Transcription
For a brief moment, Perplexity looked unstoppable. AI search was the future. Investors were calling Google the next blockbuster. And Perplexity, a tiny startup founded in 2022, looked like it might be the Netflix of search. Fast answers, clean citations, no ads, just the truth, instantly. It felt like one of those rare moments in tech when the incumbents were asleep and a startup had finally figured out what everyone else had missed.
So, here's the question. If Perplexity was so obviously better, why did it lose, not crash, not disappear, but quietly fall out of the race to own the future of AI search? This isn't a story about a bad product. It's a story about timing, economics, and distribution, and what happens when giants wake up.
Perplexity was founded in August of 2022 by Aravan Sernvas, Dennis Yarrettz, Johnny Hoe, and Andy Kwinsky. Serinas wasn't some outsider chasing hype. He'd worked at Google, DeepMind, and Open AI, the exact institutions shaping modern AI. He understood search. He understood large language models. and most importantly, he understood where things were breaking. Their pitch was simple and compelling. Google doesn't answer questions. It gives you links. Perplexity would do the opposite. You ask a question, it finds an answer and shows you the sources so you can verify everything yourself. There would be no SEO games, no sponsored placements, and no digging through 10 blue links written for algorithms instead of humans. That alone made it feel revolutionary.
By early 2023, Perplexity already had millions of users and investors started whispering something dangerous. Search is broken and Perplexity just fixed it. But under the hood, Perplexity had a problem it could never fully escape. It didn't train a massive foundational model from scratch. Instead, it relied on existing models from OpenAI, Anthropic, Meta, Google, and others and layered its own system on top. From a user perspective, that didn't matter. But from an industry perspective, it mattered a lot because it led critics to a brutal question. Is Perplexity actually doing AI or is it just an AI rapper?
That criticism stuck fairly or not. And here's the uncomfortable truth. By 2024, Chad Chippy had added citations. Google added AI overviews directly into search. Claude, Gemini, and others followed the same path. Perplexity's core differentiator, answers with sources, stopped being unique. Once that happened, the value proposition got shaky. Why pay for Perplexity and Chad GPT and still use Google when Google and Chad GPT were offering Perplexity's same product.
This is where the story turns because search isn't just about intelligence. It's about cost. Every Perplexity query runs a live search, invokes a large language model, and synthesizes results in real time. That costs real money every single time. But Google doesn't work that way. Most Google searches cost fractions of a cent and don't require expensive LLMs. Even when Google adds AI summaries, it does so selectively. only when it makes economic sense and at massive scale. Perplexity, meanwhile, was burning compute on every query while it competed against a company that optimized search economics for 25 years. That's not a product problem, that's a physics problem.
And to be clear, Perplexity did grow. By mid 2025, Serinas publicly stated that Perplexity was processing hundreds of millions of queries per month with strong month-over-month growth. That's impressive, but scale is relative. Google handles tens of billions of searches per day. Even optimistic perplexity numbers represent a tiny fraction of global search volume, well under 1%. That doesn't mean that Perplexity failed. It means the idea that it would replace Google was never realistic.
There was another problem Perplexity couldn't escape. Monetization. For decades, Google search has been subsidized by one thing, ads that convert. High intent queries, clear attribution, and massive margins. Perplexity tried to rethink that model. Instead of traditional search ads, it experimented with subtle trust preserving formats like sponsored follow-up questions and light brand integrations that didn't interrupt answers. And then in late 2025, Perplexity hit pause. They stopped onboarding new advertisers to rethink their entire advertising strategy. But why? Because the numbers didn't justify the trade-off. According to reports, Perplexity generated roughly $20,000 in ad revenue out of $34 million total revenue. Not per month, not per quarter, but in total. Advertisers wanted performance, but Perplexity's products were optimized for answers, not buying behavior. So, Perplexity faced an impossible choice. Add more ads and risk user trust or protect the experience and accept weak monetization. Google never had to make that choice. Its ad engine was already built, battle tested, and embedded into how the internet works. Perplexity was trying to invent a cleaner version of search economics while competing against the most profitable ad engine ever created.
Then came the lawsuits. Publishers accused Perplexity of scraping proprietary content, reproducing articles too closely, and attributing hallucinated quotes to real outlets. The New York Times sent a cease and desist. The BBC threatened legal action, and Reddit sued. Perplexity responded by saying it aggregates, not plagiarizes, and offered revenue sharing agreements. But perception matters. Google can negotiate from a position of power, but a startup cannot. Instead of being the ethical alternative to Google, Perplexi started looking like another platform built off of other people's work. That damaged trust with publishers, regulators, and enterprise customers. And once that trust erodess, growth stops compounding.
Perplexity tried to escape search by building Comet, its AI browser. The logic made sense. Own the interface, own the users, and control distribution. After all, this is how Google won in the first place. But history here is brutal. Microsoft integrated chat GPT directly into Bing in early 2023. They had default placement on Windows, exclusive access to GPT4 and billions in marketing spend. If anyone was positioned to force behavior change, it was Microsoft. But years later, Bing's market share barely moved. If Microsoft couldn't shift user behavior meaningfully, what chance did perplexity have? Browsers in search are sticky. People don't switch unless the improvement is overwhelming. Not incremental, not slightly better, but lifechanging. By the time Comet arrived, AI search was already table stakes. Google didn't win because its AI was better. It won because of its massive ecosystem of products like Google Drive and YouTube. Search isn't a product, it's infrastructure. Perplexity was trying to replace one brick in a fortress, but Google embedded AI into the entire castle.
So, did Perplexity lose? Not exactly. Perplexity didn't die. It didn't collapse, and it didn't vanish. But it lost the war that it set out to fight. It didn't replace Google, redefine search, or become the default. Instead, it became a powerful research tool, a niche product for power users, and a feature that incumbents copied and absorbed, which is still impressive, just not revolutionary. Perplexity's real mistake wasn't the product. It was believing that being better was enough. In tech, timing matters, distribution matters, and economics matter. And most of all, when giants are asleep, startups look like killers. But when giants wake up, they stand no chance. Thanks so much for watching. If you enjoy learning about the intersection of business and tech, please give this video a like and comment which company we should explore next.