Menlo Ventures: enterprise AI deployment spend hit $177B in 2026, and it is growing fastest where agents get connected systems, context and checkable results

Menlo Ventures published its 2026 State of Enterprise AI report on November 3. It estimates that enterprises spent $176.8B putting AI to work by September 2026, up from $33B a year earlier, which works out to about $1.70 spent on using models for every $1 spent on building them. The report is based on a Morning Consult survey of nearly 400 U.S. enterprise decision-makers plus Menlo's own market model. Menlo is an investor in Anthropic and several other companies it names.

The part that matters most for FDEs is Menlo's explanation of where agent spending sticks. It says agents succeed when they have systems they can act in, the full context of how work gets done, and results they can verify. Most enterprise functions lack those conditions, so the opportunity is to build them: connect agents to fragmented systems, write down tacit knowledge and procedures, and design evals. That is a fair description of day-to-day forward deployed work.

The report names this model directly. In customer support, a $2.7B market that grew 4.3x, it says Sierra and Decagon embed forward-deployed engineers or agent PMs to design procedures, connect systems and codify workflows. By contrast, functions where results are harder to verify, such as HR and marketing, grew more slowly.

Other findings: Anthropic's share of enterprise LLM API spend rose to 55% and OpenAI's to 31%, while coding ($34.0B) and chatbots ($25.0B) make up 76% of the $77.3B application market. For engineers weighing FDE roles, the report suggests that demand will follow companies that can turn messy, unverifiable workflows into environments agents can actually work in.