* * *
An ad supported web grows quickly and is open to those who can’t or won’t pay. But it has at least four downsides as a default business model.
First, while advertising without surveillance is possible—unverifiable advertising was the only type of advertising through most of the 20th century—it’s hard to imagine online advertising without surveillance. The primary benefit of online advertising is the ability to see who’s looking at an ad. Simply paying for online advertising requires surveillance, if only to eliminate clickfraud. And if Cegłowski’s theory is true, there’s no apparent escape from escalating surveillance to create more attractive business propositions.
Second, not only does advertising lead to surveillance through the “investor storytime” mechanism, it creates incentives to produce and share content that generates pageviews and mouse clicks, but little thoughtful engagement. Clickbait has become so prominent that even Upworthy, popularizer of spreadable media as a tool for social change, is asking advertisers to consider how much attention readers are paying to content, rather than how many pageviews it generates. Some new media empires are so attached to advertising metrics that they are giving writers days off from “traffic whoring” duty to allow them to produce content that has greater social and informational value. While many newspapers are shielding their reporters from statistics about whether their stories are being read, the increasing importance of digital news outlets to the public sphere suggests we may get less news that helps us engage as citizens and more news designed to get us to click the “next page” button.
Third, the advertising model tends to centralize the web. Advertisers are desperate to reach large audiences as the reach of any individual channel shrinks. A generation ago, you could reach a significant fraction the the American population by buying ad time on four television networks. Very few companies can offer that “Superbowl ad” reach today. Advertisers purchase ads scattered across hundreds of sites, buying demographic targeting at the lowest rates available. Companies like Facebook want get as much of that money as possible, which means chasing users and reach. Using cash from investors and ad sales, they can acquire smaller companies that are starting to build rival networks. (See Facebook’s acquisition of Instagram and, to a lesser extent, Whats App.) This centralization has dangers for online speech—it means decisions these platforms take to ban speech are as powerful as decisions made by governments, as Rebecca MacKinnon has eloquently documented.
Finally, even attempts to mitigate advertising’s downsides have consequences. To compensate us for our experience of continual surveillance, many websites promise personalization of content to match our interests and tastes. (By giving platforms information on our interests, we are, of course, generating more ad targeting information.)
This personalization means that two readers of The New York Times may seen a very different picture of the world, and that two users of Facebook certainly do, shaped both by our choice of friends and by Facebook’s algorithms. Research suggests that these personalized sites may lead us into echo chambers, filter bubbles, or other forms of ideological isolation that divide us into rival camps that cannot agree on anything, including a set of common facts on which we could build a debate. While many have written on this topic (and I wrote a book on it), few have shown the implications of overpersonalization as well as Gilad Lotan did in this recent analysis of media consumption in Israel and Palestine, where he describes the view participants in the current Gaza war have of the conflict as “personalized propaganda.”
It’s easier to rant about technology than it is to propose solutions. To Cegłowski’s credit, he closes his talk with a set of practical suggestions about limits we might put on the use of digital data by advertisers. He demands that we be given a right to review and delete data companies hold on us, proposes a time limit on how long data can be held and how it can be shared. Implementing these regulations, of course, would require finding a regulator with teeth, and it’s not clear that the Federal Trade Commission would be willing to enforce such constraints on companies that are becoming powerful actors in Washington.
More importantly, Cegłowski offers us a way forward through his own actions. Cegłowski wrote and maintains Pinboard.in, a simple and powerful bookmarking service with an unusual business model. Each user of the service pays a one-time fee, which rises a fraction of a cent with each new user. (When I signed up for Pinboard, it cost $5, and now costs a bit more than $10.) The cost has the benefit of keeping the service spam-free—Metafilter has seen some of the same benefits from their nominal membership fee—and has meant that the service has been profitable since it was launched. Users can upgrade to a $25-per-year version that archives every webpage you bookmark, creating a permanent, searchable archive of your journeys through the web. Cegłowski promises that he will never sell ads on the site and never sell data to third parties, reminding us, “If you’re not paying for your bookmarking, then someone else is, and their interests may not be aligned with yours.”
Pinboard launched in 2009, in part in reaction to changes at Del.icio.us, a beloved bookmarking site started by Joshua Schacter and sold to Yahoo, which slowly ran the site into the ground. Many of Pinboard’s policies can be read as Cegłowski’s attempts to protect himself—and anyone else—from having precious personal data held hostage when a company changes hands. But these principles are also an invitation to think about, and perhaps to create, a web that works very differently.
The web is celebrating a 25th anniversary, but that celebrates the invention of the HTTP protocol by Tim Berners Lee. In practical terms, the web as we know it is less than 20 years old. Many of the services we rely on, like Twitter, are less than 10 years old. Yet it’s often hard to imagine making deep, structural changes to the web. It’s easy to assume that aspects of the web’s architecture and business model are inevitable: We will inexorably move towards a web that is centralized, ad supported and heavily surveilled.
Part of the celebration of the Web’s 25th anniversary is The Web We Want, a campaign to open the dialog about how the Internet is structured and governed to voices from around the globe. I think it’s at least as important to consider how we want the web to make money, as these decisions have powerful unintended consequences.
One simple way forward is to charge for services and protect users’ privacy, as Cegłowski is doing with Pinboard. What would it cost to subscribe to an ad-free Facebook and receive a verifiable promise that your content and metadata wasn’t being resold, and would be deleted within a fixed window? Google now does this with its enterprise and educational email tools, promising paying users that their email is now exempt from the creepy content-based ad targeting that characterizes its free product. Would Google allow users to may a modest subscription fee and opt out of this obvious, heavy-handed surveillance?