Enterprises in Menlo Ventures’ survey spent an average of 4.7 times their 2025 AI spending this year, against a planned 1.9 times, or roughly 2.5 times plan, Menlo said in 2026: The State of Enterprise AI, published on 3 November. By our arithmetic, the average multiple is about 2.1 times the median multiple of 2.2x.
Menlo says the report draws on a survey of nearly 400 U.S. enterprise decision-makers responsible for more than $4 billion in AI spending, fielded by Morning Consult from September 8 to 26, 2026. Menlo Ventures is an investor in Anthropic, and the report marks its other portfolio companies with an asterisk.
Business Pill · AI Agents
A short explainer on AI agents: software that moves from answering to doing. It teaches the general idea only and does not discuss Menlo’s report.
The key insight: By our arithmetic, the average 2026 spending multiple in Menlo Ventures’ survey (4.7x) is about 2.1 times the median (2.2x), while the planned multiples were 1.9x and 1.8x.
The section opens: “The average enterprise planned to nearly double their AI spend in 2026. They ended up paying far more”.
The median moved less. Menlo writes: “Median spending also exceeded expectations, though not nearly as dramatically: Spend reached 2.2x last year’s level against a 1.8x expectation”. Menlo’s chart labels the average overshoot +147% and the median overshoot +22%.
Menlo’s explanation is the shift to agents: “As for the difference between actual and planned spend, agents explain the gap. They work independently, run in parallel, and keep going even after people step away, so spending scales with work rather than headcount”.
It adds that “Spend per enterprise nearly quintupled while adoption among enterprises rose only from 55% to 65%”. Across the market, Menlo puts enterprise AI deployment spending at $176.8 billion by September 2026, up 5.4x.
The bills rose even as unit prices fell. Menlo says the blended price of closed-model tokens fell 24% to $0.70 per million, and that “As agents account for more of the usage, the average price per token falls”.
A Fat Head of Power Users
Menlo reads the split between average and median as concentration: “The difference between average and median reflects AI’s “fat head”: a small group of power users accounts for a disproportionate share of AI spend and grew far faster than most enterprises”.
By our arithmetic, the average multiple (4.7x) is about 2.1 times the median multiple (2.2x), while the planned multiples were close together at 1.9x and 1.8x. Menlo’s own reading of that pattern: “With so much AI spend coming from power users, both adoption and usage still have significant room to grow across the long tail of the rest of the economy”.
Against Menlo’s Earlier Reports
Menlo’s December 2025 report closed by saying: “Enterprises, seeing real returns, are doubling down”. Its 2024 report found that “60% of enterprise generative AI investments come from innovation budgets”. The 2026 report measures something different, plan against outcome, so the figures are not a continuous series.
The Structural Read
As we read Menlo’s numbers, the plans were close together and the outcomes were not: the average and median budgets were 1.9x and 1.8x, while actual spending came in at 4.7x and 2.2x.
As we read the report, spending moved with usage rather than with seats: Menlo says spend per enterprise nearly quintupled while adoption rose from 55% to 65%, and it ties the gap to agents that keep working after people step away.
Menlo Ventures, 2026: The State of Enterprise AI
“As for the difference between actual and planned spend, agents explain the gap.”
Three Implications
FOR FINANCE TEAMS Menlo found average 2026 AI spending at 4.7x the 2025 level against a planned 1.9x, roughly 2.5x plan.
FOR AI VENDORS Menlo says a small group of power users accounts for a disproportionate share of AI spend and grew far faster than most enterprises.
FOR THE WIDER MARKET Menlo puts enterprise AI adoption at 65%, up from 55%, and deployment spending at $176.8 billion.
The framework argues: “As companies move from chatbots to thousands of autonomous agentic workflows running 24/7, the line item that used to be a rounding error becomes the budget.”
As we read it, Menlo’s 4.7x average against a 1.9x plan is that line item outrunning the budget. The framework is the owner’s argument and does not discuss Menlo’s report.
The multiples are self-reported by survey respondents. The report text gives the average and the median but not the spread of responses behind them, so the size of the power-user group is not stated.
Menlo’s link between agents and the overshoot is its own explanation, not a measured split of the gap. Menlo Ventures is an investor in Anthropic, and the report marks its other portfolio companies with an asterisk. We did not contact Menlo.
This Week In The AI Supercycle
Anthropic’s Reality Check, The Capital Squeeze, & The New World Order
This week’s issue: for two years the AI build-out was priced as a growthstory. This week it started being priced as a credit story.
Menlo Ventures found that enterprises spent an average of 4.7 times their 2025 AI spending against a plan of 1.9 times, while the median reached 2.2 times against 1.8 times. By our arithmetic, the average multiple is about 2.1 times the median, the gap Menlo attributes to a small group of power users.
A note on sourcing. The spending multiples, survey details and quotes come from Menlo Ventures’ 2026 State of Enterprise AI report, published on 3 November 2026, and its 2025 and 2024 reports. The chart is Menlo’s own. The 2.1x ratio of average to median is our arithmetic on Menlo’s figures. Menlo Ventures is an investor in Anthropic and in other companies named in the report. We did not contact Menlo.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
