Twitter: whether Elon Musk wins his case or not, metrics should be reconsidered

Twitter says spam and fake accounts make up less than 5 per cent of its monetisable daily users. Elon Musk is trying to extract himself from a $44bn deal to buy the social media company by arguing that this figure is a gross underestimate. A whistleblower has since claimed that Twitter deliberately misled regulators about security. Neither claim guarantees Musk’s case. But there is reason to think Twitter’s valuation is too high.

The company says that it has 237.8mn daily active users — aka mDAUs. Revenue in the last quarter, most of which comes from advertisers paying to show their adverts to users, was $1.18bn. That means the average revenue per user is just under $5 — about half as much as Facebook parent Meta.

Musk has mused that the real fake account figure could be at least 20 per cent. If true, then advertisers are paying over $6 per real user. If they paid $5 per user, revenue would be below $1bn in the last quarter. On the current revenue multiple, Twitter’s enterprise value would drop to $24bn. This still looks generous compared to peers. On the same forward ebitda multiple as Meta, Twitter’s enterprise value would be closer to $9bn.

Yet Musk’s hopes of using such calculations to get out of the deal or force a discount are muddied by the fact that it is nearly impossible for social media companies to count their user numbers accurately, something they acknowledge. Advertisers are also aware of this fact and appear not to care.

Twitter’s central user metric, mDAU, is not based on any standardised industry methodology. Nor is there an industry-wide definition of bots. Or an agreed measure to calculate real user numbers.

For years, social media companies have had things their own way. They have encouraged growth by making it extremely easy to open accounts, something that also makes it easy to create fake accounts. Meanwhile, they have enticed investors with high user-growth metrics while also admitting that these are hard to calculate precisely. The sector’s biggest companies are more than a decade old. Time to stop promoting flawed metrics.