Coupa Helped Pearson Increase Spend Compliance from 22% to 85%: How AI Agent Transforms Procurement

Pearson, a giant in education and publishing, faced a typical problem: a complex procurement process leading to low spend compliance. In just six months after implementing Coupa's AI agent, Pearson managed to boost this metric from a critical 22% to an impressive 85%, with 87% of users reporting high satisfaction with the system.

Procurement isn't just about spending money; it's a strategic process that can either optimize costs or become a budget black hole. Thousands of small decisions, manual document processing, and a lack of a unified view all lead to inefficiency, missed opportunities, and high risks. It used to be considered an unavoidable evil, but today, AI agents are changing the game, transforming chaos into an ordered system.

The hidden costs of traditional procurement

In an organization as large as Pearson, procurement impacts many departments and geographical regions. Before the AI agent implementation, the process was fragmented and required significant manual effort. This led to low spend transparency, difficulties in tracking contracts, and consequently, a low level of compliance with established spending rules – just 22%.

Employees spent hours searching for suitable suppliers, cross-referencing prices, placing orders, and coordinating documents. This not only slowed down work but also created budget risks: the lack of a unified view of expenditures made it difficult to identify inefficient spending and potential overpayments. With £300 million sterling flowing through the system, even a small percentage of inefficiency translated into millions in losses.

Why previous tools fell short

Pearson already had systems for procurement management, but they could not provide the required level of automation and control. These systems were focused on executing individual steps of the process but lacked the ability for comprehensive analysis, forecasting, and autonomous decision-making. They required constant manual intervention and could not adapt to changing market conditions or internal needs without significant effort from staff.

The company needed a solution that could not just automate individual tasks, but take on the entire source-to-pay cycle, providing proactive analytics and risk management. This is why the choice fell on an AI agent capable of working with vast amounts of data and making informed decisions based on that data.

How the AI agent for Coupa was designed

Coupa's AI agent was developed as a platform for building, deploying, and orchestrating specialized agents. At its core is a unique approach that utilizes "Community Intelligence" – over $10 trillion in transaction data accumulated over 20 years. This data allows the AI agent to act not as a novice, but as an experienced professional with many years of B2B procurement experience.

The agent was designed to perform several key functions: providing savings recommendations, identifying supplier risks, benchmarking expenditures, and automating routine operations. This includes creating purchase requisitions, comparing bids, monitoring risks, sanction screening, and scenario ranking. Through integration with intelligent document processing technologies (e.g., Rossum) and orchestration (Tonkean), the AI agent can autonomously process invoices, manage task flows, and interact with various systems.

Implementation and adaptation

The implementation of the AI agent at Pearson was fast and efficient. In just 6 months after launch, the company was able to process £300 million sterling in expenditures through the new system. A key success factor was not only the agent's functionality but also its ability to seamlessly integrate into existing processes and its user-friendliness for end-users.

The AI agent took on the most labor-intensive and routine tasks, freeing up employee time for more strategic work. This not only increased efficiency but also improved the quality of decisions made, as the agent provided deep insights based on a huge volume of historical data and community benchmarks.

Implementation results

Achieving 85% spend compliance and 87% user satisfaction in just six months is an outstanding result that demonstrates the transformative potential of AI agents in procurement. The company not only gained significant savings and control over expenditures but also increased productivity and employee morale by freeing them from routine tasks.

How to implement this in your company

If your company is facing inefficiencies in procurement, an AI agent can be a powerful tool for optimization. Here's where to start:

• Identify pain points. Determine where in your procurement process extra hours are spent, where risks arise, and where there is the greatest potential for savings.
• Utilize data. Ensure you have access to high-quality historical procurement data. The more data, the "smarter" your AI agent will be.
• Start small. Don't try to automate everything at once. Choose one or two key processes where AI can quickly show results, such as bid analysis or supplier risk monitoring.
• Integrate the agent into existing systems. The more seamlessly the AI agent integrates into your current workflow, the faster employees will adopt it, and the higher the return will be.

If this case sounds like what's happening in your company, our manager can help: he'll analyze your business and niche for free and point out where an AI agent would bring a real result in your case. Message the manager