Ask!-ing for a Legacy

Billions spent shaping how we search and talk to technology, and nothing left to show for it.

Visiting the Ask offices in summer 2011 remains one of the biggest disappointments of my professional career. Looking back, it was also one of the most useful reality checks I’ve ever had.

At the time, I was 22 and had just started working at 77Agency on Parker Street in Holborn. It was one of those odd central London roads where recognisable companies sat hidden behind anonymous office doors. Premier Models was upstairs. There were agencies, production companies, tiny firms posing as big firms, and big firms posing as tiny firms. But the one that always caught my attention was Ask.

Not Ask.com. Ask. Though it will always be Ask Jeeves to me.

Every lunchtime, I’d walk past the office on my way to Sainsbury’s for a salad, and every single time I did, I’d glance over at it. Because for people who grew up online in the late 1990s and early 2000s, Ask Jeeves wasn’t just another search engine. It was one of the internet’s first real characters.

Google was clean and functional. Yahoo felt corporate. AltaVista felt technical. But Ask Jeeves had something the others didn’t. A butler answering your questions made the internet feel less mechanical and less intimidating. It felt less like operating a system and more like interacting with something. And Jeeves belonged to a very specific era of the web that no longer exists. The same era as MSN Messenger screen names, GeoCities fan sites, LimeWire destroying the family PC, badly coded forums with glowing signatures, and evenings spent aimlessly wandering around the internet because there was still something mysterious about it.

Eventually, curiosity got the better of me, too.

One afternoon, I rang the buzzer.

Nothing.

I tried again another day. Still nothing.

This went on for months. Then, sometime around late September, somebody finally answered. After a slightly confused conversation downstairs, they let me come up and have a look around.

What I discovered completely shattered the mythology I’d built up in my head.

Ask’s entire UK operation was effectively four people. One marketing person overseeing locked-in agreements. One bookkeeper chasing invoices that were never getting paid. And two in telesales who tried, I guess. That was it. And somehow the reality was even sadder than the number itself. Only the marketing person really used the office with any regularity, and even then, it was mostly to occasionally check in on tenants. By that point, the Ask offices had effectively become a makeshift coworking space. Renting out desks was more commercially viable than whatever remained of Ask’s actual business in Britain.

The office itself felt frozen. Not in a charming retro way. In a forgotten way. There were old Ask-branded graphics still mounted on walls that looked untouched for years. Mostly empty desks, except for a few tenants, including a couple of administrators from the YMCA. Fluorescent office lighting. That stale silence develops in offices when people stop properly inhabiting them.

I remember standing there thinking: this can’t possibly be real.

This was Ask Jeeves. One of the most recognisable brands on the early internet. A company that had spent years trying to fight Google through acquisitions, pivots, relaunches, rebrands, and increasingly desperate reinventions. Billions had been spent trying to keep it relevant.

And this was what remained.

Four people and rented desks in Holborn. The latter still being the case.

That visit taught me something brutal about the internet: almost nothing stays on top for long. Especially online. One missed shift, one bad decade, one competitor that understands people slightly better than you do, and suddenly the company that once looked untouchable is reduced to a half-empty office renting desks to strangers.

And the strange thing is, Ask isn’t even unique in that. Even some of my old employers, 77Agency and Starcom MediaVest, both of which were doing genuinely pioneering digital work at the time, have been scrubbed from the internet, too. Whole eras of online culture and advertising history just vanished. Websites 404 Not Found themselves. Acquisitions happen. Rebrands happen. Search results collapse. People move on. Eventually, only the people who were actually there remember these companies existed at all. And even then, most of them have forgotten too.

I learned that lesson at twenty-two years old, wandering through the empty husk of Jeeves’ old servant quarters.

Then on 1 May 2026, after decades of launches, shutdowns, revivals, acquisitions, pivots, rebrands, and billions of dollars spent trying to keep the whole thing alive, IAC finally retired Ask Jeeves for good.

In truth, I’m not sure IAC was ever the right home for it.

Ask Jeeves only really worked when it still felt slightly whimsical, slightly odd, slightly human. The entire appeal was that Jeeves made search feel approachable at a time when the internet still intimidated people a little. But once Ask became another asset within a large holding company, most of that charm slowly wore away. The butler became baggage. The personality became embarrassing. The weirdness got managed out of it. And without that personality, Ask was just another search engine competing with Google.

Which was impossible.

Google won because it was cleaner, faster, and brutally efficient. Ask was never going to beat that by becoming a worse version of Google while simultaneously abandoning what made people care about it in the first place. By the end, the whole thing felt less like a company being maintained and more like an ageing relative being kept comfortable because nobody quite knew what else to do with them.

And then, after years of revivals, rebrands, pivots, and attempts to make Jeeves culturally relevant again, IAC finally took poor old, frail Jeeves out back and put him down.

Still, if there is anything genuinely worth remembering about Ask Jeeves, it is that it introduced several ideas that ended up shaping the modern internet far more than most people realise. The first was the idea of the “answer engine”. Ask Jeeves was one of the first major search platforms built around the premise that users did not simply want a list of websites. They wanted answers. That distinction feels obvious now because almost the entire modern search industry is built around it, but in the late 1990s, it was a genuinely different way of thinking about retrieval. Early search was still heavily influenced by directories, categories, keywords, and technical querying. Ask instead pushed the idea that a search engine should interpret intent and attempt to resolve a question.

That expectation now underpins almost everything. Featured snippets. AI Overviews. Voice assistants. Conversational AI. Large language models. Even the growing expectation that Google should simply tell you something instead of sending you elsewhere traces back to the same behavioural shift Ask helped normalise.

The second was natural language search itself. Ask Jeeves actively encouraged users to type full questions into a machine at a time when most people still searched like they were operating a database terminal. “Where can I buy concert tickets?” instead of “concert tickets london”. It sounds trivial now because modern interfaces have absorbed that behaviour for years, but culturally it mattered. Ask helped train people to speak to computers in a conversational way, long before conversational computing was mature enough to respond properly.

Without that shift, modern AI interfaces likely look very different. Siri, Alexa, ChatGPT, Gemini, Claude, and virtually every conversational interface today rely on an instinct Ask Jeeves helped establish decades ago: the idea that it is normal to phrase requests to a machine as though you are speaking to another person.

And finally, there was Jeeves himself. That part sounds superficial until you remember what most technology branding looked like at the time. Search engines largely presented themselves as cold technical utilities. Engineering products. Functional systems. Jeeves introduced personality into search at scale. Not just branding, but character. The butler made search feel approachable to ordinary people in a way many early technology companies simply did not understand.

Looking back now, it is difficult not to see traces of Jeeves everywhere. Virtual assistants with names. Chatbots with personalities. AI companies are obsessing over tone, warmth, friendliness, and conversational behaviour. Entire industries are trying to make machines feel socially legible rather than mechanically intimidating.

Ask Jeeves lost the Search Wars spectacularly. There is no rewriting that part. But history is full of companies that failed commercially while still shaping the future in ways their successors benefited from enormously. And in that sense, Jeeves deserved a much better funeral than this.

Dillan Gandhi works at the intersection of search engines, business models, brand discovery, cultural relevance and user experience. He is the Principal Consultant at Searchlight London and Dillan Gandhi Consulting. Connect with him here.