Sunday, March 1, 2009

A Closer Look at Twitter Follower Faking

Mike from MPD57 Blog wrote to mention some interesting posts set for this week, and I encourage you to visit.

As often happens, we got to talking. Mike is a good guy to have around, because he is skeptical and asks questions that make you think carefully about your assumptions.*

He wrote:

"Thinking back to your twitter fakes article. Why would anyone want to create fake followers when they're so easy to come by (real ones that is). Such strange behavior for so little gain. That's what I find disturbing. That anyone would go to those lengths for so little useful purpose."

A great point. I bet others have had thoughts along similar lines.

The answer is quite interesting and revealing. I'll start by referring you to the original post: Fake Data and Remarkable Patterns in Selected Webcomics. You may want to read it if you missed it, though it is longer than an average post.

First, remember that Twitter cares about fake characters a lot more than I do (source: Twitter staff). I'm interested in how fake Twitter followers tend to flag people as cheaters in other areas. Twitter cheaters are more likely to be engaged in shady behavior, like inflating their web site traffic count.**

For new arrivals, let's remember: Fake Twitter followers are when you log out, create a new identity, have it follow you and perhaps your pals, then log back in with stats showing you looking a bit more popular.

It does not include spam, or fake characters someone put in your profile without your permission, or characters you run to represent your comic (but everyone knows it's you). Fakes are intentional creations for self-gain.

So, to Mike's question. It certainly does seem like a lot of effort for nothing.

The first part of the answer is a minor point: when you get smooth at something, it goes faster. Setting up a second browser, bookmarking friends' profiles, knowing where to pick up a lot of targets fast: those all help.

The main part of the answer is more interesting. I think Mike is right, and that few people would make fakes at the apparent rate of return.

For our most prolific fakers, however, the rate of return is actually much higher. They don't simply put themselves on a fake's follow list, because that would look fishy. Picking strangers might draw unwanted attention.

So, they pick friends. A typical fake might have five, ten or more pals on its following list. Those pals will get the email (unless they've opted out) that "FAKE EXAMPLE" is now following them on Twitter. You can often tell by the aliases and other subtle details who made the fake.

These pals make fakes too, so when they have a fake follow people, they remember who made fakes for them, and add them. One fake following ten pals might result in those ten people adding you to their next fakes. With a big enough circle (and it is big) it's not out of reach to make two fakes but get twenty followers back.

That's where it gets interesting to people who like being celebrities, being in charge, and being thought of as speaking for everyone else in webcomics.

We might call this the "tattoo effect." I don't know if it holds true any more, since tattoos have become more popular, but for decades criminologists knew that one of the only reliable indicators of whether a person was more likely to be a criminal was whether they sported a tattoo.

The tattoo effect on Twitter is that it's a quick and surprisingly effective way to separate professional, socially interested, honest cartoonists from those you shouldn't leave unattended near your grandmother's jewelry.

Mike's question also reminds me of another question, which is why cartoonists with a lot to lose gamble their reputations (though some gamble with more enthusiasm than others).

I don't know the answer to that with certainty, but I can tell you the direction the data is trending.

It's quite simple. It looks like they don't have nearly as much to lose as it once appeared.

A sweeping recalculation of webcomic popularity may require a substantial rethinking of long-held assumptions, and a major reordering of some people's top lists. (Might be wise to slow down that "Best Webcomics of All Time" project at ComixTalk lest it prove embarrassing in unanticipated ways.)

This is why, when some of these blowhards descend here the first thing they do is try is to convince me how much happier I will be if I simply drop the matter and draw pretty pictures with my wife.


* Easy to do these days, though the search engines are out to crush it. Cheat scripts and devices are cheap and easily available, and you'll be noticing a lot more comics going from 5,000 readers to 250,000 in just a few months before it's resolved. Making traffic unreliable, of course, makes advertising impossible, which does major damage to Google, Yahoo and the internet economy. Word is, Google bans anyone caught cheating. If some site drops off your radar, that may be why.

** I can't tell you how lucky I am to have smart, skeptical cartoonists who take the time to critique my work. They improve the quality immensely, make me strengthen weak spots and suggest avenues of exploration I missed. Even a few who help only occasionally have been of major assistance, and all of them work jobs, do challenging comics and still make time to help. Cheers to 'em all. After we get past the stage where people like me get blackballed, I'll see if we can't coax a few into the spotlight. (Blackballed is too strong a word, but you do get harassed for reporting facts some people want kept secret.)

 Subscribe in a reader