Understanding Mega-Rounds and Deal Sizes

Mega deals require mega funds. As the price of top AI deals skyrockets, is AI investing turning into a "rich get richer" game? Andreessen Horowitz is the current King of the Hill – with the most AI investments out of any investor and early investments into OpenAI, Databricks, xAI, Anduril, Safe Superintelligence, Shield AI, and Mistral AI. The firm's rumored fresh $20B fund positions the firm to continue to dominate the space. AI companies are attracting unprecedented investment levels, with deal sizes that dwarf those in other sectors. Q1'25 saw global venture funding rise to $121B, with AI mega-rounds ($100M+) accounting for 70% of all funding. Why? 1) High Costs Building cutting-edge AI requires enormous capital investment across infrastructure, talent, and energy costs. 2) High Multiples AI companies command significantly higher valuation multiples than other sectors: median revenue multiple for AI companies is approximately 29.7x with LLM vendors specifically commanding even higher multiples at 54.8x revenue. The AI race has created intense competition among investors – all seeking to establish strategic positions in the evolving AI ecosystem. Already, a16z has invested multiple rounds into 11 AI companies that have each raised over $1B+ in total funding. Which growth-stage companies are the next mega deal targets?

Is venture capital too concentrated in mega deals? Lots of the talk around VC-backed startups focuses on pre-seed, seed, Series A — these early stages where the ideas are unproven and the dreams are big. But if you look at where capital is actually being deployed, a rising share of total VC is going to companies in mega rounds where at least $100M is raised. So far in 2024, 45 cents of every dollar pumped into VC-backed companies was invested in a $100M+ round (only 62 deals qualified). Couple thoughts: • This is the main reply to any question about dry powder. Founders should not read the headlines of $100B, $200B, even $250B waiting on the sidelines and expect much of that to go to small, early-stage deals. It's not happening.    • Some would probably question whether rounds of this size even qualify as "venture" rounds. Has the asset class change so much over the past decade that we need to concretize sub-segments as their own thing (early-stage, growth equity, pre-IPO)?    • Wild to watch the concentration of capital actually shoot back upwards this year - looks like it's a secular trend in both high and low fundraising periods (and yes, a lot of the change this year has been in major AI rounds). It may not feel like it to many early founders, but there's actually a lot of capital sloshing around in private markets these days. Debatable whether it's "too much" but there's certainly enough to make the competitive dynamics of hot rounds detrimental to many funds economics. And this trend is set to continue with so much of recent VC fundraising going into large funds. Check size dictates a whole lot of other factors. Tale of two cities I guess. #startups #founders #venturecapital

AI deals - stretching into the billions ($) - have pushed the AI sector’s average deal size significantly. In 2023, this came in at $23.4M, up 21% YoY. However, the median deal size held steady YoY at $4.4M. Early-stage rounds soared past the $100M+ mark. In Q4’23, 3 startups raised mega-rounds: • Mistral AI ($415M Series A) • Baichuan AI ($300M Series A) • Together ($103M Series A) Interestingly, all 3 of these companies develop open-source models for generative AI. The soaring sizes of later-stage and even early-stage AI deals demonstrates intense investor demand. As more capital floods into AI startups, founders should understand and prepare for rising prices.