TL;DR
AI-exposed listed companies traded at about 22 times forward revenue in Q1 2026, while a February 2026 NBER survey found 90% of firms reported no measurable AI productivity impact. The gap matters because investors and managers have already priced in gains that have yet to appear in many operating results.
AI-exposed listed companies traded at a median of about 22 times forward revenue in Q1 2026, while a February 2026 NBER survey cited in the original Thorsten Meyer AI analysis found 90% of firms reported no measurable productivity impact from AI, highlighting a gap between market expectations and current business results.
The source material says the S&P 500 traded near 7 times forward revenue during the same period, far below the median multiple for AI-exposed listed companies. That spread implies investors expect AI-linked companies to deliver faster growth, higher margins, or both.
At the company level, AI adoption remains easier to show than productivity. Thorsten Meyer AI cited earnings-call activity showing 76% of firms mentioned AI, while the NBER survey found executives projected only a 1.4% median future productivity gain.
The confirmed data does not show that AI is failing. It shows that many firms cannot yet connect AI spending to measurable gains in revenue per employee, cycle time, margins, service quality, or customer outcomes.
Valuations Depend on Delivered Output
The productivity gap matters because capital, budgets, and staffing plans have moved ahead of clear operating proof. If AI tools raise output only in narrow tasks but not across full workflows, companies may struggle to justify high software, compute, integration, and training costs.
For investors, the risk is multiple compression if earnings do not catch up with expectations. For workers and managers, the risk is a wave of automation plans built around assumed savings that have not yet reached the income statement.
AI productivity monitoring software
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
From AI Mentions to Margins
Thorsten Meyer AI defines the AI bubble productivity gap as the distance between AI promises and measurable productivity gains. The source argues that valuation debate alone is a weaker signal than productivity because valuations measure expectations, while productivity measures output.
The source says gains are more visible in narrow workflows, including code generation, tier-one support, document extraction, marketing drafts, and contract review. It also says those task-level improvements do not automatically become business-unit gains if bottlenecks move to pricing, legal review, compliance, approvals, or sales conversion.
“1.4% median future gain”
— Executives in the NBER survey, as cited by Thorsten Meyer AI
As an affiliate, we earn on qualifying purchases.
Margins Still Lack Clear Proof
It is not yet clear how much of the gap reflects weak AI returns, poor measurement, slow integration, or a delay between tool adoption and financial results. The NBER survey captures reported measurable impact at a point in time; it does not rule out later gains.
It is also unclear whether high AI multiples reflect realistic future earnings, scarcity value, investor enthusiasm, or some mix of all three. More company-level disclosure is needed to separate durable gains from activity metrics.
As an affiliate, we earn on qualifying purchases.
2027 Plans Face a Test
The next test is whether companies can show gains for at least two quarters in revenue per employee, margins, error rates, approval speed, and customer outcomes. Thorsten Meyer AI says leaders should stress-test 2027 plans at a 0.7% productivity gain and audit AI results by business unit before scaling budgets.
Signals to watch include stalled revenue per employee, cuts to AI-related capital spending, and falling valuation multiples. Any one signal may have several causes, but a combination would suggest the productivity gap is becoming a financial problem.
As an affiliate, we earn on qualifying purchases.
Key Questions
What is the AI productivity gap?
It is the difference between expected AI gains and productivity that companies can measure in operations or financial results.
Does this mean AI is not useful?
No. The data shows many firms have not yet measured broad productivity gains. It does not prove that AI lacks value.
Where are AI gains showing up?
The source material points to narrower workflows such as code generation, support, document extraction, marketing drafts, and contract review.
Why do valuations matter here?
High revenue multiples depend on future growth or margin gains. If productivity does not arrive, those valuations may be harder to defend.
What should readers watch next?
Watch whether companies connect AI use to revenue per employee, margins, cycle time, service quality, and customer outcomes over several quarters.
Source: Thorsten Meyer AI
