How Coupa and MIT Use AI to Forecast Economic Trends
One of the world's leading platforms for autonomous spend management, Coupa, has partnered with MIT Data Science Lab to release the Business Spend Index (BSI) Report 2026 Edition.
The index uses AI and a proprietary dataset of US$10tn in business transactions to forecast economic trends.
The company unveiled the report at its Inspire World Tour in London, UK. The BSI applies machine learning to millions of procurement transactions to predict business spending patterns up to 90 days in advance.
AI models predict economic shifts
According to Coupa and MIT Data Science Lab, traditional economic indicators rely on retrospective data that undergoes frequent revision. The BSI captures real spending behaviour through transactional data rather than sentiment surveys.
The index tracked capital deployment across Coupa's network of more than 10 million buyers and suppliers. The AI model detected a manufacturing turning point three months before the Institute for Supply Management Purchasing Managers' Index (ISM PMI) registered the shift.
The system analyses committed spend transactions from Coupa's proprietary dataset. It operates on the premise that purchasing actions precede executive sentiment.
The methodology eliminates subjective assessments to focus on actual B2B purchasing patterns. This could mean the index delivers more precise economic signals than traditional forecasting methods.
Machine learning applications in forecasting
Kevin Laquinto, Chief Marketing Officer at Coupa, says: "With visibility into trillions of dollars of actual business spend, you don't guess where the economy is going, you see it unfolding in real-time.
"By leveraging Coupa's unmatched proprietary dataset and combining it with the best macroeconomic indicators and AI, MIT Data Science Lab and Coupa have collaborated to deliver a report that has tremendous business value for procurement, supply chain and other business leaders who are debating and actioning future spend decisions."
David Simchi-Levi, the William Barton Rogers Professor and retired head of MIT Data Science Lab, adds: "Coupa has a very unique dataset that offers the basis for incredible economic insights.
"By applying rigorous econometric modelling and statistical analysis, we have mathematically observed that Coupa's data has strong leading signals relative to numerous established macroeconomic indicators, such as FRED, IDEA, ISM PMI and Real GDP."
The collaboration between Coupa and MIT Data Science Lab combines proprietary transactional data with established macroeconomic indicators.
The AI system processes millions of data points to generate forward-looking forecasts.
According to the report, the BSI identified month-over-month spend declines across five sectors in April 2026.
Manufacturing registered a 0.8% decline while Business Services contracted by 3.3%. In comparison, ISM PMI’s Manufacturing & Services surveys predicted the industry would expand in April.
Dataset reveals enterprise AI demand
The index forecasts procurement spend contractions in high tech, financial services and healthcare and life sciences over summer. According to the BSI, financial services could see the largest contraction.
High tech procurement spend has climbed more than 40% since mid-2024 to reach a four-year cycle high. The data tracks enterprise software and services procurement.
According to the report, the BSI provided an early signal that enterprise demand is catching up to AI infrastructure investments. This could show that corporate spending on AI tools is translating into measurable procurement activity.
According to the BSI, smaller US manufacturing firms reduced spend by 17.5% during 2025 tariff volatility. Larger manufacturing firms cut spending by 14.6% during the same period.
The February 20 2026 Supreme Court tariff ruling introduced policy stabilisation. This allowed businesses to return to long-term investments in automation and AI-enabled execution systems.
The BSI showed that late 2025 manufacturing spend increased while new customer demand remained soft. The July 2025 One Big Beautiful Bill Act included full expensing provisions.
According to the data, manufacturers used tax breaks to invest in automation, capability and supply chain resilience.
BSI is therefore rather useful to inform decision-making with an inference of how tariffs, geopolitics, trade and economic factors will impact supply, pricing and demand.
