Instagram Strategy to Radically Reduce Traffic: Kill all the spambots!
RIP to my fallen robot followers on Instagram, if there's a heaven for robot instagram users, you guys are in there
— alldaychubbyboy (@Allday)
How do you scale to handle increased user traffic? Have less traffic. No, this is not a koan. The best way to deal with traffic is not to have it.
In a two day span Instagram disappeared 18.9 million users or more than 29 percent of their "followers." Justin Bieber lost 3.5 million followers (15 percent), Kim Kardashian lost 1.3 million followers (5.5 percent), Rihanna lost 1.2 million followers.
Instagram explains this dramatic reckoning was achieved by "removing deactivated spam accounts and accounts that violated its community guidelines."
In an age when high user counts and tantalizing engagement metrics are more valuable than bitcoins, this can't have been an easy decision, but it was made after being bought by Facebook.
Why? Gabe Madway, an Instagram spokesman, tells us why: We totally get that it’s uncomfortable for people. The overall goal is we want it to be perceived that the people following you are real.
Uncomfortable is an understatement. A BuzzFeed article nicely captured some of the anger, here's just one example (could be NSFW):
These sentiments were emphatically repeated by many victims of the purge. Facebook has always been a real identity refuge, so maybe the drive for Instagram authenticity is part of that vision? The thought being real identities are the key to more profitable ad campaigns.
The dirty little fear in the industry is that advertisers are paying a lot of money to advertise to spam accounts and someday they may get tired of it and revolt. Though other industries have long had the notion of dead money as just part of the process. Retail businesses expect a certain amount of shrinkage, that is loss of products due to damage, and, well, theft. Wine makers have their Angel's share, which is the loss of wine due to evaporation during the aging process. Farmers have what they call God's portion, which is the 20% or so of crop loss that is lost to animals, insects, and bad weather. So farmers always plant extra and businesses make allowances. Maybe spambots are like that? That is until most of your users are spombots.
There's a flourishing industry on the supply side of followers and likes. Molly McHugh wrote a great article showing just how easy it is to pump up your follower accounts. From You want likes? We got likes! Getting dirty in Instagram’s seedy spambot underbelly:
But it works. It really, really works. After running Instamacro for 30 minutes, I opened up Instagram to see I had 79 new likes and six new followers. Five minutes after that, another 19 likes. Whatever this says about my popularity, I’ve never seen numbers like that. Of course, a great many of them came from accounts that had bot written all over it. To be fair, many of the accounts were real people and brands as well – which is perhaps the craziest part to me: That reciprocity is that effective on Instagram. Simply by fake-liking these users’ photos, they liked mine back, and many started following me. Some of them liked more than 10 photos, only because I’d, assumingly, done so for them.
It may not be as bad as it seems. Zach Allia created a great chart of the top 100 accounts on who lost what. While there were some biggest losers, on average the loss was only 7.6 percent, it's just that Instagram has so many users the absolute loss numbers appear huge.
What's curious is how deeply users feel about the loss of fake followers. In the cult of popularity numbers are real so the losses are experienced as real. Social media is largely a fictional world to begin with, a mythic realm, where we present a particular version of ourselves to the world. So everyone has entered a willing suspension of disbelief when it comes to what we see on social media and that extends to follower counts.
The great spambot purge was a slap in the face, an uncomfortable reminder that social media is in the real world after all.
What's missing from this story is how Instragram went about identifying spambots and their generated content. Hopefully they will give us that story someday. In searching through HighScalability I noticed spam and fraud handling are often talked about, but little is given in the way of detail. In the "Related Articles" section of this post you'll see references to some specifics.
If anyone has good sources on spambot related infrastructure please let me know. I do know it's a huge part of any popular system and lots of resources go into handling spam, fraud etc., but it's little talked about in detail.
Spam fighting spells are a technology forged deep in the heart of Mordor, using a deep alchemy, hidden knowledge and secret means.
It would be nice to change that, for in the on-line arms race we can only expect faux people to get better and better at being faux. We now know it may not be so hard to pass the Turing test after all. A disturbing thought is we may not really mind the faux world being built by spam world builders. It can make each of us a star, but only if we choose to believe.
Related Articles
- Large Scale Spam Detection
- The Instagram Architecture Facebook Bought For A Cool Billion Dollars
- Instagram Improved Their App's Performance. Here's How.
- Fighting spam with BotMaker (Twitter)
- The Faving spam counter-measures
- Scaling Spam Eradication Using Purposeful Games: Die Spammer Die!
- Paper: Spamalytics: An Empirical Analysis of Spam Marketing Conversion
- Mollom Architecture - Killing Over 373 Million Spams At 100 Requests Per Second
- Mailinator Architecture
- 11 Common Web Use Cases Solved In Redis
- Reddit: Lessons Learned From Mistakes Made Scaling To 1 Billion Pageviews A Month
- Can Cloud Computing Smite Down Evil Zombie Botnet Armies?
- MocoSpace Architecture - 3 Billion Mobile Page Views A Month
- Prismatic Architecture - Using Machine Learning On Social Networks To Figure Out What You Should Read On The Web
- Social Media Bots Offer Phony Friends and Real Profit
- Why do we have such a love/hate relationship with cyberspace?
- Exploration of gaps in Bitly's spam detection and relevant counter measures
- Spam Detection on Twitter Using Traditional Classifiers
- Large Scale Deep Learning