The right AI applied to the right operation to achieve the right return
Companies continue to face constant pressure from unanticipated market disruption. From inflationary input costs to geopolitically induced bottlenecks, supply chains are changing the way they operate to adapt to variations in supply and demand. Fortunately, artificial intelligence (AI) has also undergone its own evolution, growing in both breadth and depth of capability, with hopes of offering new levels of clarity and insight amid operations uncertainty.
Based on our latest research, 47% of leaders view AI as having the greatest value creation potential in supply chain optimization and cost reduction1. Over the years, AI has proven to be an effective enabler of operations excellence, arming organizations with sophisticated capabilities to improve service, cost, and inventory. With the advent of generative AI, its impact has extended to productivity, augmenting the workforce with an ability to generate greater output with less effort. The role of AI and its impact on operations is becoming more diverse and pervasive, requiring a more measured and practical approach for the supply chain.
Numerous opportunities abound for AI to improve supply chain performance, but a library of use cases does not necessarily translate into a viable return on investment. To unlock the right value, opportunities must strike the right balance between value and feasibility:
Operational value is predominantly defined by a measurable impact to supply chain performance, but also by the ability to change an organization’s way of working and scale across a network of products and channels.
AI feasibility is heavily reliant on the availability, accessibility, and accuracy of supply chain data; but keep in mind that in turn, the data requirement itself is dictated by the type of AI being used to address the opportunity.
Algorithms are at the heart of AI, with three types that are frequently used in supply chain operations: discriminative, prescriptive, and generative. Each serves a distinct purpose and contributes a unique impact:
Discriminative AI, or what is often referred to as “traditional AI,” excels in identifying patterns to accurately predict outcomes from structured data like product SKUs or customer orders. This is often useful for anticipating shipment arrival times or sharpening forecasts during periods of volatile demand.
Prescriptive models go a step further in the decision-making process by weighing criteria to recommend actions from structured data. By simulating an operation with real-world constraints, it can suggest an action such as moving a scheduled shipment to meet a delivery date at the lowest cost within warehouse constraints.
Generative algorithms stand out with their ability to create new content and insight from unstructured data like language and imagery. Currently the most nascent of the three, this algorithm is often found generating contract language to improve supplier compliance or product images to automate quality inspection.
An AI portfolio helps address the diverse needs of a supply chain and ultimately yields better outcomes. The key is in knowing when to apply the right AI to the right operation to achieve the right value.
1
Organizational change
In this new era of generative AI, the level of impact on the workforce has never been greater. At a strategic level, there is a strong imperative where supply chain executives are pursuing a deeper level of understanding to effectively scrutinize AI investment cases. While at a tactical level, there is a push to reap returns from these investments2 with expectations for a productivity increase from employees like procurement professionals, production operators, and warehouse associates. The change management journey from organizational awareness to acceptance cannot be understated.
2 KPMG GenAI Study: The path to sustainable returns, KPMG LLP, Mar 2024
2
Build-or-buy strategy
Manufacturers strive to find the right balance between when to make versus buy their products and assets. But when it comes to emerging technology investments like AI, the make-or-buy approach has begun to shift towards a third option of strategic partnerships. Building AI that is fit-for-purpose versus acquiring AI that is embedded in a vendor platform remains a hot topic. But the debate is less about which option to invest in, and more about how to leverage either in a partner ecosystem to outmaneuver supply chain disruption.3
3 Transforming the enterprise of the future: The new champions in a digital era, KPMG International, Mar 2024
3
Operations scalability
Supply chain organizations are known for taking what works well at one location and implementing at another, thus scaling best practices while delivering economies of scale. For AI, the concept of scalability takes on a much deeper meaning. Deploying one AI to many sites is easier said than done. Each facility may carry a different operating model, resulting in additional algorithm training and tuning. In addition to scaling the network, there is an opportunity to scale the process, where multiple AI algorithms work together across multiple supply chain functions to deliver more end-to-end outcomes.
AI has emerged not only as an enabler of operational excellence but also as a catalyst for innovation and productivity across the supply chain. Moving forward, it is imperative for supply chain leaders to:
1
2
3
With a focus on these actions, organizations can be better equipped to navigate challenges with a more practical approach to incorporating AI capabilities that can withstand future disruptions.
According to senior buyers of consulting services who participated in the Source study, Perceptions of Consulting in the US in 2024, KPMG ranked No. 1 for quality in AI advice and implementation services.
Take a deeper dive into our supply chain insights
What sets apart good from great supply chain leaders? It's their ability to identify not only broad but also deeper opportunities for enhanced visibility and better decision-making.
Popular category topics
KPMG offers a technology-enabled planning transformation journey supported by a time-tested six-layer operating model that ensures accurate segmentation analysis and includes a demand plan and a data assessment. We offer an AI portfolio, a set of algorithms for supply chains, from augmenting your workforce and optimizing costs to making inventory management more efficient and assisting regulatory compliance. Let KPMG guide, accelerate, and de-risk your supply chain with purpose-built assets and accelerators designed exclusively for supply chain operations.
Sign up to receive newsletters on KPMG's latest research, analysis, news and events to help operations executives innovate.
From supply chain planning strategists to AI experts, our people have the expertise and technology to guide your supply chain through challenges and equip you to make the most of opportunities.