The inventory challenge
From the June 2021 print edition
Running shops faced an unprecedented supply problem last winter when COVID-19-related closures forced health club goers out into the cold for their workouts.
“We were sold out of men’s winter running tights by November,” says Lynn Bourque, owner of The Runners Shop in Toronto, “and there was no way I could get more. None. And this wasn’t just one brand – I’m talking all of the brands.”
For retailers like Bourque who order their gear a year in advance, these have been challenging times. Prolonged shortages of everything from masks to bicycles to building materials, and the seeming inability of manufacturers and their distribution networks to respond to changing circumstances, have been remarkably persistent.
The kneejerk reaction, reinforced by the popular press, has been that low inventory strategies are to blame. Industry stats, however, suggest otherwise. According to Statistics Canada, the inventory-to-sales ratio for Canadian manufacturers has risen steadily over the past decade – a trend that has continued during the pandemic.
“There are more people holding more inventory than they would have 12 months ago to be able
to have that confidence that they’ll have the product when the customer wants it,” says Shari Diaz, innovation, strategy, and operations director, IBM Sterling, based in Columbus, Ohio. Low interest rates, Diaz notes, have recently moved that trend upwards.
Inventory is always a balancing act, of course. While insufficient reserves can make a company vulnerable to market fluctuations, excessive inventory can tie a company’s hands when unexpected circumstances arise. When consumer behaviour changes in unpredictable ways, as it has during the pandemic, making the right call gets more difficult.
In a whitepaper published at the outset of the pandemic, Robert Martichenko, CEO of LeanCor Supply Chain Group, addressed the alleged toilet paper shortage, noting that it was erratic consumer behaviour, not insufficient inventories, that was to blame for the empty shelves.
“There is no shortage of toilet paper,” Martichenko wrote. “The inventory has simply been replaced into the homes of individuals who feared grocery stores may close.”
“Over the last year, demand was all over the place as well as supply,” says Mississauga-based Gavin Davidson, product marketing director at ERP software vendor Oracle NetSuite. “So I think it’s really about visibility into the supply chain more than anything.”
The uncertainty has affected all levels of the supply chain. Last October, over 400 manufacturing executives responded to a Thomas Industrial Survey titled “A Year in Review: How the Manufacturing Industry Adapted in 2020”. Overall, 87 per cent reported that key markets were disrupted by COVID-19, and 53 per cent reported supply chain disruptions.
One of the key lessons from the pandemic was that planning stocking levels based on the previous year’s sales can be a recipe for disaster. Machine learning and other artificial intelligence (AI) technologies are helping companies improve both the speed and the quality of their predictions.
Amazon has been the poster child for such methods, deploying machine learning and other AI technology in decisions around warehousing, distribution, transportation and procurement. A key advantage of AI is that models can be built with an unprecedented variety of inputs including structured data from ERP systems, IoT sensors, partner data, and external data from weather and social media sites. Furthermore, AI allows decision makers to remove human biases that may limit their choices.
Most companies, however, are a long way away from being able to take advantage of these capabilities. Many are taking the first step of determining what inventory they have. The pandemic-mandated shift to e-commerce has accelerated such efforts.
“Grocery stores had to really switch to the buy-at-the-curb pickup model,” says Shari Diaz, innovation, strategy, and operations director, IBM Sterling, based in Columbus, Ohio, “but they didn’t know what was on the shelf. But how can you let someone fill a virtual shopping cart when you don’t know what you have in the store?”
Knowing what you have sounds simple enough, but when data is siloed, for example, between channels, or between different divisions of a company that grew by acquisition, significant work to integrate that data may be required. Software vendors supporting the omni-channel approach, which seeks to unify the customer experience across different channels, are rapidly evolving new tools to create a single view of inventory across the organization.
AI tools are also rapidly becoming accessible to mid-sized companies. “Our approach has always been to use those advanced technologies to deliver out-of-the-box solutions to customers,” says Davidson. “So, any NetSuite customer can basically enable a feature that, behind the scenes, engages Oracle artificial intelligence / machine learning (AI/ML) technologies to start modeling their organizational data. One example is that you can start to look at what we call predictive risks – what’s the risk that this purchase order is going to be late?”
Dealing with fragility
Some of the critics who blame pandemic shortages on low inventory point a finger at the just-in-time (JIT) approach, which is a pillar of the lean manufacturing system pioneered by Toyota. JIT, they claim, is purely a cost cutting strategy.
This argument reflects a fundamental misunderstanding of JIT according to Ken Eakin, an Ottawa-based management consultant. “Low inventory is a by-product of a production strategy, but it’s not the goal,” says Eakin. “The goal is to enable faster production, and to enable flexibility by producing in smaller batches.”
The real issue with shortages, Eakin says, is North America’s dependence on long, fragile supply chains built on moving goods halfway around the globe in container ships. “Making stuff in China for the North American or European markets is the opposite of JIT,” says Eakin, “because it’s got to go on trucks and trains, and through ports and depots, and anything that goes wrong at any point in that chain would cause the system to break down. With JIT, you minimize fragility by bringing suppliers closer to you.”
The Thomas Industrial Survey revealed some anecdotal concerns about supply chain fragility, and a surprising 52 per cent reported that they plan to re-shore or nearshore their supply chains. “Reshoring or nearshoring, or having some redundancy built into the supply chain, are things people are thinking about,” says Davidson.
JIT can be a major force in making reshoring economically viable. “By working with less inventory through a JIT production system, you can save money on transportation, and you reduce the risks of long supply chains,” says Eakin. “So, I think reshoring could be done in a lot of industries – maybe not for commodities like paper towels, but certainly for many other things. When you look at the total math of the entire supply chain, it makes a lot of sense to bring it back domestically, or perhaps to Mexico.”
JIT could also be part of the solution for relieving pandemic shortages. “A JIT producer can switch within a few days to a different model that might be in more demand because they don’t have all the sunk cost of pre-purchased inventory,” says Eakin.
A JIT producer might have been able, for example, to help alleviate last November’s winter running tights shortage.
JIT, of course, is not a magic bullet. Agile companies that work with short lead times and low inventories need timely, accurate information to make fast decisions. High visibility of inventory, therefore, will become even more critical for companies that move in that direction.
“We’re maybe a long way in many cases from being able to leverage technologies like AI and machine learning,” says Davidson, “but a good first step might be understanding what data you have available to you, and also, what information your suppliers can provide.”