Spam, Microspam, and Social Problems
I just watched a really interesting situation unfold on app.net, the nascent social networking site. For those of you unfamiliar with it, app.net is a very interesting experiment in social networking, designed to be more of a framework, than a networking site itself, set apart by the fact that users have to pay $36 a year for their account.
This pay wall (to post, not to read) has had the side effect of largely keeping out spammers, because paying $36 for an account that will soon be banned or universally muted simply doesn’t pay off. Compared to a place like twitter, or any mention of a keyword like “iPhone” or “mortgage” gets you spammy mentions from five or 10 different robot accounts, this lack of spammers is a wonderful change, and it’s something that people are jealously guarding.
However, today, a very small firestorm (“App.net’s first spam post?!?”) erupted when an account posted the below message advertising some sort of app.net aggregating service that he’s developing (links removed):
VIEW’s have grown to 1000+/day
Contact me to build VIEW’s into your app
-link- German @view5
-link- Japanese日本语 @view6
-link- Arabic العربية @view14
[This part is a list of @ mentions of 10 prominent app.net develops,and by mentioning them, that message made its way into all of their streams whether or not they follow that particular individual.]
What is spam?
Most of the people we would consider spammers operate in a very distinct way. Their messages are not targeted, but broadcasted. Their identities are hidden, not shared. Their goal is not quality, but passing through the filter systems out there. They take over other people’s computers to improve the number of messages they send, with the hope that if they send enough messages, some of them will get through filters, and some small subset of those messages will result in a favorable outcome (a purchase, or perhaps a click on a link).
I don’t think that’s what happened here, with this individual message. This person clearly is a person, who paid for an account and uses that account for other purposes. He made no effort to conceal his identity, nor the origin of the message. It would appear to be human composed, if not copy pasted, and in this person’s mind, he is targeting these individuals because of their clout on the service. Perhaps most importantly, this person does not consider himself a spammer, and defended himself against that accusation:
It’s not spam Paul.
I’m a serious data structures engineer. Not everyone is building apps … nor should all the engineers be building the same Twitter-clone apps.
A non-redundant approach to building structure of the ecosystem.
Yet, upon reading that first message, my first instinct was to check his profile to see if he was sending that same message to everybody else, and to confirm my suspicions that he was really a spambot who’d gotten onto ADN. His message looked like spam, was formatted like spam, was addressed like spam, but somehow, knowing that information about the author, his intent, and the small scale of distribution, I could no longer call it such
Who hasn’t signed up for a local businesses’ mailing list, only to find yourself getting two messages a day advertising every new product put on the shelf? Similarly, everybody knows somebody who’s gotten involved with some kind of marketing scheme (e.g Mary Kay, Amway, Tupperware) and who, to protect their bottom line, simply won’t shut up about it. And everybody’s got that family member who is heavily involved in some campaign or another, and is constantly forwarding you emails about some political candidate, some refugees, or some local issue that you never asked for.
All of these things, if conducted by some random Romanian, would assuredly be considered spam, but coming from somebody in your community, it’s more forgivable, and we’re reluctant to use the s-word in classifying it. Knowing the person, knowing their intent, and having some idea as to why you were targeted, all somehow make something not spam.
This is a fundamentally different problem, this is an issue of person-to-person interaction and conversational norms which has been around for a long time. The difference is that over email, or over app.net, the social means of enforcement aren’t present. At the dinner table, you can politely ask your Tupperware saleswoman aunt to leave business for another occasion, and there’s a general understanding that there’s a limited time and place for such advocacy in social settings.
This kind of micro-scale spam (microspam?) is a social and behavioral problem, and is one that all social networks have to deal with, whether it’s app.net or your neighborhood book club. In the book club, it’s easier for one person to pull another aside for a quick chat, and usually there’s some person in a leadership position who will naturally do just that. Online, that’s usually not the case, and there’s not the same decorum that prevents a large number of users from jumping on our unfortunate microspammer, throwing wild accusations and writing lengthy blog posts about it (guilty as charged).
Microspam is fundamentally different than Capital-S Spam as it exists on the Internet. Spam can be addressed technologically. It can be filtered, it can be shut down, and simply disconnecting the spammers from the Internet would likely cause a net improvement for everybody involved. Microspam has to be handled socially and culturally, because simply kicking your aunt off of the Internet will have undesirable consequences, and you run the risk of silencing a productive member of the discussion who has otherwise interesting things to say, in between bouts of microspamming. Painting this micro-level with the same brush as your large-scale spammers won’t help in the battle against either phenomenon.
Unless, of course, your Tupperware selling aunt also happens to be a Ukrainian spam kingpin. Then you can solve both problems at once.