Ask.com officially shut down on May 1, 2026, after almost thirty years online. In its place, IAC put up a blank page that says, "After 30 years of answering the world's questions, Ask.com officially closed on May 1, 2026." The notice is gracious, the Jeeves nostalgia is genuine, and the strategic decision behind it — choosing to wind the business down rather than relaunch it as an AI search product — is the single largest unforced error in consumer AI I have seen in 2026.
The numbers are not subtle. Ask Media Group, the unit inside IAC's Search segment that owned Ask.com, generated $320.9 million in revenue across 2024, according to IAC's own quarterly filings. The full Search segment — Ask Media Group plus the legacy Desktop business — pulled in roughly $387.8 million for the year and produced positive Adjusted EBITDA in every quarter, including $6.0 million in Q4 2024. Even on the way down, IAC told investors in February that it expected Search to generate $240 to $295 million of revenue in 2025 with $10 to $15 million in EBITDA. This was a profitable, cash-generative business when the decision to kill it was made.
The catalyst was structural rather than financial. Ask Media Group's revenue model has, for years, been an arbitrage: IAC bought traffic, sent it to ad-supported search and content pages, and collected a share of the paid-listing revenue from a services agreement with Google. IAC renewed that Google agreement on January 20, 2025, and extended it only through March 31, 2026. Ask.com went dark on May 1, 2026 — five weeks after that contract expired. The shutdown is the punchline; the contract was the script.
Key facts
- 01Ask.com. A key thread of reporting in this story.
- 02IAC. A key thread of reporting in this story.
- 03Perplexity. A key thread of reporting in this story.
That backdrop is what makes the decision strategically interesting. IAC was not surprised by the end of its Google deal. The company had at least fourteen months of warning, a recognizable consumer brand, a domain that has been online since 1996, and roughly $1.8 billion in cash on hand at year-end 2024. The question is not whether they could afford to pivot. The question is what return profile they thought a pivot would offer compared to walking away.
“Ask Media Group did $321 million of revenue in 2024 with a 30-year-old brand and a question-and-answer heritage that maps almost exactly onto the chatbot UX. IAC's response was to turn off the lights.”— Jaeden Schafer
The comp set says the return profile would have been excellent. Perplexity, the cleanest pure-play AI search wrapper, hit roughly $500 million of annualized revenue in April 2026 and was last priced at a $21.2 billion valuation in its Series E-6 round. The company has 100 million monthly active users and tens of thousands of enterprise clients. You.com, a smaller wrapper with a fraction of the brand recognition, raised a $100 million Series C at a $1.5 billion valuation off roughly $50 million of ARR. Andi, Brave, and a handful of others are building real businesses on top of frontier-model APIs with consumer brands the broad public has never heard of. The wrapper category is no longer speculative. It is a durable, fundable, cash-generating layer of the AI stack.
Ask.com's strategic position relative to these companies is, frankly, embarrassing for the people who chose to shut it down. The site has nearly thirty years of brand equity, a name that is literally a verb for the action a chatbot performs, and a product heritage — natural-language Q&A with a friendly persona — that maps almost perfectly onto the modern LLM UX. The original 1996 Ask Jeeves pitch and the 2026 Perplexity pitch are, functionally, the same pitch. One has $500 million of ARR and a $21 billion valuation. The other is now a blank memorial page.
The relaunch math is not hard. A reasonable AI search rebuild on Ask.com — frontier model API costs, a small product team, a paid subscription tier at $20 per month plus a free ad-supported tier on top of the existing traffic — would have required, generously, $30 to $50 million of investment over two years. Even capturing 5 percent of Perplexity's user base, on a domain Americans have already typed into their browsers for thirty years, plausibly produces $30 to $50 million of subscription ARR within the first 18 months and a multi-billion-dollar mark on a category-leader basis. The downside case is "we wrote off the same number we wrote off by shutting it down." The upside case is "we just printed a public comp."
There is a defensible story for IAC's actual decision and it is worth steelmanning. IAC under Joey Levin and Barry Diller has historically been ruthless about exiting capital-intensive businesses with deteriorating economics — the spin-off of Match Group, the planned spin-off of Angi, the systematic wind-down of Ask Media Group's traffic acquisition spend over the past three years. The Search segment was a mature, declining ad-arbitrage business with structural exposure to a single customer (Google) and a depreciating contract. Pivoting to AI search would have required IAC to underwrite a competitive product roadmap against well-capitalized incumbents — Perplexity, Google's own AI Overviews, OpenAI's ChatGPT search, Anthropic's Claude — in a category IAC has not operated a real product team in for years. Capital allocators reasonably asked whether IAC was the right owner for that bet. The honest answer is probably no.
But "we are not the right owner" is not the same as "this asset has no value." The right move, if IAC truly did not want to operate an AI search business, was to sell the Ask.com domain, brand, and traffic to someone who did. Domain comparables alone make the case. AI.com sold for $70 million in February 2026, the largest domain transaction in history. Voice.com previously held the record at $30 million. Chat.com sold for $15.5 million. Ask.com — a four-letter, dictionary-word, three-decade-active domain that perfectly describes the consumer AI category — would have cleared eight figures in a competitive auction without anyone breaking a sweat. Adding the existing inbound traffic, brand recall, and SEO equity to the package would have pushed the bidding into low-nine-figures territory from any of the wrapper-stage AI startups currently struggling to manufacture brand awareness from a standing start.
Instead, the URL now redirects to an obituary. The domain registration sits behind IAC's wall until it expires. The traffic that did show up directly to Ask.com — and Similarweb's data confirms it was overwhelmingly direct, not Google-referred — is now bouncing off a blank page that thanks them for thirty years of curiosity. Some of those people will find Perplexity. Some will find Google's AI Overviews. None of them will find IAC, because IAC chose to be invisible in this category.
The deeper read here is about how legacy media companies see AI. IAC ran a search business that, on paper, was a perfect candidate to be repackaged as an AI product. The leadership team chose to treat it as a 1996-era ad-arbitrage business approaching contractual end-of-life rather than as a brand-and-distribution asset that could carry an entirely new product category. That framing decision is the fumble. The execution of the shutdown is just what the fumble looks like in public.
The AI Chat Daily read: in five years, the consumer AI search category will be a multi-tens-of-billions market, dominated by two or three brand-name players. None of them will be Ask.com. That is not because Ask.com couldn't have competed. It is because IAC declined to find out. Ask.com goes into the same case file as Yahoo's failure to acquire Google in 2002 and Blockbuster's failure to acquire Netflix in 2000 — not the most expensive miss on the list, but a clean, cleanly preventable one. Jeeves did not need to retire. He just needed a new uniform.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
