A growing percentage of people are rejecting “official” sources of information about computer products, relying increasingly instead on user opinions to make their buying decisions.
It’s not perfect, but for more than a few aspects, seeking user opinion of products is a wise idea.
That might not last much longer, though . . . .
Say Hello To Stealth Marketing
You only want to hear from real people, not fake marketeers. Well, the marketeers’ answer is to is to
fake being “real people” to sell their products.
Unfortunately for marketing, the current inadequate state of robotic design means you still need wetware to fake being a type of person.
Fortunately for marketing, cyberspace has evolved past such anarchronisms, and they are starting to take advantage of this, through software offered by companies like Activebuddy.
A number of years ago, the New Yorker had a cartoon with the caption, “On the Internet, nobody knows you’re a dog.”
No danger of canine comments in your favorite forum, but what about other nonhuman comments, from silicon- rather than carbon-based beings?
They’re already among us in cyberspace; they’re called bots.
The software used by ActiveBuddy is essentially a very sophisticated bot.
Computer Evolution Meets Human Devolution
If I have technology sophisticated enough to keep teenage girls coming back for beauty info, I certainly have technology sophisticated enough to keep teenage boys coming back for computer info.
The problem with bots is that they often don’t “comprehend” the question well enough to give an appropriate response. The wider the potential range of subjects, the worse it gets.
Bots do better when the subject material is restricted. If anything, the range of subjects within a computer forum is narrower than what that teenage girl bot has to deal with, and indeed, can be made even narrower. If the bot just works in, say, socket A motherboards, no one would find that unusual.
Now imagine our appropriately reprogrammed bot being unleashed on a computer forum. A bot posting in a computer forum would have a far easier task than talking to a teenage girl.
A bot talking to a teenage girl has to be interactive. No matter what the girl says, the bot must come up with an answer. In a computer forum, a poster doesn’t always have to post a reply to a question, sometimes yes, always, no.
Secondly, in a conversation, people expect a pretty high degree of relevancy and sense in a response. No one will ever accuse a computer forum of rigid relevancy or reason to the subject at hand :), so an occasional irrelevant or even nonsensical response is much less likely to attract attention.
Matters that might disqualify the bot in other circles, like grammar and spelling, hardly matter in this arena, indeed, this might be yet another area to dumb-down a bit.
Personality matters little, also. Build a degree of excitement and enthusiasm into the responses, and that will do.
Indeed, I suspect the bot might have to be “dumbed-down” a bit to fit in better, or at least have “Cursing Out People Who Disagree With You” modules.
So we have increasingly sophisticated artificial intelligence programs facing relatively unsophisticated natural intelligence systems.
(Sorry, folks, but when you have forums specializing in restricted areas of subject with high levels of irrelevancy, a prevalence of short answers, and often low levels of logic, attention spans, consistency and grammar, a bot is going to have a much easier time of “passing” than in, say, a forum dedicated to literary criticism. (You might also say that there’s enough fanboys already out there out who post just like bots.) It’s just easier to do, and any mistakes made aren’t as obvious as they might be elsewhere.)
We have the technology.
Now imagine using that technology and not telling anybody that you are.
You’re marketing for a mobo manufacturer. You know that mobo buying runs in streaks; a particular board becomes the “hot board” and most people buy it for a while.
Executing marketing offensive using the review websties is old hat to you, but you find that while it’s essential, it’s not quite sufficient to guarantee success. Word-of-mouth is the other main marketing element to success.
So how are you going to get that? Send in the bots.
Imagine sending in fifty or a hundred bots into a few key forums, with at least initially a little knowledgable human supervision.
I think it’s doable. Very, very doable.
Frankly, the areas where the marketeers are likely to get caught fall within shortcomings of natural rather than artificial intelligence.
The bots we’d be likely to see would probably sound a little too correct, a little too straight-laced, and more than a little too much like a salesman.
To do it right, the programming would have to be closely supervised by someone with a background in forums rather than marketing. Having a wide range of differing responses multiple bots could give to the same inquiries wouldn’t be difficult; it would just take more work and time and cost more.
Finally, you’d probably want a human being monitoring what was going out, sift out the more ridiculous responses, and provide a human response occasionally. You’d also have to put that person on a very long leash and allow occasional less-than-glowingly-positive responses.
All this would either cost more and/or put a few bureaucratic noses out of joint, so what would probably happen is that the company would cut costs (like maybe have all the bots come from the same virtual place), and end up losing ten or more times the money they saved when the project blew up in their faces.
But done right, I think it’s a scam that can be pulled off.
And that should scare you.