J-Ventures Co-founder Oded Hermoni: ARR Multiples Are Losing Their Proxy Meaning in the AI Era

J-Ventures co-founder and managing partner Oded Hermoni, in a recent interview with ZDNet Korea, put forward a specific valuation-methodology claim: AI is eroding the proxy meaning of ARR multiples in startup valuation.

His reasoning runs as follows. First, AI will reduce the capital companies need to generate actual revenue, so early-round sizes will shrink and many companies will skip rounds entirely. Second, when a 20-person team can reach $20 million in ARR, employee count is losing its meaning as a proxy for company size. Third, the key risk is revenue quality—what he calls 'vibe revenue,' income that AI helps produce rapidly but that can disappear just as quickly if a better AI tool emerges.

Based on that judgment, Hermoni proposed a set of metrics he looks at much more heavily: revenue per employee, gross margin after accounting for AI compute costs, net revenue retention (NRR), and how defensible a company's data and distribution channels are. He considers these more informative than ARR multiples for assessing an AI-era startup.

More broadly, Hermoni believes VC value will depend less on the size of the check and more on access—connecting portfolio companies with customers, channels, talent, and strategic partners. When the hard part is no longer supplying capital but choosing which companies to back, whoever can help a startup land its first customers and follow-on investors becomes more valuable.