Supply Chains and AI: Reduce Value-Eroding Behaviors Without Stifling Relationships

Let’s take a moment and recognize that deceptive practices aren’t necessarily intended to be malicious or purposeful.

Getty Images 935226186 Artificial Intelligence
Getty Images

Many businesses struggle with value-eroding behavior in their supply chains. Although some impose too many controls with the intent of standardizing processes and improving buying power, others implement too few in the hopes of agility and fast-moving capabilities. Both approaches risk inviting behaviors that reduce efficiency and erode value — think fraud and gamification.

Let’s start with organizations with complex, control-heavy systems, ones that centralize procurement and intensify buying power. These types of firms often inadvertently encourage gamification of their bidding processes through change-order abuse.

The situation usually follows this pattern: A vendor submits a lower-than-usual bid in order to secure a specific contract. Although this initial pricing is attractive, the vendor gradually makes change-order amendments — which generally face less scrutiny than bids and contract-awarding — to recover profits not obtained from the bid itself.

Due to the time taken to amend contracts (and the deceptive nature of these bids), these change-order abuses are much more costly for companies in the long run. Let’s take a moment and recognize that deceptive practices aren’t necessarily intended to be malicious or purposeful.

This type of behavior is conditioned into the market, however, and many companies might not even fully recognize its pervasiveness: More than 40% of United Kingdom-based businesses have received fraudulent invoices driven by gamification and cost-recovery incentives. Unfortunately, identifying these problems in advance is difficult. How can you truly distinguish between a valid change order and one that’s created solely to recapture costs later?

At the other end of the spectrum, companies that leverage few controls leave themselves exposed to human error and make it harder to identify fraudulent transactions. A lack of controls might make for a more agile environment — but if invoices arrive via email, fax and paper forms, then it’s virtually impossible for a company to monitor its supply chain properly.

This issue doesn’t stop at invoice format. There’s also a lack of maturity in the systems processing the invoices themselves. Many companies deploy some form of enterprise resource planning (such as QuickBooks), but lighter controls in this area mean they likely don’t monitor and observe transactions in a way that helps them swiftly recognize fraudulent activity. These process gaps make it easy for value-eroding behavior to slip through unnoticed.

When nearly 45% of fraud cases are only discovered by accident or because of a tipoff, this means companies need better systems in place to mitigate supply chain risks. Artificial intelligence provides an answer.

Getting to the Root of Supply Chain Fraud

Some obvious value-eroding behaviors are easy to spot, but more complex examples require too much oversight for manual detection.

Consider a field services company that delivers products to a designated site: Let’s say its delivery trucks can hold 10 units and a company purchases 100 units to be delivered weekly. This task should only require 10 trucks, but an invoice arrives once a month with an 11th delivery fee.

When pressed, the vendor mentions that loss occurred at the site, which warranted another truck run. The buyer suspects the company is underfilling its trucks rather than facing spillage, but it lacks the resources to prove it. Because of the supplier’s bargaining power in a supply-constrained market, the buyer’s options are limited. It’s forced to pay up.

This is where AI comes in. Remember that human behavior is never perfectly random. AI can spot patterns in that behavior and identify new or unanticipated value-eroding behavior through its pattern-recognition capabilities. In turn, companies can better predict when this behavior will occur and leverage nonstandard approaches — something other than position power or the ability to apply pressure — to resolve any issues. 

More and more companies now recognize this and have implemented analytics to monitor supply chain waste and fraud and prevent expected types of value-eroding behavior. Just over a quarter of companies used analytics in 2014, and this figure rose to 35% in 2017.

We should also note that fraudulent business-to-business dealings aren’t like credit card fraud, where ostensibly fraudulent transactions are denied immediately. Spending analysis is great for looking backward (but not forward) — and reconciliation costs much more than taking preventive action.

In all, an AI-powered supply chain digitizes trust by using data to provide a complete picture of the supply chain, and emotion doesn’t factor into its judgment. Humans can miss minor errors or rationalize them away, but an AI-based algorithm will more easily spot mistakes. It can find the needle that nobody knew existed in the haystack no one ever saw.

Using AI for Supply Chain Optimization

Although it provides clear benefits, AI alone is not a supply chain panacea. In order to effectively identify and quell value-eroding behaviors, it must be implemented and trained correctly. High-control buyers have to recognize the threat of gamification, and low-control buyers need to identify the ways suppliers might take advantage of their lack of structure.

Good data is also key to managing a supply chain effectively. An AI algorithm is a tool at its most basic level, meaning it needs the right background to be useful. Often, existing procurement and financial systems have to be improved to keep up with the technology. And in the most extreme cases, it could shift the way supply chains function entirely.

AI can find the gaps in existing management and drive supply chain optimization, all by giving businesses more insight into value-eroding behaviors. Knowing how to offset those behaviors is the subsequent challenge, and it might mean making changes to company culture. In these cases, companies might have to implement additional supply chain controls (or even loosen existing ones).

The oversight AI can provide to a business will shift the balance of power in its favor. By identifying value-eroding behaviors, AI provides opportunities for companies to engage most effectively with vendors and mitigate supply chain risk. However, shifting the balance of power is not about taking control of the supply chain environment — but empowering an organization with the tools needed to incentivize collaboration and co-value creation.

In turn, business leaders must think about the behaviors they want to see from their suppliers and how they can incentivize and monitor those. AI can do more than just monitor for fraud and single out deceptiveness: It can help revolutionize a company’s entire supply chain and add transparency. The information gleaned can be shared with supply chain partners to serve as the basis for digitized, data-driven trust. 

By elevating the relationship between vendors and buyers and ensuring mutual value, implementing AI into supply chain operations spells a prosperous future for all parties.