How AI startups are inflating their ARR metrics
Last month, Scott Stevenson, co-founder and CEO of legal AI startup Spellbook, exposed a "massive fraud" growing among AI startups on the X platform. He claims that many companies are artificially inflating their Annual Recurring Revenue (ARR) figures. "The world's largest funds are supporting these false metrics by misleading journalists for PR purposes," Stevenson wrote. This is reported by Techcrunch.com reports .
ARR is traditionally used to calculate annual revenue from active contract-based customers. However, Stevenson and many other industry experts believe that AI companies have distorted this metric beyond recognition. TechCrunch interviewed over a dozen founders, investors, and financial experts to investigate how widespread this issue is.
Sources confirm that inflating ARR in public statements has become common practice. The main trick is to present the value of not-yet-realized, merely agreed-upon contracts (Contracted ARR or CARR) simply as ARR. According to one investor, if one startup in a category does this, others are forced to use the same method to remain competitive.
ARR has been a reliable metric for evaluating product sales since the cloud computing era. However, this figure is not officially audited by accountants because GAAP focuses on historical revenue already collected, not future revenue. This allows AI startups to portray their financial position as better than it actually is.
