Who pays for content and services on the internet?
My friend Bo Peabody thinks we should be asking not just whether ad-supported journalism is feasible, but whether ad-supported social networks will work. In a Washington Post op-ed titled “Twitter.org?“, Bo leverages his experience founding and running Tripod.com to suggest that social networking sites are misunderstood as content sites, and won’t be profitable as ad-supported properties. He suggests that, because these spaces are critically important digital public spheres, we should consider supporting them as nonprofits if necessary, but shouldn’t expect them to sustain themselves based on advertising. As I look more closely at Bo’s thinking, I’m concerned that advertising may not be a viable model to support anything other than search online, and that systems we are increasingly reliant on may be supported by the shakiest of foundations.
Bo may not be right that social networks need to become nonprofits – I’m interested in communities where participants are willing to pay for membership (see Dreamwidth or Metafilter as examples), or communities that might thrive via an alternative revenue stream (see Brian McConnell’s suggestion for how Skype could run a highly profitable Facebook or Twitter and generate more call traffic in the process.) But I’m increasingly convinced he’s right that advertising is a lousy way to support social network sites.
Internet advertising works extremely well in the context of a search engine. Many searches are intended to lead to transactions, so matching a paid ad to a query is sometimes a good user experience. Advertising can work well in the context of niche content – a website focused on cross-country skiing is a great place to advertise to cross-country skiiers, and there’s a decent chance they’re going to be interested in learning about your ski wax. Ads on sites like Facebook work much less well, and while targetting those ads based on demographics may make them more effective, that targeting doesn’t fix the core problem: people are using social network sites to communicate, not to consume content, and they don’t want to be bothered by ads when they’re communicating.
The good news – for users annoyed by ads, not for advertisers – is that we appear to learn very quickly how to ignore online advertising. comScore, a company that monitors user behavior on the web for advertisers, reported in 2007 that only 32% of internet users clicked on banner ads in a given month. By 2009, that number had fallen to 16% of internet users, and that a core 8% of all internet users – “Natural Born Clickers” (yes, that’s what they called the studies) – are responsible for 85% of all banner clicks on the web.
There’s at least two ways to spin this finding. comScore, which exists to provide information to advertisers and would be out of business if people stopped buying online ads, uses this data to make the case that advertisers should stop obsessing over clickthrough rates:
“The act of clicking on a display ad is experiencing rapid attrition in the current digital marketplace,” said Linda Anderson, comScore VP of marketing solutions and author of the study. “Today, marketers who attempt to optimize their advertising campaigns solely around the click are assigning no value to the 84 percent of Internet users who don’t click on an ad. That’s precisely the wrong thing to do, because other comScore research has shown that non-clicked ads can also have a significant impact.”
Anderson may be referring to this study by Gian M. Fulgoni and Marie Pauline Mörn, which finds a modest increase in users visits to an advertised website based on being exposed to that site in banner ads, even if they didn’t click them. The argument is a traditional advertising one – you can’t know whether that particular billboard led a customer to find you, but we know that exposure to ads builds your brand, so buy more billboards. And you may or may not be surprised to learn that Fulgoni is the co-founder and CEO of comScore.
There’s another response to the clickthrough study: ask yourself whether you, personally, ever look at banner ads on the web. You probably don’t – you’re “banner-blind“. Usability expert Jakob Nielsen uses this term to explain a wealth of eye-tracking studies that illustrate web users’ almost uncanny ability to sift through a webpage and focus only on the parts that contain actual content. (He’s reported on this behavior since 1997.) Nielsen concludes that web users are so good at avoiding paying attention to ads that the only way to make an ad banner effective is to be deceptive and disguise it as content. At the same time, his studies suggest that search ads – ads that are sometimes helpful to users – aren’t filtered out in the same way.
comScore’s study suggests we – collectively – may be becoming more banner-blind over time. If only half as many users click banner ads as did two years back, we might conclude that those users have learned how to ignore banners in the interim. If comScore would release demographic data on the 8% who are inclined to click, we might be able to confirm these suspicions. If those 8% are new internet users, it suggests a future internet with mature users too savvy to pay attention to most forms of advertising.
In the meantime, here’s a thought, this one from danah boyd – anyone building a new, ad-supported social network is building a business on that 8%. Assume for the moment that I’m right and that those 8% are the newest and most naive users. We’re at 74% internet penetration in the US – there just aren’t that many new users who can come online and click those ads. Instead, that 8% may well represent new users from other parts of the world, where internet penetration is much lower and where new, naive users are still coming online.
Companies like Facebook aren’t planning the future of their business around these users. As Brad Stone and Miguel Helft pointed out in a New York Times article, “In Developing Countries, Web Grows Without Profit“, some social network sites are beginning to question whether they’ll be able to continue providing services to users outside the US, Europe and other markets they perceive to be lucrative. The article points to efforts at MySpace and Facebook to provide lower-bandwidth products for developing nations, both to improve user experience and to cut costs in serving these markets. It’s possible to imagine a future in which Facebook, strapped for cash, focuses on providing services only to users their advertisers are interested in reaching. Technorati recently relaunched their blog search engine with a near-exclusive focus on English-language content, de-listing prominent non-English blogs – my guess is that the change reflects advertiser demands.
Internet users all over the world have access to a vast array of powerful publishing and communication tools. While some premium users pay for access to these tools, the vast majority do not. Whether we believe these tools can lead towards more transparent and democratic governance, or whether we’re skeptical of such cyberutopian ideas, it’s clear the internet would be a very different place if these tools weren’t available for free. If Facebook weren’t free, it would likely be orders of magnitude smaller… which would increase exclusivity, but lose some of its utility as a powerful tool for reconecting with lost friends. It would include fewer users from developing nations where credit cards are significantly less common. Optimised for membership revenues rather than for ad views, it would be a deeply different place.
Revenue models have a deep impact on digital spaces. Why’s Twitter growing so fast? My guess is that it’s because the founders are following the traditional social media playbook: attract a ton of users, promise to monetize them through targeted advertising, sell the company to a larger one for billions and never confront the difficulty of monetizing that ad space. We can imagine a different Twitter, one that decided to focus on digerati and first-movers – that space might have used invitations to control access or membership fees to limit growth. It would be less ubiquitous, more exclusive and have a different utility curve. Or consider a company like Demand Media, which publishes more that four thousand articles and video clips a day, all intended to answer commonly asked questions on search engines and create targeted advertising inventory. We tend to think of the Internet as a place where questions are answered by random people all over the world, organized into a useful collection by Google. What if those questions were answered hastily and poorly, all by the same company, through content commissioned for $20 a video? Demand Media focuses on the business model first, and appears to be positioned to reshape the biggest internet space of all – the search and content space – in the process.
Fernando Bermejo sent me a paper of his, “Audience manufacture in historical perspective: from broadcasting to Google“, which suggests that researchers have a “blind spot” when it comes to considering the power of revenue models in media environments. He references a debate, sparked by Canadian social scientist Dallas Smythe, who suggested that communications research overfocused on the cultural side of communications and didn’t pay enough attention to the economic dimensions. Fernando worries that we’re doing the same thing today, ignoring the pervasive influence advertising has on the contemporary internet environment.
I suspect he’s right. We’re far more likely to discuss peer production, open-source models or collaboration at the Berkman Center than we are to discuss how advertising might shape the future of Facebook. I spend far more time trying to figure out how activists are finding clever ways to use social media and how those uses may be shaping these tools that I do considering how ad models are shaping these tools. “Blind spot” is putting it mildly
In our defense – it’s hard to study advertising. The data’s hard to get – it’s carefully controlled and tends to be released with large price tags on it, while participatory media projects tend to release usage data and welcome analysis. And researchers tend to be biased towards what we’re inspired by – I’m fascinated and inspired by independent and citizen media, so I pay attention to them, even if most of the use of social network tools is for communication, not for media publishing,
What if the social internet as we know it is being built on sand, on ads that almost no one looks at now and fewer will look at in two years? What if we’re optimizing tools for advertising audiences that don’t exist and turning aside models for social media built on membership fees or premium services? What if my assertions and speculations are wrong, and advertising’s a sure-fire way to build the social web?
I’m realizing that I (and probably anyone studying social media) need to understand at a much deeper level how advertising really works, because it shapes the systems I study, the systems we increasingly rely on. We need to know who those 8% of users who were “born to click” are, and we need to think about what happens if they stop clicking.