Background - News & Events (Landing) 2016

Media Contacts

Search NewsRoom

Advanced Search >

Tech & Brand Law Bulletin

Does Elon Musk Have a Point? The Impact of Bots on Twitter Revenue.

June 9, 2022

The news of Elon Musk’s $42-44 billion offer to purchase Twitter, and his apparent cold feet, have spread far and wide. Speculation has swirled that his offer was a politically-motivated stunt and that he never intended to actually follow-through with the deal. More recently, however, Musk has publicly demanded more information from Twitter regarding its so-called “bots” and he has publicly suggested that up to 20% of Twitter’s active user base is comprised of fake accounts, in contrast to the 5% that Twitter itself has claimed in its latest annual report. So, is Musk’s stated concern real or a coverup for cold feet?

While some have derided Musk’s demands as pretextual, he has a valid point that the number of real, daily active users on the Twitter platform – actual human eyeballs – is critical to the value of Twitter. A difference of 15% in estimates of bots is likely to be material. Twitter itself has said so.

To understand why requires a bit of background.

While Twitter is a digital company, it makes money the way many publishers have done for centuries – by selling advertising. As stated in Twitter’s 2021 Annual Report. “The substantial majority of our revenue is currently generated from third parties advertising on Twitter. We generate substantially all of our advertising revenue through the sale of our Promoted Products: Promoted Ads, Twitter Amplify, Follower Ads and Twitter Takeover. As is common in our industry, our advertisers do not have long-term advertising commitments with us.” The prices charged by Twitter for these services is directly correlated with the number of active users who might be exposed to the advertising. It’s simple, the greater variety and larger number of potential consumers the ads reach, the more advertisers are willing to pay. But bots don’t buy products. Humans do.

It is for this reason that Twitter pays close attention to what it calls “Monetizable Daily Active Users”, or “MDAU” for short. MDAU are actual human beings using the platform – well, not just humans, but “Monetizable Daily Active Users” as stated in the dry, Wharton-derived finance parlance often used in Silicon Valley. Twitter measures and reports on MDAU on a monthly basis and it has stated its aspiration to grow MDAU by 20% on a compound annual basis by 2023.

This is why bots are a problem. Bots are fake or automated accounts that certainly exist on Twitter, and apparently in some fairly large number ranging in the U.S. from about 2 million per month (5% of MDAU according to Twitter’s estimates) to about 8 million (20% of MDAU according to Musk’s estimate). The $42 billion question is how much, if any, of Twitter’s MDAU estimate is comprised of bots?

This is important for at least three reasons. First, if you are an advertiser paying Twitter to run your ads, you certainly want to know the potential real reach of your ads. And if those estimates are overstated by 15%, you would not pay Twitter as much money. Twitter makes most of its money by selling ads. Its valuation depends on how well it does that core function.

Second, proliferation of bots on the platform degrades the user experience for real users, potentially driving them away. MDAU will drop if real people quit the platform.

Third, advertisers who have historically placed ads on Twitter, might ask for their money back if they learn MDAU was substantially overstated when they purchased – and perhaps even sue.

This places considerable pressure on Twitter to do a good job estimating MDAU. We say “estimating” because directly “measuring” is not possible. Like attempting to assess how many tuna there are in the sea, one can only sample and draw inferences from each sample (think dipping a net in the water and pulling out a few fish) regarding the size of the larger population. With the astronomical number of accounts on the site and no simple, direct way to test each and every account for humanity, Twitter reports that it employs a variety of algorithmic methods to observe account activity for indications of “bot-like” behavior such as spamming and “suspicious” behavior. Twitter says it provided information about how it carries out these estimates to Musk’s team, but Musk apparently disagrees that it is adequate. Jumping into the fray, Texas Attorney General Ken Paxton has recently announced an investigation into the Twitter MDAU calculations as well. We’ll never know AG Paxton’s motives for investigating, but perhaps this has something to do with Musk’s recent relocation of Tesla to Texas.

All that said, Twitter’s apparent difficulty producing estimates of MDAU that satisfy Musk could have substantial ramifications for Twitter’s revenues and its future ownership. The difficulty in measuring actual eyeballs viewing ads is a perpetual conundrum in the advertising industry. It is often swept under the rug or minimized by self-interested actors in the adtech space. Nevertheless, given the possibly large amounts of money being regularly spent for services not provided, this issue could spiral into litigation in the near future.

UPDATE: On June 8, it was reported in national media that Twitter had agreed to provide Musk’s team with “full ‘firehose,’ the massive stream of data comprising more than 500 million tweets posted each day.”

For more information, please contact the professional(s) listed below, or your regular Crowell & Moring contact.

Christopher A. Cole
Partner – Washington, D.C.
Phone: +1.202.624.2701
Email: ccole@crowell.com