I never used Perplexity. Not once. While the tech world was buzzing about the "Google killer," I kept typing my queries into the same search bar I've used for over a decade. And honestly, I think that instinct is proving right.

The promise

Perplexity AI burst onto the scene with a genuinely compelling pitch: instead of giving you a wall of blue links, it would just give you the answer, complete with citations. It was fast, conversational, and felt like the future. By May 2025, the company was processing 780 million queries per month, had raised over $900 million in funding, and was valued at nearly $9 billion. Investors were confident. The narrative was clear: traditional search is dead. But narratives have a way of colliding with reality.

Google wakes up

The turning point, in my view, was Google AI Overviews. When Google first rolled out AI-generated answers above its traditional search results, the responses were rough. There was the infamous suggestion to put glue on pizza. Perplexity's CEO, Aravind Srinivas, publicly mocked Google's stumbles. It felt like the incumbent was fumbling and the upstart was winning. Then Google did what Google always does: it iterated, improved, and leveraged its incomprehensible scale. AI Overviews got better. Google added an "AI mode" for longer, more complex queries, essentially mirroring the chatbot-style interface that was supposed to be Perplexity's differentiator. And Google did all of this while sitting on top of the world's largest search index, the most comprehensive maps data, shopping integration, and 89.7% global market share. According to BrightEdge data reported by Yahoo Finance, Google's market share, which had been slowly eroding since December 2024, recently reversed course. For the first time since tracking began, AI search engines like Perplexity, ChatGPT, and Grok all lost share simultaneously as Google gained it back.

The numbers tell the story

Perplexity's traffic peaked and then started to slide. SimilarWeb data shows that the site received 170 million visits in January 2026, down 5.31% from the previous month. For a company built on growth, any decline is a red flag. The financial picture got complicated too. In March 2025, Srinivas took to Reddit to deny that the company was under serious financial pressure, responding to user theories that Perplexity was "doing horribly financially" and cutting costs aggressively. When a CEO has to publicly deny financial trouble on Reddit, that itself becomes the story. Meanwhile, Perplexity pivoted away from advertising entirely, betting its future on subscriptions and business sales. In early 2026, executives confirmed they weren't pursuing any ad deals, a sharp contrast to OpenAI's push into advertising. It's a principled choice, but it also means Perplexity needs to convince millions of users to pay for something Google gives away for free.

Why I still prefer Google

My reason for sticking with Google is simple: I like seeing the search results myself. When I search for something, I don't just want an answer. I want to see the landscape. I want to scan headlines, recognize sources I trust, click into a primary document, and form my own understanding. That process of browsing, evaluating, and synthesizing is part of how I think. An AI-generated summary, no matter how well-cited, short-circuits that process. There's a deeper issue too. Perplexity doesn't own any search infrastructure or exclusive data. As one commenter on Hacker News put it, "They're not doing anything that can't be replicated by bigger players with more resources." That's the fundamental problem. When your entire product is a better interface on top of someone else's data, and that someone else decides to build the same interface, your moat evaporates.

The structural problem

Perplexity faces a challenge that no amount of funding can easily solve. Google spends over $12 billion per quarter on AI research and infrastructure. Perplexity, despite its impressive fundraising, operates at a fraction of that scale. In a competition defined by data, compute, and distribution, those gaps matter. And distribution is perhaps the most underrated factor. Google is the default search engine on virtually every browser and phone. Users don't choose Google each morning, they simply don't un-choose it. Perplexity has to convince people to actively change a deeply ingrained habit. That's an enormous ask, especially when the incumbent is rapidly closing the quality gap. The legal challenges haven't helped either. Publishers have pushed back against Perplexity for scraping and summarizing their content without adequate attribution or compensation, a fight that adds cost and complexity to an already difficult business model.

What this means

I don't think Perplexity will disappear. It still has loyal users, real revenue, and a product that genuinely works well for certain types of queries, particularly research-heavy, conceptual questions where you want a synthesized answer rather than a list of links. But the "Google killer" narrative is dead. What we're witnessing is a familiar pattern in tech: a nimble startup identifies a genuine weakness in an incumbent's product, builds something exciting around it, and then watches as the incumbent absorbs the lesson and fights back with advantages the startup simply cannot match. Google fell behind on AI. That much is true. But "behind" for Google means something very different than it does for everyone else. When you have the infrastructure, the data, the distribution, and the resources that Google has, catching up is just a matter of time and focus. And once Google caught up, the question stopped being "Can Perplexity beat Google?" and started being "Can Perplexity survive next to Google?" For me, the answer was always obvious. I never left.

References