Last week in our Error 402 series on the history of web monetization, we wrote about the earliest forms of web advertising: banner ads. As we noted, this “simple” way of making money seemed to derail other forms of monetization, including early attempts at paywalls (which had many other problems and were destined to fail). Either way, though, the success of ads as a way of funding content online in the mid to late 90s of the original dot com boom certainly meant that much less attention was paid to other ways to monetize content on the web.
Of course, much of the early dot com boom in advertising was funded by startups with massively inflated bank accounts from venture capitalists, and those companies, in turn, spent the money on ads. Sometimes, this all turned criminal, such as the time AOL agreed to help an ad buyer show fake revenue in order to keep the company purchasing ads.
The boom in online ads also created its own industry of middlemen, with DoubleClick being the largest. DoubleClick became a kind of marketplace connecting advertisers with websites, and then tracking how many views those ads received. The earliest ads were paid (like most other forms of advertising before it) based on views, not clicks (despite DoubleClick’s name).
But another type of online advertising quickly surpassed it.
Search on the web was seen as a big business from early on, with many players. Yahoo (which was initially just a human-created index that later bolted on a search engine), Lycos, Altavista, Excite, Infoseek, WebCrawler, Dogpile… the list went on and on. And most of them made money by utilizing those banner ads we discussed last week. In fact, most of them shifted from being “search engines” to being “portals” where they hoped you’d spend more and more time (and see more and more ads).
Then came two companies that changed a lot: Google (obviously) and GoTo. Lots of people will know about Google, which bucked the trend of all those others in two important ways: first, it said that it didn’t want to be a portal, but just to provide search and send you on your way (how things have changed, huh?) and second, it made search much better through innovations like PageRank (again, how things have changed…).
In the early days, Google had no business model. Indeed, there was a point at which they thought their business model would be selling their search tools to enterprises, or just licensing their search code to portals like Yahoo’s (which they did for a time). Eventually, Google hit on its text relevant ads that were (initially) based off of the keywords you searched for and were (initially) in a separate and distinct column on the results page. This was originally called AdWords and by 2002 it was based partly on how much people were willing to pay to advertise, but also on performance (the more people clicking through, the more your ads would rise to the top).
That pricing mechanism was at least marginally similar to another company’s. And for that, we turn to the story of Goto. Goto was launched in 1997 (before Google!) and was one of dozens of companies that spun out of Bill Gross’ Idealabs, which was a sort of incubator for early internet ideas. Goto was also a search engine, but it’s big differentiator was that every result would be a paid result. The only way to get into the Goto search results was to pay to be there.
And, as part of that, Goto had to figure out how to handle ranking the results, and decided to do it… via an auction system. Those who wanted to be listed in Goto would put in the keywords they wanted to show up for, as well as the maximum price they’d pay for a click. And the search results would be ranked accordingly.
But who wants to go to a search engine that is all advertising? Turns out not many people. What ended up happening was that Goto quickly got into the search spamming arbitrage game. They’d buy ads on Yahoo or AOL for valuable keywords on Goto, but at a price less than advertisers on Goto would pay, and then direct results through to Goto. Buy the keyword on Yahoo for 5 cents, sell it on Goto for 10. Profit.
Eventually, as Google’s advertising grew by using a similar auction model, but (again, initially!) keeping the ads separate from the organic search results, basically all the other search engines started relying on Goto not as its own search engine but to power keyword based ads to compete with Google. Goto changed its name to Overture (and sued Google for patent infringement), and basically became the alternative to Google ads for many search engines, including Yahoo. Eventually Yahoo bought the company outright.
But “keyword” advertising had been shown to be a money printing machine, and Google (and, later, Overture/Yahoo) looked to expand it beyond search. Twenty years ago, Google launched AdSense, with the claim that it would help publishers who were having trouble setting up their own advertising leverage Google search keyword advertising. It wasn’t based on keywords directly (since publishers didn’t have keywords in the same way search did), but took that underlying concept to target ads based on the content on a publisher’s site, tied back to the keyword-based advertising Google had become so dominant in.
Of course, that also continued to grow and grow. Google eventually swallowed DoubleClick, and basically became the center of gravity for all kinds of internet advertising. Yahoo and Overture faded, and were eventually bought for pennies by Verizon who merged it into AOL (which it had also recently bought), and then sold it off again in a deal no one remembers.
But the dominance of internet advertising for publishers became basically unstoppable as a business model. I still remember, in the early 2000s when Techdirt resisted putting ads on the site, someone emailing us incredulously asking how stupid we were for not cashing in on the advertising boom.
At the very least, though, it dampened any discussion of more creative business models for content. And that’s what we’ll discuss in the next installment (which will be in two weeks, as next week is Thanksgiving).