Twitter data suggests safemoon and BNB supported by bots

Twitter data gathered from the Utradea financial information API suggests that interactions with tweets concerning BNB and safemoon are heavily driven by bots.

Data showing the ten most posted tickers on Twitter over the last 24 hours shows two outliers with unlikely tweet engagement metrics.

Namely, BNB and safemoon have had significantly less posts published on Twitter than bitcoin (BTC) and ethereum (ETH), but their number of likes and tweets were so high that they even exceeded the number of impressions — which is rather improbable.

The exponential chart above shows how the numbers of impressions, posts, likes and retweets per every one of the top ten most posted cryptocurrencies over the last 24 hours compares.

Ethereum has the most posts, bitcoin dominates the impressions, BNB has the most retweets and likes with ethereum not far behind in both those metrics.

Still, the data concerning BNB and safemoon strikes as rather odd to those familiar with the concept of impressions and those who have viewed social media analytics before. Impressions are the number of times users see a piece of content on a social media platform.

Impressions are counted every time a tweet is displayed, regardless of whether the user interacts with the content.

Twitter’s help center notes that impressions are not counted for community tweets, Twitter circle tweets and older tweets, so part of the discrepancy could be explained by an anomalous number of interactions with old tweets (it is unlikely that community tweets and circle tweets play any role in this data).

Twitter also clearly states that one user may count as multiple impressions by viewing a tweet multiple times and even the tweet’s author counts, making those numbers even more unexplainable. Tweet impressions have been public since the end of 2022.

Indeed, the unrealistic engagement numbers suggest that the BNB and safemoon communities have a disproportionate number of bots interacting with and sharing posts — very possibly in a way that does not count toward the impression count.

Indeed artificially driving up engagement numbers by this much would be significantly more cumbersome from a computing and bandwith perspective if the bots had to completely load the content, so those systems may end up not triggering an impression based on how they are counted.

YouTube detective and crypto personality Coffezilla, famous for numerous successful investigations, described safemoon as ‘a billion dollar fraud’ in a video viewed 3.5 million times nine months ago.

The class action fraud lawsuit is still ongoing, so it would not be a stretch to expect a fraudulent project to artificially pump its social media engagement numbers — but what about BNB?

While it cannot be ruled out that BNB’s team is involved in pumping its token’s engagement numbers this aggressively, for this token there is another likely explanation. BNB hosts a great number of particularly speculative — and sometimes outright fraudulent — tokens, including safemoon itself.

A quick look at Twitter’s page dedicated to the mentions of the $BNB ticker shows that it is often cited despite being unrelated in the tweet — presumably in an attempt to reach potential investors who already engage with the BNB blockchain.

For this reason it is possible that BNB’s engagement numbers are indeed significantly inflated by bot activity, but without any involvement from the central development team.

Another datapoint where safemoon significantly stands out is the average account age of the accounts that interact with tweets featuring its ticker.

While bitcoin has an average account age of 1,233 days, ethereum 1,183, BNB 893, safemoon has an average account age of 2,211 — well over six years. So far there is no explanation for this anomaly in the on-chain data.

The report follows crypto scam project promoted by Logan Paul — who also promoted Safemoon, which got him a lawsuit — Crypto $ZOO increasing in value following an expose by Coffezilla in January.