Can technology save the retail supply chain?

We’ve all now witnessed firsthand what can happen when a basic, bulky household staple suddenly sees wildly increased demand (for reasons both real and imagined). Shoppers have snapped up rolls of Charmin and Cottonelle every time they are restocked; while rumors of continued shortages and images of stockpiling lead to ever-more panic-induced buying.

However, the Toilet Paper Problem, experts say, is only the most notable and visible example of the basic supply chain challenges retailers have already faced in the era of the COVID-19 crisis. In fact, much of retail’s coronavirus-related challenges have to do with supply chain issues, according to a recent survey by Digital Commerce 360.

“This is that black swan event that a lot of retail supply chains couldn’t possibly plan for,” says Bob Ferrari, managing director, the Ferrari Consulting and Research Corp. “Demand for some products has been as much as three, six or 12 months of normal supply, evaporating in a matter of days.”

Typically, after the “pantry loading” that occurs, say, when a product goes on sale, shoppers don’t buy that product again for a while. However, in today’s era of prolonged confinement, the purchases continue out of fear of shortages and also because household consumption does actually go up, says Lana Klein, partner and strategic analytics practice lead at Fractal, which offers AI-driven analytics to retail and CPG companies. “It’s extraordinary, with kids at home and people eating more, how overall consumption of certain products is increasing,” she says.

On the supply side, there is also the “bullwhip effect,” adds Ferrari. That is, big swings in consumer demand reverberate up the supply chain, leading retailers, distributors, wholesalers and manufacturers to order more goods, causing a domino effect that leads to shortages or inefficiencies.

“It’s a complex challenge that companies are facing now,” says Kevin Sterneckert, chief marketing officer for Synchrony AI. “Most of today’s forecasting supply and demand systems are built around triggers that help understand consumer behaviors and provide insight into what the likely demand will be, as well as figuring out how to meet the optimal result given the supply.” Now, however, there is a huge demand in a variety of categories, with resulting real struggles to meet those demands.

[ Related: 5 ways ecommerce retailers can boost security ]

Historically, Thanksgiving and Christmas have always been modeled as the ultimate demand scenarios, says Siva Venkataramani, AVP at consulting firm Cognizant, who specializes in retail and CPG supply chain issues. “This moment is blowing all of that apart,” he says, though it does create a new data point that can be used in the future.

Large retailers have shored up supply chain technology

Large retailers have invested in shoring up their supply chains with modern technology in recent years, including cloud-based omnichannel platforms, AI-driven analytics and IoT sensors that reduce forecasting errors; provide real-time insights into machine functioning; track assets all along the supply chain; and achieve order accuracy.

Adoption went mainstream from 2012-2014 in areas such as inventory management, demand planning, supplier management/sourcing and sales and distribution, in response to macro trends including technically-savvy and convenience-seeking consumers with increased demands of the retail shopping experience. Same-day delivery, BOPIS (buy online, pick up in store) and other capabilities have become table stakes for most retailers.

“A lot of the big box retailers are well-prepared technology-wise, particularly Walmart, Target and digital-born Amazon,” says Venkataramani. “From an operational efficiency standpoint, most of these companies are pretty ready.” 

Companies that have already been on the journey of a digital transformation do have a distinct advantage right now, agrees Sterneckert, though he adds that on a digital maturity scale of 1-5, most retailers fall somewhere in the middle. “Those farther along have a leg up because they can begin to understand their inventory position and supply and demand at every point, with near real-time updates,” he says. AI-based systems, he explains, allow companies to tag special high-demand events so when the event ends it returns to a normal model so that there isn’t an oversupply of the product later on.

Those retailers that have the digital technology to support flexible fulfillment options, such as ship-to-store, order pickup and ship-from-store, will also be ahead of the game, says Carlos Castelán, managing director of retail consulting firm The Navio Group. “Maybe if you have some inventory in stores in one part of the country and demand coming from another, retailers can ship to those customers and still sell the product at full price; they don’t have to discount and take a hit to their margin,” he says.

One problem now, though, is that modern supply chain management, and the technology that supports it, has primarily been focused on “just-in-time” and lean production, producing and delivering finished goods 'just in time' to be sold. “Those firms that were the most efficient, had the lowest inventories and the most consolidated vendors are being hurt the most,” says Klein.

In addition, while advanced, multi-tier inventory management systems offer increased visibility up and down the supply chain, most companies still have disconnects and do not have end-to-end demand-to-supply visibility, from the shelf all the way to production.

They may have good visibility into their Tier 1, or primary product suppliers, but remain blinded by the lower tiers — the components that make up the product, says Ferrari.

Take, for example, a popular brand of hand sanitizer, which includes the actual product, the packaging in glass bottles or dispensers, and the FDA-required aluminum cap that seals the bottle. According to Ferrari, while the brand recently described going 24x7 around the clock to make up the difference in manufacturing due to extraordinary demand, they ran out of aluminum tops, which stopped them in their tracks. “They had to switch to the aerosol version that did not need the seal, but had to scramble for a supplier,” he says.

Also, most regional and mom-and-pop retailers, which typically have low margins, do not have the money to invest in today’s advanced cloud-based digital technologies that solve many supply chain issues. “Large CPG brands may have more leeway, so companies such as Procter & Gamble and Nestle are getting there, while high-margin businesses such as equipment manufacturers have more to invest,” says Ferrari. “When smaller retailers run out of stock, they rely on all their suppliers to have those capabilities.”

What technology can do for retailers now -- and where it hits limits

For now, retailers selling the most in-demand products will have to accept that their models may be wildly inaccurate right now and outlier detection technologies may not pick up the necessary signals. “I think down the line technology will definitely help more retailers, but right now if you had your price sensitivity model forecasts based on current promotions, that goes out the window,” says Klein. “Retailers will need to manually intervene unless they have those sophisticated AI-based systems.”

If possible, retailers should be investing in new software today for tomorrow, says Steneckert, to make sure they start to understand the real demand for the items that appear to have the largest increases in sales. “There is a list of items in a typical grocer that have had the highest increases in sales,” he says. “Those items need special care and feeding.”

Amazon, for example, is ramping up to meet its increased demand, hiring 100,000 new workers. “That’s because they have AI-based systems that understand what’s happening,” he explains. “They also changed delivery times on items so they can adjust the supply chain to focus on items of immediate need rather than those that are secondary. That’s the kind of intelligent thing you’d expect from an AI-based set of retail capabilities.” 

Yet, there is no doubt that even the retailers with the best technology are facing stiff supply chain challenges, says Ferrari, who points out that there is only so much product that can be shipped and manufactured in this time of unprecedented demand. Where the technology will help going forward, he explains, is offering early warnings and boosted visibility when subsequent disruptions occur. 

“There will inevitably be new use cases from what has happened with COVID-19, and the incorporation of machine learning technology in supply chain planning,” he says. “We’ll see some of that start to play out when all the dust settles and companies start to talk about where it helped them.”

There are already things to learn from data coming out of China, he adds, such as how companies including Alibaba have used supply chain technology capabilities, including sensors that track shipments and truck movements, to respond to conditions where everything was locked down. “They could not get deliveries and trucks to where they were needed in the Hubei province,” he says. “They got around those challenges and technology played a role.”

In the immediate term, retailers are scrambling to keep shelves stocked. “Analytics won’t help you right now, there is no time,” says Kleine. “You have to focus on the products that are really flying off the shelf. You may have 200 SKUs but five of those products are in high demand. You may have to bypass distribution centers and have suppliers go straight to the store.”

Back to the Toilet Paper Problem: If demand is twice the manufacturing capacity, retailers may have trouble even finding new sources of supply. “I’m not sure if that’s feasible, even if I go overseas,” says Venkataramani. “Will I really airlift toilet paper? And even if you go all the way back to pulp, how much can they support? Companies will have to do their best with the card they’ve been dealt.”

In the future, companies will take lessons they learned from the effects of the pandemic on the supply chain and learn to model similar scenarios supported bt using advanced technologies, says Venkataramani. “As sophisticated a country as we are, the first time dealing with this kind of planning is going to be a monumental challenge.” he says. “At the end of the day, everyone is still learning.”