Flam: Friends' Lovingly Annoying Messages

Spam is the name we give to unsolicited emails from unknown people. We shouldn't call spam what our friends send to us, but we have the same problem, that of having to wade through too many unimportant messages in a limited amount of time.

This thought reached me following an effort by Forrester Research analyst Josh Bernoff to audit all the PR email he was receiving. He counted 114 emails in the last two weeks of January; for each message, he analyzed what research area it addressed, whether a human had consciously sent it to him in particular, and how easy it was to unsubscribe. He concluded that three-quarters of the PR email he received was irrelevant. He did everything but give it a name. Without a name, it's difficult to get others track the problem, let alone suggest remedies for it.

I hereby call it flam. Backronym: Friends' Lovingly Annoying Messages. Well-intended (lovingly) but otherwise off-topic (annoying). (With apologies to the village in Norway by the name of Flåm. It looks charming; I hope to visit sometime.)

It's not merely a problem with PR agencies, or with email for that matter. The new frontier of communications on the Internet is status messages: brief notes, often with hyperlink. Twitter popularized this; Facebook realized from Twitter's success that status message updates were their under-utilized resource, and this past week re-tooled its web interface to encourage their use. Just like an email inbox, users now see a stream of status updates from Facebook or “tweets” from Twitter. What makes flam economical is the same as for spam: there is no monetary cost for a sender to reach more users. (Unlike spam, every piece of flam is the result of an “opt-in” relationship). There is, of course, the cost of attention,: the subscriber must spend increasing amounts of time scanning messages, or risk missing some. Or they can just drop the most flammy of their friends.

This has ramifications as more people turn to status messages, not just for personal communications, but for their professional lives as well. I will address some solutions below. First I wanted to explain how I came up with the term.

Etymology

“Spam” is a word of amazing economy. It is it is one letter removed from scam / sham. It has inspired anti-spam software, the CAN-SPAM legislation, and quickly made the jump from jargon to language. The New Oxford English Dictionary added it in 1998.

In 2007, some computer users in Pittsburgh at the PodCamp conference sought a word to indicate the automated emails that come from services like Facebook; they kept with the porcine metaphor and chose “bacn.” This even reached the New York Times as a buzzword for 2007, but it hasn't caught on very well in common language. Bacn doesn't easily conjugate to a verb (as spam has). While Hormel-brand SPAM is just as much an esoteric cultural artifact as it processed ham, the most common connotation of “bacon” is that it's on the breakfast menu. The website of bacn.com is used by enthusiasts of the food; the promoters of “bacn” had to settle with bacn2.com. The posted four posts to the blog in three days and effectively abandoned it thereafter. They listed no description of the problem, let alone any solution.

Ironically the problem, as originally stated, got the very thing we're solving here: “how some people we know twitter so much we have to turn off notifications for them.” That may have well solved the email problem, but it doesn't solve the problem of having one's Twitter feed clogged!

I wrote Josh Bernoff after he released his report last month to see if he he could suggest a name for the term. He suggested “pram” since he wanted to emphasize the PR nature. I suggested “spim” because it sounded like a weakened spam, as well as the word spin (also a PR connotation). I since learned that spim has already been used for spam-like, unsolicited IM messages. But, as above, I didn't think this problem was particular to the PR industry.

Cram? Possibly, the image of squeezing something in a tight space seemed to fit what I was after. Of course, cramming already has connotation in popular culture: studying all night before an exam – squeezing as much knowledge into your short-term memory as you can. Cramming and slamming were also adopted a decade ago by the phone industry to describe the practice of adding phone charges, and switching phone providers, respectively. (A common practice by competitive carriers after the 1996 Telecom deregulation; the incumbent carriers very swiftly were able to pass laws against this and.)

Tsam? I got this by taking the word ending of flotsam and jetsam, the junk from ships floating at sea. But TS- is a most unusual combination to start a word with; almost all of the words in English that do are from a foreign language (tsar, tsunami, tsetse, tzadik). Not that being of apparent foreign origin hurt the adoption of “wiki,” but nonetheless, I didn't think a foreign-sounding would catch on. And I couldn't come up with an easy backronym.

So I got back to flam. It sounds like flan, the tasty crème caramel dessert from Spain. Tasty, fluffy, yummy – but all in moderation! A flimflam is a confidence game – though it avoids the harsh sound of scam (flip, flap, fluff, flight, flake – these words all suggest levity). Rex Stout's fictional detective Nero Wolfe regularly favored “flummery” as an idiom to mean nonsense. And, as noted before, the backronym: Friends' Lovingly Annoying Messages. Flam it is.

I did pause to consider that the gerund form “flamming” looks like a misspelling of flaming, the old Internet pasttime of harassing someone online. A research paper from Patrick O'Sullivan and Andrew Flanagin of Illinois State provides a helpful examination of the practice. They suggest that the truest example of a flame is where sender intends to insult the recipient, and the recipient and third party perceive it as such. I am well aware that disruptive and injurious speech persist online (see Categories of Harmful Speech Online), but I sense it is becoming more the exception than the rule. The pre-social Internet of “no one knows you're a dog” ethos featured lots of anonymous users, and many others with unknown reputations (IT professionals, college students, etc.). It was easy to flame then and wreck civilized discourse. On today's social Internet, you often know very well whom you're talking with– their professions, their peers, their pets. Flamers used to dominate Usenet channels and bulletin boards. People moved elsewhere to more controlled forums. More conversations happen in blog comments, in which blog publishers have a lower tolerance for attacks. Twitter has ad hoc channels, but reputations are known, and violators are subject to a centralized code of conduct.

In other words, flaming will be yielding to flamming as a central communications problem of the Internet.

Problem & Solution

The flam rate is the amount of unimportant messages from a sender . If the rate is too high, you stop paying attention, and you unsubscribe.

On Facebook, you have reciprocal relations with all of your linked friends. You might have more common interests, and less flammery material. In addition, under the new design, you effectively unsubscribe from a flammy friend's updates – without sacrificing your friend connection.

Twitter was a bit innovative in tossing out reciprocity. You can follow people without them following you. Oddly, Twitter has never straightened out the terms for this. The inverse of follower is given as a following, which doesn't quite sound right: there is no plural. Hubspot's State of the Twittersphere Q4 2008 reveals that 88% of people have less than 100 “Number Following.” This is linguistically confounding, since it easily can be confused for “Number of Followers.” I would humbly suggest “fave” and use it here. You have followers, and you have faves (people you follow). If you have a reciprocal relationship, it can be described as a friend in the Facebook sense.

Therefore, flam can just as well stand for “Faves' Lovingly Annoying Messages.” Each status has meaning to the sender, but not to ever reader. The updates from a fave in an airport are meaningless to most followers, unless you are in the destination city as the traveler and would like to meet him or her (I have joked that someone should assemble a compilation book titled Idlewild: Bloggers and twitter users passing the time at airports.)

The flam problem doesn't just annoy current users. It also prevents people from joining a service in the first place. A recent poker gathering with some friends drew the sort of improper Bostonians crowd that Twitter would love to have aboard: technology workers, early adopters, VC advisors, iPhone users. None saw any particular compelling benefit to Twitter; all were under the impression that it would be a new stream of off-topic messages to deal with.

My long delay in adopting Twitter (17 months) was based on the fact that Facebook was superior in defining a rich vocabulary of updates (beyond status, users post events, photos, links, etc.). Granted, once I joined Twitter I realized that I could follow faves that weren't necessarily my friends. I could follow many more faves than I had ever done over RSS. I'd happily subscribe to a company or charity's Twitter feed if I never got another email or paper mail from them.

I just have one caveat: they need to construct their posts in a way that can help me address the flam problem. In Semantic Social Media Construction, I describe how different types of semantic updates could be coded. In Star Priority Notation, I provide some implementation examples for use in Twitter. (These will remain unused until greater appreciation for flam problem comes about). I'd also like to get set a threshold for how many updates in a day I get from a particular sender, lest my message inbox gets flooded with their flam. Asking them to self-prioritize their notes would drive which of those actually I get to see. Event planned? Yes. Something you've written? Yes. Something you're linking to? Maybe. You're in an airport? Only if you're going to be in the city I'm in.

Recognizing flam and keeping score will help – the trust problem. With the social internet, we have less worry about trusting people to give us reliable information; we have the freedom and ability to check multiple source. The question is whether we trust them with our time. The best measures we have on the Internet for gauging popularity (Google for websites, Technorati for blogs; Twitter Grader) are solely functions of popularity. The power law taxes effect: people follow channels simply because other people are following them. Including a measure along the lines of a flam rate might help readers make better decisions as to whom to follow.