AI search startup Perplexity may have employed some accounting tactics that transform what would have been negative profitability into an impressive 60% gross margin, financial documents obtained by The Information reveal. The $9 billion startup shifted millions in costs away from its income statement as it courts investors for even higher valuations.
Financial statements reviewed by the publication show Perplexity spent at least $57 million last year on AI models and web services, with $33 million supporting non-paying users. Rather than counting these as direct costs, the company classified them as research and development expenses.
Had Perplexity counted costs for free-tier and trial users as “cost of revenue,” the company would have shown a negative gross profit margin instead of the 60% it reports.
The San Francisco-based company has seen explosive growth since its 2022 founding. Its valuation jumped from $500 million in early 2024 to $9 billion by December, with recent talks potentially pushing it to $18 billion.
Perplexity reported $63 million in annual recurring revenue at 2024’s end — an 800% year-over-year increase. Sources familiar with the company say it’s now approaching $100 million.
Perplexity’s reported margins compare favorably to competitor OpenAI’s 40% gross profit margin, though they fall short of the 75% target the company previously shared with investors. The company claims infrastructure cost advantages from using open-source models instead of relying entirely on commercial APIs like those from OpenAI.
Perplexity projects $656 million in annual recurring revenue by 2026, a tenfold increase. This aggressive growth forecast helps justify its premium valuation of approximately 143 times revenue.
Many growth-stage startups use similar accounting interpretations, but it still raises red flags about about Perplexity’s business sustainability. Analysts say increased scrutiny may threaten the company’s ability to maintain its valuation as it faces mounting competition from Google, Microsoft and other AI companies.
Information for this story was found via the sources and companies mentioned. The author has no securities or affiliations related to the organizations discussed. Not a recommendation to buy or sell. Always do additional research and consult a professional before purchasing a security. The author holds no licenses.
