GoTo.com / Overture

What GoTo.com actually was

GoTo.com was the search engine that invented paid search as everyone now knows it: an open auction in which advertisers bid for placement against a keyword and pay only when someone clicks. Bill Gross, whose Pasadena incubator Idealab spun the company out, presented the idea at TED in February 1998 to an audience reported as confused and in places actively hostile. His argument was that a search engine's ranking could be sold transparently rather than editorialised, and that charging per click rather than per impression aligned what the advertiser paid with what the advertiser actually wanted.

It is filed here as no longer available, and the reason is corporate rather than technical: the company was acquired by Yahoo on 7 October 2003, and the Overture brand was retired on 18 April 2005 when the business was renamed Yahoo! Search Marketing. Today the goto.com domain belongs to an entirely unrelated cloud communications and IT support company. There is no continuity whatsoever between it and the search engine.

GoTo never won as a consumer destination. It won by becoming the plumbing: it syndicated its paid listings into other companies' search results — Yahoo, MSN, AOL, Excite — and took a share of the revenue. By 2003 it was supplying a large fraction of the commercial advertising on web search, had more than 88,000 advertisers, and was bought by Yahoo for a headline $1.63 billion.

Where its results came from: Inktomi, not a crawler

This is the part most often written wrongly, and it settles the engine's historical index position. GoTo.com never built a general web crawler for its consumer product. Its own "results" were the paid listings it sold at auction. The unpaid algorithmic results beneath them were backfilled from Inktomi from 15 June 1998 — bought in from a third-party index.

For the whole of its life as GoTo and most of its life as Overture, then, its historical position was resold results: an advertising marketplace running on top of somebody else's index. Even the domain's original search asset was bought rather than built — Gross acquired the World Wide Web Worm, one of the earliest web search engines, and repurposed the property instead of writing a crawler.

That changed only at the very end. In February 2003 Overture bought both surviving independent Western crawlers within a week of each other:

• AltaVista, agreed 18 February 2003. The announced consideration was $80 million in stock plus $60 million in cash, roughly $140 million plus assumed liabilities; one timeline reports the closing value as $106 million on 28 April 2003. The gap is consistent with stock-price movement between announcement and close, and both figures are given here rather than averaged.
• FAST's web search division, AllTheWeb.com, agreed 25 February 2003, for $70 million in cash plus up to $30 million in performance incentives over three years.

Overture's chief executive Ted Meisel framed both deals as combining paid placement with advanced algorithmic search technology from AltaVista and multilingual capability from FAST, in order to strengthen the Yahoo and MSN syndication relationships. Months later Yahoo bought Overture — and thereby AltaVista and AllTheWeb too. So the honest summary of the historical position is: no index of its own for five years, then a hybrid for a matter of months, then gone.

The invention: auction, per-click, self-serve

The mechanism GoTo invented is the mechanism that funds the modern internet, and it had three components that were novel in combination in 1998.

1. The keyword auction. Advertisers opened accounts, chose keywords, wrote a title, description and URL, and entered a bid. The system ranked competing listings for a term by bid amount, highest first. Bids started at one cent. The company's own analogy was the Yellow Pages.
2. Pay-per-click, not pay-per-impression. The advertiser paid only when a user clicked, which transferred risk from advertiser to publisher and made the price directly comparable to the cost of a lead. Top bids reached $1 per click by July 1998.
3. Self-serve at scale. From 1 June 1999 GoTo shipped tooling that let advertisers manage their own bids and listings without a sales representative. This is the underrated part: an auction only scales if tens of thousands of small advertisers can operate it unaided, and GoTo built that.

Revenue in practice came from syndication rather than from goto.com's own traffic. A "GoTo-in-a-Box" agreement dated 15 April 2001 placed its paid listings on Microsoft properties; from 20 November 2001 Overture supplied featured paid listings on Yahoo's results pages; AOL and Excite were among the other partners. By 2003 the distribution network covered MSN, Yahoo and AOL simultaneously. Overture mattered far more than its brand recognition suggested because it sold monetisation to everyone who had audience and no way to earn from it.

The early economics were deliberately loss-making. Gross bought traffic on Yahoo and AOL at around four cents a click and routed it to GoTo's results page, which did not pay for itself until keyword prices rose above the acquisition cost. Secondary reporting holds that Overture came to account for over 20% of Yahoo's revenue before the acquisition — a figure worth attributing rather than stating flatly, since it has not been confirmed against Yahoo's filings.

The patent, and what the litigation did and did not decide

On 28 May 1999 GoTo.com filed the application that became US Patent 6,269,361, "System and method for influencing a position on a search result list generated by a computer network search engine," granted on 31 July 2001 with ten named inventors. It covers bid-for-placement ranking of search listings with per-click charging.

Overture sued FindWhat.com in January 2002 and Google in April 2002, the latter over AdWords Select, which Google had launched that February. The Google case settled in August 2004: Google took a perpetual licence to the '361 patent and related Overture patents covering bid-for-placement ranking and account management tools, and issued 2.7 million shares of common stock to Yahoo, which dismissed the suit. A separate dispute over shares owed under a 2000 services agreement was settled at the same time. The timing mattered enormously: the settlement landed immediately before Google's IPO and removed the largest legal overhang from the offering.

Three things must not be asserted, because the record does not support them. Google never admitted infringement — a licence and a settlement are not a finding, and no court ever ruled that AdWords infringed. The patent's validity was never adjudicated: the one case that reached a jury, against FindWhat, ended in a mistrial in May 2005 when a California jury deadlocked, with an allegation that the patent was unenforceable for inequitable conduct left undecided; that case then settled in August 2005, with MIVA, formerly FindWhat.com, paying $8 million within five days plus ongoing revenue-based royalties in exchange for a worldwide non-exclusive licence. And no authoritative dollar value exists for what Google paid: it issued shares, not cash, and the figure depends entirely on which price is applied and when. Any monetary figure attached to that settlement should be treated as an estimate, attributed, and dated.

The '361 patent is today assigned to R2 Solutions LLC, having passed out of Yahoo's hands into a patent-holding entity.

What Google took, and what it added

The defensible version of the influence story is simple and does not need the litigation at all. Bill Gross has been open that GoTo invented the model. Google's founders were initially and publicly opposed to letting advertising drive results, and Google's first AdWords, launched in October 2000, was sold on a cost-per-thousand-impressions basis rather than as an auction. Google moved to a keyword auction with AdWords Select in February 2002, after GoTo's model had visibly worked. "Google copied GoTo's business model" is a defensible claim; "Google stole a patented invention" is not.

What Google added is the more interesting half. Google ranked ads by bid multiplied by quality and clickthrough, not by bid alone. That single refinement is not in GoTo's patent, and it made Google's version simultaneously better for users and more profitable, because a more relevant ad is clicked more often and therefore earns more per impression even at a lower bid.

Overture never fixed this. It ranked purely by bid to the end, and Yahoo did not replace that system with quality-weighted ranking until Panama, launched 5 February 2007 — five years after AdWords Select. That five-year gap is arguably the single most consequential product decision in the history of search advertising.

The sale to Yahoo, and two sets of numbers

Yahoo announced its acquisition of Overture on 14 July 2003, and the deal completed on 7 October 2003. Wikipedia's phrasing captures the strategic logic in six words: Overture was acquired by its biggest customer.

The reported price differs by source and by date, and both figures belong on the record. $1.63 billion is the value at announcement on 14 July 2003, when the all-stock deal was struck. $2.2 billion in stock and cash is reported at completion on 7 October 2003. Neither is wrong; they are the same transaction measured at different points, in a deal whose value moved with Yahoo's share price over three months.

What Yahoo was buying was the fastest-growing segment of internet advertising: commercial search was projected at the time to grow from roughly $2 billion at the end of 2003 to roughly $5 billion by 2006. Yahoo was itself an Overture distribution partner and had discovered that its supplier accounted for a large share of its own revenue, so it bought the supplier. The Overture Keyword Selector Tool, a free advertiser utility that a generation of search marketers used, disappeared with the transition; it was a by-product of the advertising system, not the business.

Timeline

• 1997 — GoTo.com established under Bill Gross's Idealab; Gross acquires the World Wide Web Worm and repurposes it.
• 21 February 1998 — Gross presents pay-for-placement at TED8, to a poor reception.
• February 1998 — GoTo.com launches bidding for search-result placement, charged per click.
• 15 June 1998 — Inktomi begins supplying the algorithmic results alongside GoTo's paid listings.
• 28 May 1999 — Patent application filed that becomes US 6,269,361.
• 1 June 1999 — Self-serve advertiser tools launched.
• June 1999 — IPO; GoTo.com becomes a public company.
• October 2000 — Google launches AdWords, sold on a per-impression basis.
• 15 April 2001 — "GoTo-in-a-Box" agreement places GoTo listings on Microsoft properties.
• 31 July 2001 — US 6,269,361 granted.
• 8 October 2001 — GoTo.com renames itself Overture Services, Inc.
• 20 November 2001 — Overture supplies featured paid listings on Yahoo.
• January 2002 — Overture sues FindWhat.com over the '361 patent.
• February 2002 — Google launches AdWords Select, a keyword auction.
• April 2002 — Overture sues Google over AdWords Select.
• 18 and 25 February 2003 — Overture agrees to buy AltaVista, then FAST's AllTheWeb.
• 14 July 2003 — Yahoo announces the acquisition of Overture.
• 7 October 2003 — Acquisition completes.
• August 2004 — Google and Yahoo settle; perpetual licence to Google, 2.7 million shares to Yahoo, immediately before Google's IPO.
• 18 April 2005 — Overture Services renamed Yahoo! Search Marketing; the brand ends.
• May and August 2005 — Overture v FindWhat ends in a mistrial, then settles with MIVA for $8 million plus royalties.
• 5 February 2007 — Yahoo launches Panama, replacing bid-only ranking with quality-weighted ranking.

What it got right, what it got wrong

What it got right is enormous. GoTo.com solved the problem that had defeated every search company before it: how to make web search pay. Banner advertising did not work. Portals did not work. Subscriptions did not work. An open, self-serve, pay-per-click keyword auction did, and it worked so well that it became the economic foundation of the consumer internet. Gross deserves the credit and frequently does not get it, because the company that perfected the model is not the company that invented it.

What it got wrong were two things that compound. First, it ranked purely by bid. That maximises revenue per click in the short run and corrodes the results page in the long run, because the highest bidder is not the most useful answer and users learn to distrust the page. Google's quality-weighted auction produced better results and more revenue per search at the same time. Overture never fixed it, and Yahoo took until 2007, by which point the market had gone.

Second, it never owned any search technology. It bought its algorithmic results from Inktomi, then in 2003 bought AltaVista and AllTheWeb — the two best independent indexes left — and was itself acquired eight months later. It had built a magnificent monetisation layer on rented foundations, and when the company that owned its own foundations built a monetisation layer of its own, Overture had nothing left to defend with except a patent. A patent is not a product.

The instructive lesson is not that the model was bad. The model won completely. It is that inventing the business model is not the same as owning the market. GoTo invented paid search and captured a fraction of its value; Google combined the same auction with the best index and the best relevance and took nearly all of it.

Common misconceptions

"Google invented pay-per-click search advertising." No. GoTo.com did, launching in February 1998 and filing the patent in May 1999. Google's AdWords launched in October 2000 on a per-impression basis and only became a keyword auction with AdWords Select in February 2002.

"Google was found to have infringed Overture's patent." No. The case settled in August 2004 with a perpetual licence and a share issuance. There was never a finding of infringement, and the patent's validity was never adjudicated — the one trial that happened ended in a mistrial.

"Overture was a search engine with its own index." No, and this is the important structural fact. Its own results were the auctioned advertisements; the algorithmic results were licensed from Inktomi from June 1998. It acquired real crawling technology only by buying AltaVista and AllTheWeb in 2003.

"Yahoo bought Overture for $2.2 billion" or "for $1.63 billion." Both figures circulate and both are right for their dates: $1.63bn at announcement on 14 July 2003, $2.2bn reported at completion on 7 October 2003.

"GoTo.com is still around at goto.com." No. As of 19 August 2026 that domain belongs to an unrelated cloud communications and IT support business. "GoTo.com is the same as Go.com or Go2Net." Also no — Go.com was Disney's portal and Go2Net a separate company, and the record carries an explicit note distinguishing them.

Frequently asked questions

Sources

• en.wikipedia.org/wiki/GoTo.com
• patents.google.com/patent/US6269361B1/en
• thehistoryoftheweb.com/goto-forgotten-search-engine/
• bgtheory.com/blog/goto-to-overture-to-ysm-timeline/
• computerworld.com/article/1720542/yahoo-to-buy-overture-for-1-63b.html
• newsbreaks.infotoday.com/NewsBreaks/Overture-Acquires-Two-Major-Web-Search-E…
• searchenginewatch.com/2004/08/08/google-yahoo-settle-patent-and-share-disput…
• theregister.com/2005/05/17/overture_ad_patent_mistrial/