The Impact of Fake News on the Internet of Things and Other B2B Conversations

Ahead of the news from Facebook on Cambridge Analytica abusing the data of American citizens, Twitter started suspending accounts for a practice called Tweetdecking, where accounts would solicit and at times pay for accounts to artificially boost their classic influence metrics of reach and followers.  Over the weekend of 18 March, some of the Internet of Things most prolific influencers were also suspended, at least for a while.  Many of them are back on the Twitter-waves, pushing content like before but the artificial amplification they received by Tweetdecking has diminished to mortal levels.

At Argus Insights, we have been doing our own analysis of the rampant pay to post shenanigans in the various B2B markets we track.  I first identified the issue when we saw Brocade jump in 2015 after a product announcement.  Our client at the time, HPE, was concerned about the amount of attention that Brocade was getting for their announcement of offering free Network Functional Virtualization solutions.  We dug into the data, initially by hand, and found the bulk of the lift Brocade demonstrated came from two sources, their own employees (Corporate Narcississm) and what appeared to be bot accounts from Saudi Arabia, tweets from people that had no interest in the telecom space, based on their past social engagements.  In short, someone, maybe Brocade, maybe their agency, or someone else, was paying to push their message out, inflate their metrics and give the perception that the whole market cared about their announcement.  Turns out the market didn’t care.  Without the bots and employees boosting the interest, the chatter around Brocade died down to normal levels almost the next day.

We had a client looking to boost their followers (the boss challenged them to beat his follower count in a few weeks) and ignored our advice to build their following organically.  They instead spent money on Twitter ads and saw their followers go from 92 to over 40,ooo in just a few weeks.  The boss cried foul and asked them to prove these were real followers.  They came back to Argus Insights, hat in hand, and asked if we could help.  By looking at which of their followers had actively participated in the market (NFV) in the last six months, we could say which of their 40k followers were likely to be real.  It was only 271.  Of over 40k followers they had grabbed with their ad campaign, less than 1% mattered to their business.  The client started managing to their True Followers metric instead and saw their authentic influence grow, even as their overall follower count dropped.

But not everyone that participates in a market is part of that market.  In November of 2017, after being frustrated with the amount of poor content that was topping the charts of our analysis within IoT, I developed some metrics to gauge whether an account fit into a few different categories.  Were they a brand, pushing out content mostly from a single domain with a good level of active dialogue with the rest of market?  Were they a broadcaster, just sharing content from others?  We also identified content farms, accounts that tend to talk about themselves a lot and retweet content of their clients.  The most nefarious type were the compromised accounts.  These accounts are basically owned by content farms and seek only to artificially boost their ‘influence’ in the marketplace.

Once we had applied this account types, we found that over 75% of all IoT content was published by the compromised accounts.  This means that 3 out of 4 tweets about a multi-billion dollar market are not authentic and serve only to misguide and misdirect the 25% of the content that is more likely legitimate.

It gets worse.  One of the most prolific IoT influencers pushed out almost 700 tweets, over 10% were self promotion, a clear sign of a content farm.  More nefarious is that of his thousands of retweets, 85%, eighty-five percent, came from the aforementioned compromised accounts.  This means that only 15% of his “influence” is legitimate.  This means his clients that rely on his reach to bolster their own market awareness are paying to push content to compromised accounts.  Not customers, not influencers, not thought leaders, but accounts owned by others whose sole purpose is to buy and sell attention from others.

B2B marketing on Twitter is broken.  Broken by those that would game the system and misrepresent their own influence.  Broken by those firms that pay for influence rather than earning it.  Broken for those who see Twitter as a source of what is happening in their market.  Broken for those looking to see what trends are driving their markets.  Broken for you…