An Inflection Point in the Consumer Goods to Retail Supply Chain? Part 2

I can't believe it's been more than two months.

In September, I wrote column titled An Inflection Point in the Consumer Goods to Retail Supply Chain? that a lot of people read, tweated and commented on. I ran out of space and promised a part 2, but a lot of other things got in the way and I am finally delivering that follow up this week.

At least I made it before the end of the year.

So here is the summary of part 1:

In my view, the consumer goods to retail supply chain - the most prominent one we have - has been characterized by a high degree of variability since its inception. This variability occurs for a variety of reasons, as I will discuss in a bit.

But retailers live with a level of variability today that is just unimaginable in most other supply chains, with wide variance in accepted delivery windows, performance against those windows, fill rates and other aspects of vendor shipments (labeling, EDI, etc.)

At the highest level, this variability has persisted because the effects are largely hidden. Is there a clear view of supplier variability and its connection to out-of-stocks on the retail shelf? Not really. Heck, much of the time with in-store out-of-stocks the merchandise is actually in the backroom, so if retailers can't connect those dots effectively, how can they connect OOS's to supplier and PO variability?

Most retailers don't know how much of a PO a vendor actually shipped until an advance ship notice arrives shortly before the goods, and in some cases not until the goods arrive. That might be OK if the shipment was on-time and in-fill most of the time, but in retail it is simply not.

Other sectors of course have supply variability too, a challenge exacerbated by offshoring, but nothing like retail. And most non-retail sectors have been working to reduce variability for years. Variability was one of the "3V's" (along with visibility and velocity) Art Mesher famously wrote about in the late 1990s while at Gartner , but I think most retailers missed that discussion. Meanwhile, I remember writing in the very early days of SCDigest how defense contractor Raytheon was using Six Sigma to substantially reduce supplier variability. Haven't exactly heard many stories like that coming out of the retail sector.

There are other odd dynamics. For example, Greg Holder, CEO of Compliance Networks, has talked for many years about how retailers shouldn't set their "chargeback" levels too low for failure to meet thresholds for things like on-time and fill rates, because if they do, vendors short on inventory will ship to the retailers with higher penalties and under serve the ones with more lenient chargebacks

Put that in your supply chain cap and ponder for awhile.

But as I said in September, the times may indeed be a changing. Some developments:

A major regional grocery chain has about finished rolling out a highly integrated store-level planning system - among the most comprehensive that has been deployed by any retailer to date. Now that it has leaned out its own store-level inventory and replenishment planning approach, where is the inventory risk? "Vendor failure," it says.

Now when a vendor shipment is late or the fill rate is less than 100%, the chance of an out-of-stock rises versus the previous approach, which had a lot more buffers - and the grocer is deploying a vendor compliance program not common in the grocery sector to reduce that potential vendor variability.

In June, Tom Shortt, Home Depot's senior vice president of supply chain, was quoted as saying its stores should "Get comfortable with days of inventory, not weeks." HD is targeting sales growth of nearly 15% by 2018, but wants to keep inventory levels flat or slightly down. That Leaning out of inventory can only work if supplier variability is reduced, right?

Also this past summer, Target's COO John Mulligan said the chain was going to attack the out-of-stock problem in large part by reducing "variability" in its distribution centers, especially with regard to supplier deliveries. 

Mulligan said that "An unacceptable number of vendor shipments were received by our DCs either too early or too late," adding that "We have been collaborating with our vendors to increase the percent of shipments that arrive on the correct date and we have already seen meaningful progress."

That progress includes a 50% reduction in out-of-stocks for ecommerce-only items its DCs.

A bit earlier, rival Walmart wrote in a blog post for vendors that it was reducing the window for which a shipment is considered on time from the current four days to just two, starting in February 2017.

In addition, the fill rate requirement is being raised from 90% to 95%, measured at the case fill level. Failure to hit either metric results in chargebacks - a practice that Walmart was actually very late to embrace versus other retailers.

Ok, so what we have thus far are:

(1) Operational changes - retailers going more Lean and thus requiring reduced variability in terms of supply.

(2) What appears to at long last to be a recognition by some leading retailers that variability really is the enemy, and that it can be attacked by simply focusing on the problem as well as using the stick of chargebacks and enhanced performance metrics/supplier grading to reduce that variance in terms of vendor performance.

All good, from my view. But is it that simple? Of course not.

While taking the steps outlined outlined will help, retailers also create much of their own supply chain variability.

It starts with the famous BullWhip Effect, first identified by MIT's Jay Forrester in the early 1960s, and then rediscovered in the early 1990s by Procter & Gamble and Dr. Hau Lee of Stanford. In great simplicity, the Bullwhip Effect refers to the fact the while actual consumer demand is generally fairly steady and predictable, for a variety of reasons (including simply a lack of supply chain discipline), orders to first tier suppliers and then their suppliers tend to be erratic and hard to predict.

This of course presents those suppliers with real supply chain challenges. They could bulk up on inventories to buffet the order variability, but that is expensive. Retail vendors do some of that, but more often try to cope by madly shifting production schedules around and expediting - which only goes so far. The result (surprise!): problems with on-time deliveries and fill rates, meaning high levels of variability from the retailer's perspective.

But not only are the actual orders from retailers too vendors erratic - they in general do a very poor job even of forecasting those orders, though there is really no reason they could not do a much better job of that.

This lack of visibility to future order flow is what spawned the Collaborative Planning, Forecasting and Replenishment (CPFR) "standard" (that's an overstatement) in the late 1990s. I suppose CPFR has produced some benefits, but clearly it never met early expectations - or much solved the variability problem in the consumer goods to retail supply chain.

So while Target, Walmart and maybe other retailers should be complimented for attacking part of the variability challenge, the other side of the equation also requires attention.

That means driving the supply plans truly from the shelf back by both parties, providing time-phased order plans such are commonplace in manufacturing but very rare at retailers (evolving into the concept of "supplier schedules"), and providing retail vendors the type of forecast visibility that Canadian Tire has admirably been doing for years for its retail suppliers.

I am out of room, and will thus save that discussion for the last part of this series in early 2017.

You will find some feedback below from part 1 that I think you will enjoy - would love some additional comments here on part two.

What's your reaction to Gilmore's column? Why have retailers lived with high variability for so long? But are retailers really the cause of much of their own variability? Let us know your thoughts at the Feedback button below.