An intelligent new trade promotions system for a CPG giant

As a CPG giant, our client relies on retailers to execute much of its marketing strategy. In fact, it contracts with more than 150 large retailers to carry out over 300,000 transactions totaling about $2 billion annually. Retailers claim back a portion of the promotion costs they incur, and in the past the CPG company created manual contracts based on sales force input.

It validated each claim and invoice prior to approval, but it wasn't practical to audit more than 30% of invoice value—and that led to significant over-payments. What's more, there was no centralized information repository, so it was difficult to assess performance or to plan promotions based on analytics.

Although the company had already optimized its existing trade pay process, the system was ripe for AI automation. But no off-the-shelf solutions could do the job.

Genpact's Contract Assistant uses built-in artificial intelligence to extract, analyze, and reconcile contracts with a high degree of accuracy. It applies computational linguistics, pattern recognition, and machine learning to extract contract terms and validate them against invoices and other data.

The Genpact team used a structured assessment process to adapt this solution to our client's specific trade promotion requirements. Key steps included:

• A business readiness assessment for the deployment of digital tools
• A recommended target operating model, governance model, and implementation plan
• A 10-week proof-of-concept exercise that automated the trade promotion process for two retailers

We used the Genpact Cora platform—and a holistic end-to-end approach—to automate each part of the process. Genpact Cora features a modular, interconnected mesh of technologies that tackles each business challenge and automates the following steps.

• Contract creation: The system receives promotion information from an online portal, interprets the information, creates a contract, reviews it and finalizes it.
• Trade pay: The system receives each customer invoice, extracts data from it and its associated contract, and interprets this information to identify validation criteria.
• Proof of performance: The system retrieves third-party market research data (from a separate application), performs price and quantity validation, and uploads the contract, invoice, and compliance template.
• Payment: On successful three-point validation, the system triggers payment.

Our proprietary data extraction, pattern recognition, and language correction capabilities helped overcome input issues such as poor-quality scans and damaged documents. That minimized exceptions.

In the end, we successfully demonstrated how the system would not only improve efficiency and accuracy, but also how the new centralized data repository would support predictive analysis. Even better, machine learning means the system will improve continuously.

Following the successful proof-of-concept phase, our client is going forward with a view to revamping the entire trade promotions process. Our modular approach means the deployment will go live in short phases for rapid ROI. Expected business benefits include:

• Incremental productivity gains across the end-to-end process
• Increases in net revenues resulting from a more effective three-point payment validation process

As well, our client will acquire a leading-edge data repository for assessing the performance of individual promotions. And that means it will allocate marketing resources more efficiently. It's a win-win all around.