Just caught an entry at BoingBoing about an accusation that the graphics card manufacture nVidia has been seeding online forums with people paid to endorse or recommend their products.
The prospect of such marketing alarms many people. I have no question that this practice, and similar forms of viral marketing, are becoming more common. I’m inclined to think it’s very interesting but not especially worrisome.
It reminds me a little of how “pavement radio”, popular forms of ‘rumor’, small fragments of narrative and news, is sifted in everyday conversation in parts of Africa. When I was in Zimbabwe in both 1990 and again in 1998, certain stories or narratives were commonly regarded as inventions of the government and discarded. Such stories didn’t last long in the evolutionary tree of rumor-knowledge, or if they were reproduced, they were framed as evidence of the folly of the state rather than for what the stories were literally trying to claim. Even to me as an outsider, it was easy to see what identified such stories. They contradicted everyday knowledge in some respect, their purpose was transparent, their design was crude, and they often could be found in nearly identical form somewhere in the official print media.
In short, most people were skilled at the work of everyday “textual” interpretation, whatever their background and education. Now there were, on the other hand, rumors which I (and some of the people I spoke with) believed to have an “interested” character, that either came from or were congruent with the strategic objectives of some powerful individual or group. For example, there were many stories about the “super Zezurus”, a sort of ethnic clique of powerful ruling party members gathered around Robert Mugabe, and their land holdings and wealth. Such stories certainly might have served the interests of other ruling party members trying to position themselves for a post-Mugabe world, or trying to push their way closer to the inner circle of power by pressuring the party to diversify its power base. But the stories were retold and found useful because in various ways they corresponded to what many people (in the city of Harare, at least) already knew about the social world around them, both in terms of the way politics worked and about the way they read the identities of other Zimbabweans.
So even if the story served someone’s interests, it was circulated in part because it had the ring of truth, because it had a correspondence to lived reality.
Obvious or crude attempts to seed online forums with commercial promotion are even less credible than officially sanctioned rumors. You’ve probably seen just such efforts, and consciously or not, known them to be such. Improbable enthusiasm for a particular product, speech that sounds nothing like ordinary conversation, a weirdly off-kilter kind of detailed knowledge about some commodity. Even a message thread with an anomalous number of postings to it. That kind of viral marketing isn’t going to serve any company well: quite the opposite.
A message thread loaded with shill promotions, whether it’s at eBay or epinions, is going to have the paradoxical effect of making any negative or even balanced assessment of a product vastly disproportionate in its influence. If you’ve ever read through an eBay merchant’s ratings to find a single rational, calm negative feedback message, especially one that concretely states that the merchandise was described poorly or was never shipped, that can have a huge chilling effect if you have the least concern about what you’re buying, even if there are 99 other messages saying, “Great seller! Superfast shipment!!!!” It’s because the negative message is the only thing with differentiated information in it. Everything else looks patterned, empty.
Now of course the same can be true for positive assessments in such threads. A positive review at epinions or Amazon that is detailed, naturalistic in its speech patterns, by someone who seems to know their stuff but also includes criticisms or gestures towards the overall market for a particular type of good, may clinch a sale. Some viral marketers know this, and so now what’s being seen more is shill postings where the shiller starts off trying to talk about normal, everyday subjects, or to sound more like the average online writer. That still doesn’t work: someone who starts off saying, “Man, that King Kong was a great flick! And the Steelers rock! But I’ll tell you what, my new video iPod is really the best thing that’s happened to me lately” still sets off warning bells, especially if that person has no built up reputation capital or known identity within the community where the posting appears.
Let’s suppose, though, that the viral marketer manages to create an absolutely credible, spot-on imitation of the posting that a real human with real opinions and experience might offer. That’s still nothing to worry about, because at that point marketing and truth begin to coincide. If a viral marketer plants a post that says, “Well, I really enjoy my iPod, because it looks really nice, it’s easy to use, and I can put so much of my music on it, but on the other hand, the cost is ridiculous, Apple’s digital-rights management is a scandal, the quality of AAC is low, and it’s really fragile given how expensive it is”, then they’re welcome to shill-post all they like. Subtract out of that what many iPod users know, or are coming to know, and the message loses its authenticity. Viral marketing that aims to deceive gets only one shot before it’s useless: you get the opening weekend for your movie, no more; you get the one purchase for your commodity, no more. After that, you not only lose your customer, you degrade the information channel you infiltrated. Some companies might be short-sighted enough to settle for that bargain, I suppose, but that’s their problem (and the problem of their stockholders).
The only time it’s worth really worrying about viral marketing is when a government tries to do it in the national or international public sphere. It’s not such a problem when it’s a government whose complete and total disconnect from its public is a given, as in the case of Zimbabwe: then such behavior is merely a symptom of a deep and fatal disease. It is a problem in the case of a government that inherits a healthy or at least functioning public sphere, a government that’s supposed to practice some kind of democratic transparency. Not only can viral marketing make the body politic sick, but it almost never achieves its instrumental ends over the longer haul. You can’t create a truth which has no roots in what is already known and lived; you can’t shill an alien falsehood into being simply by hiring consultants or setting up bureaucracies.
This is not to say that people in general couldn’t become even more expert or skilled interpreters of information than they are. In fact, viral marketing is one of those things that I’d say a program of courses in the humanities ought to be aimed at. Not in the relatively crude way that some left intellectuals might advise, through some polemical course or program of instruction. Learning to read and interpret anything, with the fullest battery of techniques and critical skills, would lend itself to discriminating engagement with all kinds of information in public circulation. A course in 19th Century British literature ought to sharpen a student’s ability to sift through customer reviews on Amazon looking for the one informationally-rich review in a sea of shill promotions and morons with grudges. I think many people can do that fairly well already, but practice makes perfect.
I don’t know Timothy, there’s nothing new about rumors and intentional leaking of both information and disinformation. If that makes societies sick then all of them always have been sick. This is so, but can it be otherwise?
What I see is a breakdown of the broadcast model due to technology change: 57 channels and there’s nothing on. It affects media, marketing and politics in much the same way, the decline of brands as some see it.
I think you’re missing a further class of shill, and this is the sort that gets people (okay, me) hot and bothered. Your “reasonable” viral marketer wouldn’t say anything near like what you posited; instead, it would be more like, “Well, I really enjoy my iPod, because it looks really nice, it’s easy to use, and I can put so much of my music on it. I know jfpred34 is always whining about the DRM on it ;-), but I don’t care, because I only have one computer and my ipod anyway, so I don’t care if I can only use it on 3 devices at a time. Also, check out this sweet ipod cover I got for it!”
This approach is much less likely to get our marketer tagged as a shill, especially if he’s been posting for a month or two by then, and especially if he sticks around for a little give-and-take with jfpred34 after the initial post. I’m certainly not positing that the audience for these guys is necessarily more sophisticated than the streets of Zimbabwe (one look at my inbox would put that to the lie!), but I think the people targeting them are.
From the linked article:
>I interviewed for a guerilla marketing business in San Francisco that targeted web forums.
>I was told that if I accepted the job, I was to have at LEAST 50 identities on as many forums as I could muster (they wanted 100 eventually), with a goal of 5 posts an hour. The posts had to be well thought out, and the idea was that I was to establish multiple identities with a history on the forums, so that when the timing was right a well written but subtly placed marketing post could be finessed in. And regular visitors would recognize the post as coming from a long time poster.
I think that this type of approach could be relatively effective and is something worth worrying about. I have bought things (mostly books) based on people’s recomendations on the internet. I generally trust that people are giving their real opinions on things on the internet.
>Learning to read and interpret anything, with the fullest battery of techniques and critical skills, would lend itself to discriminating engagement with all kinds of information in public circulation.
This could be true, but it is still better to arrange things so that people don’t need to use fancy critical skills. The debasing of trust in ones fellow citizens isn’t to be encouraged.
Plus, it is easy to overestimate the extent of ones critical thinking skills: “My guess is that the righteous indignation we saw on last week’s show comes from Oprah representing the less experienced readers who needed Frey’s memoir to be true in a journalistic sense.” I-am-not-suprised- when-anything-I-read-turns-out-to-be-a-lie isn’t actually critical thinking.
I agree that blanket cynicism isn’t critical thinking. I’m more talking about the kinds of concretized interpretative skills that good teaching and scholarship in the humanities can demonstrate and transmit.
The thing for me is that when a shill post approaches something that might credibly be said to be by a real human being with real experience of a product, you get to an almost Borgesian juncture. Take Emschwar’s example: the sentiment itself is pretty much what someone might actually say. Somehow it bothers us to think that someone is saying that for money rather than simply trying to share information, and it should, but the issue that concerns us there is not the content of the information and what effects it might have on our behavior. The issue is about something else, about the motivations of other humans, about the authenticity of communication.
You could not “shill post” people into making a bad film into a blockbuster, into making a serious lemon of an electronics product into a runaway success. Even something like the iPod, where I take its popularity to be at least partly about image, marketing and advertising, given its weaknesses as an actual product, has succeeded for “real” reasons as well as a marketing hook–its simplicity of design and utility, its connection to a well-developed and unique content-supply model, its aesthetics, etc. The only way I think you could shill post something genuinely bad or defective into seriously success would be to somehow completely control an information channel that readers take to be meaningful and open–say, for example, to buy off every single registered user of epinions or some such. Otherwise, what’s going to happen is that a single bad review that has substantive content is going to tip off many readers to the existence of shilling, even if the shilling is being done with great skill and effectiveness.
Most people generally like, or are predisposed to like, the things that they buy. Amazon’s recomendation engine just assumes that this is true. I can see how the few negative reviews are given greater weight because they counteract this tendency. But, this also means if people can be persuaded to buy a sub-optimum product, they will probably like it.
I notice that I tend to buy books that I see two independent sources of recomendation on the internet. The approach that is described in the article of posting on multiple boards to get a reputations under multiple screen names, then producing multiple recommendations of a product would totally work on me. Not that it would sell enough books to make this approach worthwhile, but they are faking the the kinds of things I use to make my purchasing decisions.
Even as the top end of phishers are getting better at disguising their messages, so the top end of marketing shills may be getting better at disguising theirs. And if they’re good enough, you won’t notice that there’s shilling going on. It’s a Boolean recall problem: how can you evaluate shills if you can’t identify them all?
I’ve also had the weird sensation of wishing to honestly and genuinely praise something, but of being aware that I might sound like a shill for doing so. Best way to avoid that is to include in one’s comments a neutrally-phrased description that doesn’t sound like a marketing brochure. But one doesn’t always have time to do that.