Embracing AI and Predictive Procurement to Overcome Industry Challenges

Today’s procurement teams are faced with a host of challenges: Managing more spend with fewer headcount. Dramatically improving the internal customer experience and supplier experience. Enhancing supplier performance. Balancing conflicting priorities and delivering year-over-year cost reductions. 

It’s a tall order, especially for large enterprise manufacturers with multiple unique plant locations. Ultimately, margins are tighter, prices are higher, and details matter more than ever before.

Fortunately, procurement teams can overcome these challenges — but they need higher-quality procurement data and optimized procurement processes to make it happen. In this blog, we’ll share how to improve data quality, leverage attribution frameworks, and optimize cadence and tempo to increase cost savings.

These findings are from our Founder and CSO, Edmund Zagorin’s series on Forbes Technology Council called, “The Devil’s in the Details.”

Turning Low-Quality Procurement Data Into High-Impact Cost Reductions

Many organizations are plagued by data silos and disparate systems that lack the visibility, communication, and collaboration needed to win cost reductions. Though low-quality, item-level procurement data is holding them back, it also represents a major opportunity for cost savings once they get it right.

Here are some of the most common data quality issues procurement and purchasing teams face.

Unit of Measure

Unit of measure (UOM) indicates the unit used to measure any given item, such as pound, pallet, box, truckload, etc. Unfortunately, different stakeholders may use the same UOM to express different values, which can lead to varying cost expectations.  

Item Description

Inconsistent item descriptions can cause procurement teams to miss out on big cost-saving opportunities. For example, a manufacturer might list the same item description for the item itself, and for the services associated with delivery. This can lead to problems with cost benchmarking, demand forecasting, and more.

Geographic Segmentation

Geographic segmentation is about analyzing how purchase prices vary across plant locations and geographic areas. With accurate data, this analysis helps procurement identify regional cost differences and changes. However, poor data can lead to unrealistic expectations around price and cost transformation with suppliers.

Predictive and generative AI can help procurement teams resolve these data quality issues, allowing for long-term sustainable margin improvement. These technologies can clean large sets of item master data in seconds, offering a more accurate picture of how items should be categorized, measured, segmented, and more. Such tools can also be used to create a framework for producing higher-quality data for lasting benefits.

The Role of Cost Savings Attribution in Procurement

For many procurement organizations, the hardest part of improving data quality is getting started. Item-level data quality may be so poor — especially at the plant level — that the task seems insurmountable.

Here’s where attribution mapping comes in. Tracing the source of data quality issues can create immense cost savings by pointing to the features of a given dataset that need more attention.

Let’s take a look at how attribution mapping works.

The Spend Cube

Cost is often analyzed using the spend cube — a method that analyzes historical transactions across spend. This method offers an overview of each spend category and highlights opportunities for improvement, but it also lacks important factors that impact cost, including geography, time, and seasonality. Like a bill that hasn’t been itemized, it creates challenges for procurement teams seeking to benchmark item-level costs against industry standards.

The Imperative of Attribution Frameworks

To overcome these challenges, procurement must go beyond the traditional spend cube method and analyze an itemized list of spend. At this point, teams can leverage machine learning and generative AI to execute the attribution process in minutes. 

Once specific attribution codes have been assigned to data quality issues for every transaction, procurement teams can better spot underlying problems, and then prioritize and resolve those issues with greater efficiency. 

Optimizing Order Cadence and Tempo to Achieve Procurement Savings

Cadence and tempo have the potential to greatly impact cost, efficiency, and supplier relationships in procurement management. In an ideal world, cadence and tempo strike a careful balance between being agile and flexible vs. deliberate and cautious. But in reality, many procurement teams struggle to achieve the right balance, often facing delays due to complex processes.

Thankfully, there are a few solutions to help procurement leaders get their order cadence and tempo under control, be more prepared for disruption, and achieve procurement savings.

High Costs for Delays

Spend data doesn’t usually account for costs associated with unplanned, urgent procurement tasks. This is why it’s critical to conduct a more realistic time-series analysis of transaction data. These analyses tie the source of truth back to the quoted cost components, allowing procurement professionals to quickly spot correlations between high costs and longer cycle times. Teams can identify temporal order clusters and then compare them to price variances, which offers a huge opportunity for dramatic cost reductions.

Seasonality: Procurement Events and Market Tests

Seasonal items often present a cost-saving opportunity for procurement. Rather than simply triaging reviews based on quarterly financial reporting periods, procurement can leverage the impact of seasonal demand to accurately calendar the quoting process with suppliers. This allows procurement teams to be proactive in setting desirable terms for their business partners. 

Identify Predictably-Wrong Demand Plans

Many demand plans have underestimated demand numbers that then sway the cadence of orders. In many organizations, this is because sales teams tend to prefer a forecast that estimates the number of sold units and then outperforms that estimate. But when demand and prices are both high in the market, that predictably-wrong demand plan also creates unnecessary shipping costs and inefficiencies in procurement cadence and tempo.

In these cases, it’s helpful to perform a looking-back analysis on historical demand forecasts. Let’s say forecasts have always been wrong by a margin of 3-5%, and the price of logistics has gone up. In this scenario, procurement should plan for additional capacity, instead of having to scramble for spot quotes at the last minute. 

Market Basket

Order fragmentation and consolidation is another quick win that can achieve significant cost reductions. By placing orders for the same items in bulk to the same manufacturer, procurement can take advantage of volume-based discounts or rebates. This process also simplifies transaction reporting, since companies can communicate material demands to suppliers who can easily track and respond.

In the past, these savings were impossible because smaller orders didn’t have time to wait in a queue for procurement to take action. However, in a world where decentralized purchasing has become just as digital as its centralized counterpart, it’s now possible to combine orders and unlock business benefits in the process. 

Your AI Playbook: Join Us at ISM World 2024

When procurement cost savings are a top priority, teams should do everything they can to improve data quality, leverage attribution mapping, and optimize order cadence and tempo. AI and autonomous sourcing technology make it easier to successfully execute these strategies, allowing teams to identify and act on cost-saving opportunities at scale.

For many organizations, implementing AI and other emerging technologies is easier said than done. If you’re seeking guidance on the path to AI transformation, don’t miss our ISM World 2024 workshop, “Procurement’s AI Governance Playbook: Managing AI Across Your Supply Base.” Hosted by Arkestro founder Edmund Zagorin and several guest speakers, this workshop gives participants a look at AI and predictive procurement in action, along with tips for AI governance, risks, and opportunities.

Register for ISM World 2024 today, and we’ll see you April 29-May 1 at the MGM Grand Las Vegas.