Richard Dean doesn’t like to talk about supply chains. He prefers supply chain network, community or ecosystem.
“The world does not work like a chain. It works like a community,” said Dean, chief marketing officer at cloud-based software vendor One Network. “The chain metaphor suggests sequential links. In a community, the members work together for individual and mutual benefit.”
This interconnectedness is reflected in One Network’s cloud-based supply chain network, which allows trading partners to share and monitor information with each other and their partners and connections, and to synchronize business processes and events from point of sale to raw materials sourcing.
One Network has partnered with the University of North Texas on a research project that seeks to expand the understanding of how companies can leverage real-time information on demand to determine their “theoretical minimum inventory” — the smallest stockpiles they can maintain without jeopardizing service or sacrificing sales growth.
It’s a potentially groundbreaking project with an ambitious goal: “to determine the theoretical minimum inventories in any supply chain ecosystem, presuming all information latency is eliminated, while asset availability and performance are optimized.”
Information latency is the lag in communicating information through supply networks. “We differentiate between physical and informational lead time,” Dean said. “Once you take out the information latency, you can take out more physical lead time. That takes you to your theoretical inventory minimums.”
Don’t be put off by the word “theoretical.” This research has solid dollars-and-cents relevance, said Wesley Randall, assistant professor of logistics at North Texas and leader of the research effort. “It’s very real,” he said. “There’s a tremendous amount of waste in supply chains. If companies can get real-time data that is actionable and allows them to understand their true minimum inventory levels, it could be a game-changer.”
Companies have long touted just-in-time supply chains’ role in minimizing inventory, but One Network CEO Greg Brady says there’s much to be done. “Despite all the major technology investments over the past 20 years, supply chain performance has failed to measurably improve, and in many cases has gotten worse. Even the best supply chains still operate with massive amounts of inventory,” Brady said when the research project was announced last spring. “The results of this research will bring into sharp focus the causes of this stagnation and more importantly, what can be done about it.”
The University of North Texas research into theoretical minimum inventory is related to what One Network does for companies such as Del Monte Foods.
Using the computer cloud, One Network captures details on lead times, manufacturing frequency, batch sizes, ordering policies and other data that are crunched by algorithms and demand signals to automatically adjust inventories at each level of partners’ supply chain networks.
For example, if a retailer expects to sell a certain number of lawnmowers, the information is shared not only with the supplier and manufacturer but also with their suppliers — and in a format they can use. Retail orders for 1,000 lawnmowers are translated into orders that tell suppliers the exact number of nuts, bolts and carburetors that will be needed and when they’re needed.
Squeezing out information latency allows partners in the supply network to base decisions on actual demand instead of forecasts, Dean said. “All forecasts contain errors,” he said.
In a retailer’s typical supply chain, a point-of-sale forecast is passed upstream to a series of vendors and suppliers, each of which makes its own forecast. Inevitably, forecast errors are introduced at each step, and are magnified by delays. By the time the information reaches the start of the supply chain, it’s a classic “bullwhip” effect.
To hedge against uncertainty, companies build more inventory than they’d need if they knew exactly what and when their customer would need it.
The University of North Texas research couples software such as One Network’s with an algorithm that quantifies the time needed to transmit relevant information throughout a supply chain network. The idea is to help companies leverage up-to-the-minute information to determine how close they want to come to theoretical minimum inventory.
“We’re taking software that is almost agnostic, coupled with a thought process,” Randall said. The software prompts a series of questions: What’s the demand? Variance of demand? Lead time? Standard deviation? The answers, coupled with the software solution, “allow you to see in very high fidelity the results of your supply chain processes.”
Some companies may decide that they’d rather spend more on inventory to ensure good customer service or to increase chances for future sales. But at least they’ll understand the tradeoffs involved, Randall said.
Cloud-based software makes it possible “to quantify the current inventory versus the theoretical minimum inventory, and to translate that into dollar signs,” Randall said. “That is very powerful at the CEO and CFO level.”
Randall’s North Texas team includes specialists in logistics, marketing, information technology and decision sciences. Their research so far has been based on companies’ information and observations. The next step would be to develop additional data that would allow researchers to develop a basic theorem that companies can use to determine their theoretical minimum inventories.
“We’d like to use these numbers to find out whether there are some large-scale relationships in the corporate data that we could gain some insight from,” Randall said. “If you assume real-time access to data and then you bring in this idea of theoretical minimums, you can see how much inventory you can pull out of your supply chain. From our perspective, it’s a lot.”