Beware of blockchain projects that brag about the size of their social media following. In the blink of an eye they might disappear. That’s what happened this week when Twitter removed fake followers from its statistics. Specifically, they removed “locked accounts” that were not bots exactly, but instead were accounts that Twitter suspected of being hijacked or otherwise misused. Most people and companies on Twitter lost a few followers, but this does not demonstrate anything nefarious. For example, although President Donald Trump lost 307,606 followers, that was only a .58% decline. However, The New York Times confirmed that clients of a company that sold Twitter followers saw dramatic declines this week, which proves that some companies got caught trying to inflate their popularity.
Which brings us to cryptotwitter. After looking at its internal data, Cryptoinfluencers.io reported that “truly influential people were not affected much.” Yet, they also noted that the largest drops were among business accounts like Ethereum Labs, which lost 56% of its followers almost overnight. This made us ask which other entities also saw a huge decline.
At the end of June 2018, Strategic Coin collected Twitter data about half of the 4,000+ tokens, coins and blockchain companies it tracks. Using this sample, we went back and looked at 964 accounts with over a thousand followers. We were reassured to find that most tokens and companies did not see a dramatic decline. We saw several accounts with double and triple the number of followers in the last few weeks. Whether or not they bought or earned those followers is the topic for another story. Instead, we dug deeper to find 45 entities, 14 of which are tradable on a crypto exchange, that saw dramatic declines.
Since late June 2018, GoHelpFund lost 77% of its followers and Cryptotask lost 68%. Ironically, two of the the other publicly traded coins seeing big drop are in the business of promoting trust — Verify (-63%) and VerifyUnion (-33%).
We do not know for sure why the number of followers declined, but the drops are most likely the result of Twitter’s actions. There are other possible explanations. Many Twitter users follow others in the hope that action will generate a reciprocal auto-follow enabled by commercial social media software. For example, a hypothetical company could follow all 12,573 Twitter users with #ethereum in their bio, which would result in maybe a few hundred new followers. However, if those same users were unfollowed, most would eventually drop from the hypothetical company’s numbers.
Strategic Coin looked at the team members and partners of most of the tokens included in this article but could not identify a connecting thread between them. Nor was there a significant difference in quality based on rating on ICObench. If any entity has a good explanation about why their followers dropped, please send an email to lh@strategiccoin.com.
Many token ratings sites include social media statistics as a way to measure community interest. Metacert founder Paul Walsh spends his days protecting ICOs from fraud. He told Strategic Coin that we “wouldn’t believe how many investors and exchanges rate crypto companies based on their Telegram numbers.” He warns that this is foolish because “almost all large Telegram communities are paid for. Although Strategic Coin has not independently validated this claim, reporting by Bitcoin.com points to ways bots utilizing stolen phone numbers can inundate a Telegram channel with fake users. To date, Telegram has not add
Most industry pundits know social media analysis is a seriously flawed method of assessing a project’s popularity and community strength. Yet, this does not stop anybody from collecting the data anyway. In research we have written about before, Hugo Benedetti and Leonard Kostovetsky use the number of followers as a crude approximation of a coin’s end users but found little statistically relationship with an ICO’s price. However, they note that a coin’s market capitalization increases at a greater than 1-to-1 ratio for each additional Twitter follower.
Measures of activity are more strongly correlated with the rise and fall of ICOs. The aforementioned research found that Twitter activity rises while an initial coin offering or token generation event takes place. More successful token launches see more Twitter activity, but the activity is just as likely to be the result of the community’s excitement about the token rather than a result of a successful social media campaign.
Twitter activity starts to decline precipitously soon after an ICO is completed because the need to communicate with the public dissipates. Many projects focus on executing on their roadmaps and do not have a lot of information to share. Another, more negative, interpretation is that a total cessation of tweeting means the entity has failed. In fact, there is some data to backup this assessment. As we wrote about previously, another academic paper looks at social media’s impact on the risk an investor faces when making an investment in a token. That study shows that a decline in Twitter activity is associated with a statistically significant increased chance that a token’s price will drop at least 75 percent in the first one, three or twelve month period after it is listed on an exchange.
Several companies believe that social media activity should be included in investment decisions. For example, Solume.io looks at the volume of engagement to measure sentiment. However, because each company or token has its own unique set of circumstances, we believe this type of data creates crude metrics that do not create a lot of alpha (outperforming the market). Instead, as mentioned above, these statistics may be better at uncovering beta (riskier than the market).
Founding members of ICOs should be careful how they manipulate their social media following. Tactics used to promote your personal accounts can backfire when applied to a more public entity. Make no mistake about it, although cryptocurrency and tokens may not (for the moment) be regulated like securities, they face many of the same public relations expectations.
If you decide to hire an ICO marketing service, look for a reputable company that does not promise quick results or pay-for-followers services. Adding additional Twitter followers can increase a company’s exposure, but this is only matters if the followers are real. Make sure to track actual engagement — likes, retweets, inbound inquiries — resulting from a supposed increase in audience.
For stakeholders of the tokens and companies included in these tables, be careful not to assume that large changes in followers means something unethical was done by management. Especially for the publicly traded tokens, management may be focused on day-to-day operations, while outside investors not associated with the actual project may be trying to prop up their holdings with misinformation.
Finally, if you are considering participating in an upcoming ICO, don’t think everyone is involved in a scam. Instead, make sure to do more due diligence to validate your investment premise.
Feature image: “The Naked Emperor” by Edward von Lõngus
Originally published on Strategic Coin, a website that is no longer operational.
