III. The Business Model as Belief and Reality
Why is Facebook such a repeatedly bad actor in its relationship to its users, constantly testing and probing for ways to quietly or secretly breach the privacy constraints that most of its users expect and demand, strategems to invade their carefully maintained social networks? Because it has to. That’s Facebook’s version of the Red Queen’s race, its bargain with investment capital. Facebook will keep coming back and back again with various schemes and interface trickery because if it stops, it will be the LiveJournal or BBS of 2020, a trivia answer and nostalgic memory.
That is not the inevitable fate of all social media. It is a distinctive consequence of the intersection of massive slops of surplus investment capital looking desperately for somewhere to come to rest; the character of Facebook’s niche in the ecology of social media; and the path-dependent evolution of Facebook’s interface.
Analysts and observers who are content with cliches characterize Facebook as sitting on a treasure trove of potentially valuable data about its users, which is true enough. The cliched view is that what’s valuable about that data is names associated with locations associated with jobs associated with social networks, in a very granular way. That’s not it. That data can be mined easily from a variety of sources and has been mined relentlessly for years, before social media was even an idea. If an advertiser or company or candidate wants to find “professors who live in the 19081 area code who vote Democratic and shop at Trader Joe’s in Media” they can buy that information from many vendors. If that were all Facebook was holding, it wouldn’t have any distinctive wares, even imagined, to hock. All it could do is offer them at a bargain rate–and in the global economy, you can’t undercut the real bargain sellers of information. Not that this would keep Facebook from pretending like it has something to sell, because it has a bunch of potentially angry investors ready to start burning effigies.
What Facebook is holding is a type of largely unique data that is the collaborative product of its users and its interface. But if I were a potential buyer of such data, I’d approach my purchase with a lot of caution even if Facebook managed to once and for all trick or force its users into surrendering it freely to anyone with the money to spend. If my goal is to sell something to Facebook users, or to know something about what they’re likely to do in the future, in buying Facebook’s unique data, what I’m actually learning about is a cyborg, a virtual organism, that can only fully live and express inside of Facebook’s ecology. Facebook’s distinctive informational holding is actually two things: a more nuanced view of its users’ social networks than most other data sources can provide and a record of expressive agency.
On the first of these: the social mappings aren’t easily monetized in conventional terms. Who needs to buy knowledge about any individual’s (or many individuals’) social networks? Law enforcement and intelligence services, but the former can subpeona that information when it needs to and the latter can simply steal it or demand it with some other kind of legal order. Some academics would probably love to have that data but they don’t have deep pockets and they have all sorts of pesky ethical restrictions that would keep them from using it at the granular level that makes Facebook’s information distinctive. Marketers don’t necessarily need to know that much about social networks except when they’re selling a relatively long-tail niche product. That’s a very rare situation: how often are you going to be manufacturing a TARDIS USB hub or artisanal chipotle-flavored mustache wax and not know exactly who might buy such a thing and how to reach them?
Social networks of this granularity are only good for one thing if you’re an advertiser or a marketer: simulating word-of-mouth, hollowing out a person and settling into their skin like a possessing spirit. If that’s your game, your edge, the way you think you’re going to move more toothpaste or get one more week’s box office out of a mediocre film, then Facebook is indeed an irresistable prize.
The problem is that most of us have natively good heuristics for detecting when friends and acquaintances have been turned into meme-puppets, offline and online. Most of us have had that crawling sensation while talking to someone we thought we knew and suddenly we trip across a subject or an experience that rips open what we thought we knew and lets some reservoir, some pre-programmed narrative spill out of our acquaintance: some fearful catechism, some full-formed paranoid narrative, some dogma. Or sometimes something less momentous, just that slightly amusing moment where a cliche, slogan or advertising hook speaks itself from a real person’s mouth like a squeaky little fart, usually to the embarrassment of any modestly self-aware individual.
Facebook could, probably will, eventually wear down its users’ resistance and stop labeling or marking or noting when it is allowing a paying customer to take over their identities to sell something to their social networks. We’ll still know that’s happening to a friend up until the day that an AI can use all that data to so convincingly simulate our personal distinctiveness that there’s no difference between the AI’s performance and our own. At which point, so what? Because then my accurately simulated self will just be selling or persuading on behalf of that which I would, with all my heart, sell or persuade, in the voice I would normally use to persuade with.
As long as Facebook’s potential customers want to use my social networks to sell something I wouldn’t sell, in a way I wouldn’t sell it, most of the people who “know” me through Facebook will know that it’s not me doing that, and they know that better and better proportionately in relation to the amount of information I’ve provided to them all through Facebook. (E.g., the best protection from being puppeteered is paradoxically more exposure rather than less.)
So what of the other unique information Facebook holds, a record of everything I’ve “liked”? Surely that’s information worth having (and thus worth paying Facebook for) for anyone desperate to sell me products, persuade me to join a cause, or motivate me to make a donation? Not really (or not much), for two reasons. First, because existing sources of social and demographic data are generally good enough to target potential customers. If you know who the registered Democrats with graduate-level education making more than $75,000 a year are in Delaware County Pennsylvania, you have a very good understanding of their likely buying habits and of the causes to which they are likely to donate. If you’re selling something that has a much more granular target market, it’s almost certainly more efficient and cheaper to use a more traditional media strategy or to rely on social networks to sell it for you simply because they’re interested in it. If you’re the budget-photography company YongNuo, you don’t need spend money to mine my Facebook likes and posts to see I’m interested in moving into studio-based strobist photography: existing networks of hobbyists and professionals are sufficient to acquaint me with your products. If you’re trying to sell a Minecraft pendant necklace, your potential customers are going to do a fine job of notifying each other about your product.
More to the point, if I’m trying to sell you a product or a cause and I find you through data-mining your pattern of “likes” on Facebook, what is it that I’ve found? Maybe not the “you” that actually buys things, shows up to political rallies, writes checks to a non-profit. I’ve found the algorithmic cyborg that clicks “like” on Facebook, half-human and half-interface, formed out the raw stuff of things that are clickable and linkable and feed-compliant. Which is sometimes a set that overlaps with what can be bought and done and given in the rest of our lives and sometimes is very palpably not. If my sales or success depended on the liking of Facebookborgs reliably translating into behavior elsewhere, I’d be on very thin ice. And I’d just as soon not pay much to get onto thin ice.
So what about the rest of social media? Do they have something to sell, something worth investing in? Sometimes they do, and that brings me back to Flickr and 500px, where I started this series. What Flickr and 500px have to sell, first and foremost, is not information but services: data storage, a form of publication, and access to a community with a narrower focus than “all of life”. Both of them have at least a rough model for how to make a bit more revenue on the backend, through facilitating the sale of community members’ images to any external buyers (while giving the creator of the image a cut of the revenue). That is not a business model that is going to make them megabillions, but it’s very likely a sustainably profitable enterprise when all is said and done. It rests on a fragile foundation, as Flickr in particular has discovered. Your paying customers have to care enough about the social capital they have invested in the service to pay for it, the publishing interface has to be updated to look contemporary and run on contemporary hardware, and the archive has to be searchable enough that external buyers (whether it’s someone looking for a canvas to hang on their wall or a media organization looking for stock footage) can sift through it. All of which takes work for a labor force that has to be kept lean and cheap. One slip and your users, the real source of your value, are going to pack their bags and content for the next new thing. When that starts to happen, it can cascade quickly into collapse. If you do something to try and slow the flight of content and participation, by making content difficult to extract or erase, you might spark the equivalent of a bank panic.
There’s one other social media business model that demonstrably works, if in the spirit of 21st Century financial capitalism: it’s the digital version of a pump-and-dump. Set up a specialized social media service, lure a venture firm or investor in that’s looking to bet a bit of money on the next new thing, spend a bit of money on an interface design, put on a dog-and-pony show that gets the restless digerati in the door and providing some kind of content. If dumb luck is really with you, maybe you stumble into the next YouTube or Twitter, you somehow provide a space or tool in a niche that no one knew existed. If dumb luck is sort of with you, you’re Instagram and you get bought up by bigger fish who need to prove to their investors that they’re working towards a profitable business model and are using acquisitions as a distraction from tough questions. In that case, your business model is to be someone else’s business model, only you can’t say as much without shining a spotlight on a naked emperor’s private parts. In the worst case (probably) you burn someone’s money, earn some salary, get some experience, and have a story or two to tell to your next investor–or at least build a resume that gets you hired at a real company.
Social media that provide a service that is sufficiently valuable that people will pay for it, however little, have a business model that is not only sustainable but that doesn’t require them to constantly breach the trust of their users or work against what their communities want.
Social media that have no business model except trying to monetize the information that users provide to them will, sooner or later, be required to breach trust and demolish whatever is useful in their service, to come back again and again with new interfaces and terms of service that lie or conceal or divert. Even if they get away with it for a time, they’re selling a product that is far less valuable than the innumerable info-hucksters and digital prophets (or even protectors of privacy) think it is. In some ways, it might be best if Facebook just got it over with and gave itself permission to sell every last scrap of information it’s holding: what we might all discover is that there’s hardly anyone at all who will pay for that service at the scale and price that Facebook needs them to pay.
Au contraire, marketers of most products and services DO want to know about the social networks of their actual and potential customers. This is because almost all goods, even so-called commodities, are subject to fashions in purchasing – what economists call network effects. People buy things in part or completely because others in their social network have, or wish to, or are perceived to wish to. Consumption is always partly social, even for commodities like coal. I have blogged about this topic here:
Like Peter, I think you’re underestimating the marketing value of the social graph. Very few people actually bowl alone, so it makes sense for a bowling alley to try and sell to an entire group of friends at once. I’d bet that a substantial fraction of all decisions on what entertainment to consume are made collaboratively.
You have underestimated how useful some of Facebook’s data is. There are many brands, hobbies, interests, and activities which have no significant existing networks to speak of and certainly have no usable “existing sources of social and demographic data”.
For example, I’m a musician, and I may want to advertise to fans of bands similar to my own. I can think of other place, online or offline, where I can target such an ad anywhere near as effectively as I could on Facebook.
How much that is worth to a 3rd party or investor, I couldn’t say – but there is definite value there.