one being that Google has
09:35an illegal monopoly in the general search and search advertising market. The second one around Google
09:40being a monopoly in online advertising technology. In 2024, Google lost the biggest antitrust challenge
09:47when the judge Amit Mehta found that it illegally monopolized the search market. And this month,
09:53the US government has ordered Google to sell its Chrome web browser and license search data to
09:59competitors. And by the way, this would mark the biggest forced breakup of a US company since AT&T
10:05was dismantled in 1984. So, who do you think expressed interest in purchasing Google Chrome?
10:10You guessed it. Perplexity and OpenAI. The Justice Department said that Google,
10:15whose search engine controls nearly 90% of online queries, has paid around $26 billion to maintain a
10:23monopoly over the search market by agreements with tech rivals, smartphone manufacturers, and wireless
10:28providers. In exchange for a cut of advertising revenue, those companies, including Apple and
10:33Samsung, agreed to set Google as the default on browser and mobile devices. Google illegally
10:39monopolized the market for general search and search text advertising, the ads that appear at the
10:44top of the search results page. So, what happens now? The judge proposed the following. A forced sale of
10:50Google Chrome, like I mentioned before, prohibit Google from entering into any kind of exclusive deals.
10:55And for its existing agreements, Google would be required to offer smartphone makers and wireless
11:00carriers the option to display a choice screens to users. Now, how did Google respond? Google said that
11:06it intends to appeal the judge's ruling that it has illegally monopolized the search market,
11:10which could delay a remedy for months or even years. Google also said that it plans to challenge any
11:16ruling that calls for Chrome to be divested and also proposed a counter set of remedies that would
11:21modify its default search agreements with Apple and Mozilla. It also said that it would make it easier
11:26for phone makers such as Samsung to use its Android operating system without having to make Google
11:31search the default on the devices. Now, what did perplexity offer Google? Perplexity made an audacious
11:38$34.5 billion all-cash offer, which is their first big acquisition move. Now, personally, I don't think it'll happen,
11:49but... Now, did I think there would be an M&A situation around perplexity? Yes. And I was hoping
11:55that it would be an A acquisition by, for example, Apple, and they have good reasons for it. But did
12:01I
12:02think perplexity would be offering to buy Google Chrome? No, that wasn't in my wildest dreams.
12:08Perplexity's bid was made through a letter to Alphabet's CEO, and it promises to contain Chrome as open
12:13source, invest $3 billion after two years, and, drumroll, keep Google as the default search engine.
12:22This offer would essentially allow perplexity to access Chrome's user base, 3 million, and transform
12:28perplexity's engine economics without displacing Google's core search monopoly. This would also allow
12:33perplexity to gradually introduce its AI-powered search through the browser while avoiding the direct
12:39confrontation that has limited other Google competitors. The industry analysts are very
12:44skeptical about this offer and deem it as a stunt that significantly undervalues Chrome's true worth.
12:50Many suggest that the browser could be worth upwards of $50 billion because of its unmatched data
12:56collection and user reach. Now, even if there is a universe where perplexity can pull this off,
13:01I really don't know where they're going to get $34.5 billion in cash. Perplexity's valuation is $18
13:10billion, which is why this whole move has raised serious question about financial ability of perplexity
13:16to pull this off. And at the same time, Google has shown no intention to sell Chrome. Now, the interesting
13:21thing is that OpenAI and Yahoo have also expressed interest in Chrome, which could potentially trigger
13:26a bidding war that can push valuations far beyond perplexity's resources. But let's see what happens.
13:33Conclusion. Building a better search engine isn't the hardest part. Building a profitable one is.
13:39Perplexity has created a beautiful and extremely high-quality product, but their own unit economics
13:45plays against them. Every AI-powered response costs exponentially more than a keyword match. And no
13:51amount of venture capital funding changes that math. Google's monopoly isn't only about distribution
13:56in data. It's about the fact that they discovered a search model that actually makes more money at
14:01scale. And this is why perplexity cannot beat Google search, at least not today. I hope you found this
14:07video helpful. Let us know what you think in the comments. Till next time. Bye.
Comments