The internet is vast and wild and unknowable and full of potential, unless you are a website. If you are a website, you depend on traffic. And if you depend on traffic, you know that it comes from just a few different places. Facebook is a big one, and for many sites the biggest. Pinterest is enormous, staggeringly so, for sites that overlap with Pinterest's audience. LinkedIn sends a lot of people if you write about business or self-help; Twitter sends a very modest and modestly valuable stream of people to stories about the news. In other words, in 2014, normal people read the internet mostly on their phones, usually through apps. The only honest way to assess your website is to imagine it with fat blue bars above and below your words.
From the captive publisher's perspective, Google was the Facebook of five to ten years ago. Some publishers learned how to game it and reaped great rewards. Others simply kept on doing what they were doing and were rewarded anyway: Google, in its attempt to "organize the world's information," deemed their information worth organizing. This is what happened to MetaFilter, one of the great online communities of the 2000s. Over the years, the site's "Ask MetaFilter" section accumulated a lot of questions and answers that happened to be similar to questions Google users were asking. This became a major contributor to Metafilter's traffic, and remained so for years. Until recently:
While MetaFilter approaches 15 years of being alive and kicking, the overall website saw steady growth for the first 13 of those years. A year and a half ago, we woke up one day to see a 40% decrease in revenue and traffic to Ask MetaFilter, likely the result of ongoing Google index updates. We scoured the web and took advice of reducing ads in the hopes traffic would improve but it never really did, staying steady for several months and then periodically decreasing by smaller amounts over time.
Google updates its index all the time, ostensibly in an effort to kill off spammy how-to sites and content farms. MetaFilter isn't perfect but it's neither of those things, and now it has to lay off some of its staff. MetaFilter enjoyed Google's traffic but didn't ask for it, and now that traffic is gone.
Metafilter came from two or three internets ago, when a website's core audience—people showing up there every day or every week, directly—was its main source of visitors. Google might bless a site with new visitors or take them away. Either way, it was still possible for a site's fundamentals to be strong, independent of extremely large outside referrers. What's so disconcerting now is that the new sources of readership, the apps and sites people check every day and which lead people to new posts and stories, make up a majority of total readership, and they're utterly unpredictable (they're also bigger, always bigger, every new internet is bigger). People still visit sites directly, but less. Sites still link to one another, but with diminishing results. A site that doesn't care about Facebook will nonetheless come to depend on Facebook, and if Facebook changes how Newsfeed works, or how its app works, a large fraction of total traffic could appear or disappear very quickly.
Of course a website's fortunes can change overnight. That these fortunes are tied to the whims of a very small group of very large companies, whose interests are only somewhat aligned with those of publishers, however, is sort of new. The publishing opportunity may be bigger today than it's ever been but the publisher's role is less glamorous: When did the best sites on the internet, giant and small alike, become anonymous subcontractors to tech companies that operate on entirely different scales? This is new psychological territory, working for publishers within publishers within publishers. The ones at the top barely know you exist! Anyway, internet people, remember this day in five years: It could happen to you, whether you asked for it or not.