Enabling AI in Procurement Transformation – The Case For Adoption in the Electric Utility Industry

Artificial Intelligence (AI) is taking center stage in 2023, with the emergence of generative tools such as ChatGPT, Bing Chat and Google's Bard, along with image generator tools like DALL-E 2, Midjourney and DreamStudio. These new and rapidly evolving AI tools offer myriad opportunities for performance improvements across the business enterprise.  

Industry analysts are predicting growth in the market for AI-based automated procurement process decision-making technologies will increase from 15% in 2023 to more than 50% by 2025. The areas where we see significant impact are in improved efficiencies and cost reduction due to innovations, which in turn will lead to an improved customer experience and decision making due to easier analysis of large datasets. 

Procurement Opportunities in the Electric Utility Industry 

Over the past two years, we have reviewed procurement opportunities in the electric utilities industry — presented in SupplyChainBrain as “Powering Procurement in the Electric Utility Industry” (Part 1 in July 2021 and Part 2 in August 2022). 

We highlighted how utilities should embrace best practices for streamlining procurement processes and implementing next-generation technologies. Procure-to-pay processes execute the complete procurement cycle, from the time a purchase requisition is created to when the invoice is paid. A fully integrated procure-to-pay process provides end-to-end coverage and tracking functionality. Utilities should set their procurement technology strategy to implement a fully integrated global I.T. system infrastructure that incorporates procure-to-pay operations with fully automated workflows. A consolidated data warehouse using standard formats for utilizing electronic data exchange, feeding web-based supply and demand planning systems, will provide timely and meaningful business intelligence. It will also feed forecast data, inventory levels, and project schedules with the required procurement documentation.

To maximize their return on the investment, utility industry procurement managers must continue to embrace best practices for streamlining processes and implementing enabling NextGen technologies. The drive toward streamlining and digitizing procurement operations continues to be a major focus for the utilities industry.  

The different initiatives and objectives of the procurement function across utility companies often lead to siloed operations and sub-optimal performance. Looking across business processes, technology and performance measures, business leaders often experience significant disconnects and missed opportunities for improvement. 

In our experience with procurement organizations in large electric utilities, the top 4-8% of vendors comprise 75-85% of the spend, while the bottom 80% of the vendors account for 4-8% of the spend.  The highly concentrated spend, and the wide fragmentation in the “tail” suggests two opportunities:  to maximize cost reduction efforts/strategic focus on the “top” vendors, and minimize the cost to manage vendors wagging the “tail.” Often, with limited resource levels and tools, the procurement team does not have the bandwidth to drive strategic sourcing initiatives.  The teams are often overwhelmed by meeting the ongoing day-to-day business requirements of their stakeholders.  It is key to remember that only 10-15% of procurement activities are strategic, and as much as 85-90% are transactional in nature.

AI-Driven Procurement 

With this focus on streamlining procurement operations, the emergence of AI offers impactful possibilities in supply chain and procurement performance enhancement.  What we outlined in our earlier articles was that procurement organizations in many electric utilities have significantly advanced their capabilities in the past decade, introducing concepts of strategic sourcing, and improving management of spend, suppliers, and contracts.  They have also focused on the importance of category management, sourcing, and supplier relationship management.  

This focus on the procurement organization structure for the basic procurement operation offers the greatest opportunity for significant improvement.  The advent and adoption of AI tools will only grow with accelerated adoption, and the procurement organization must step up to include these as part of their overall toolkit to improve business process efficiencies to maximize human resources. These will also help with: 

• Understanding and recognizing patterns across tasks while looking at issues in a holistic manner, without getting caught in a narrow approach to problem-solving. Data sets can be analyzed efficiently without the human error that creeps in while performing routine and repetitive, time-consuming tasks. New and inexperienced users can be guided using generative AI to navigate the maze of often complex procurement processes, including which forms to fill out and which boxes to check. 
• Information (data) mining across a multitude of ever-changing suppliers and categories. It will also help in the ability to assess global supply-and-demand gyrations, adverse price signals and supplier financial stability, to proactively manage risks in a timely manner. 
• Monitoring business processes holistically by cutting across narrow organizational siloes and helping consolidate relevant data sets. AI, using predictive analytics, can facilitate “what if” analyses, and be used to generate a first-pass document of common and routine procurement contracts, while also updating base data across existing contracts.
• Reducing the amount of time procurement professionals spend analyzing spend data across different organizational units. AI can help develop machine-learning algorithms for spend analysis, to improve and speed up spend classification and vendor matching while sifting through massive data sets of suppliers and categories. AI based algorithms rapidly sift through the huge spend data and generate meaningful recommendations. This leads to better decision-making, improved supplier relationships, reduced human error, optimal resource allocation and usage, and better management of risk. 

Some of the leading enterprise software vendors have already started building AI capabilities into their offerings. Ariba sourcing has developed capabilities for extracting data from previous sourcing events, to compile relevant supplier lists for future sourcing activities.  This can expedite the RFI, RFQ, and RFP processes without significant training effort for new procurement staff. Similarly, Fieldglass AI can be used for identifying promising matches for contingent workers by extracting keywords identified from job descriptions.  The process of describing indirect materials can be simplified.  AI based algorithms can automatically suggest category names by matching against backend systems, in order to significantly improve the performance of operational buyers. It will also reduce errors which become time-consuming and problematic to fix later.

The time-consuming process of invoice reconciliation and payment can be streamlined by assigning account attributes (GL and cost center) based on previously processed invoices.  SAP AI helps lower errors, accelerate processing time, reduce costs, and improve compliance. Coupa AI Classification takes a different approach to the daunting tasks of standardizing, normalizing and enriching spend data. Using a vast trove of more than one trillion dollars (US) of spend data across the globe, Coupa AI Classification helps clients control costs and more effectively manage the supply chain by providing deep insight into spend: what purchases are made, from whom, from where, as well as when and why. This enables companies to trust the data and make critical business decisions based on real financial facts and spend behavior.

Carpe Diem! 

Enabling and leveraging AI to support an electric utility company’s operations, and ensuring significant business impact, is not something that will happen overnight.  Companies need a short-term plan to deliver quick, high-impact AI wins as well as a long-term strategy, fostering a transformative AI culture across the organization. Using AI to improve procurement performance in the electric utility industry is an existential strategy, given the challenges exposed through the recent pandemic, climate change/weather storms, and an increasingly shrinking and tech savvy talent/labor pool. 

As an aging workforce, with significant business-related contextual information accumulated over years, continues to retire, knowledge continuity within key positions is a significant challenge for utilities. Similarly, utilities must offer NextGen information technology infrastructure services to attract and retain the newer generation workforce. Utilities must initiate procurement and technology improvement strategies to fully capitalize on the available funding and public awareness to effectively deliver power to the communities they serve. Seize the day! 

Sandeep Shah is global managing partner, procurement strategy in the global supply chain consulting & transformation practice at Tata Consultancy Services (TCS).

TK Subramanian is a director, procurement strategy in the global supply chain consulting & transformation practice at Tata Consultancy Services (TCS).