First Thoughts
  By Dan Gilmore - Editor-in-Chief  
  April 19, 2013  

Supply Chain News: A Unified Theory of Out-of-Stocks?


Ok, I am suddenly some smitten with the issue of out-of-stocks (OOS), for a variety of reasons, the most proximate being a very good videocast we had on the topic this week - more on that in a minute.

Now, when most of us think of out-of-stocks, we tend to think of the retail-consumer goods angle, either as supply chain professionals or frankly as consumers. On the latter side, it is obviously to me that the OOS problem is getting worse for many retailers of late. Anyone else go to certain stores and see as much as 75% of the pegs on a wall display empty these days? Amazing, really, here in the 21st century supply chain.

Gilmore Says:

So, mighty Walmart's supply chain is mightily challenged with OOS - no wonder the rest of us are too.

Click Here to See
Reader Feedback

E-commerce has added a wrinkle to the retail OOS challenge, as now the issue isn't just products not being on the store shelf, but now potentially "not in stock" at the e-store as well. An OOS there likely has an even bigger impact in terms of lost sales than it does at brick and mortar retail.

I remember the Gilmore family in the 1990s might call up a cataloger such as Land's End or LL Bean and place an order for an item even though we were told that item was on back order and expected in at some certain date - who does that on-line? Or who today gets a rain check at a brick and mortar store for sale items that go out-of-stock, as our moms used to do? (Thanks to Richard Wilhjem of Compliance Networks for that latter observation).

Very few, of course - you go elsewhere, or don't buy at all. So the penalty for OOS rises.

But of course, out-of-stocks are not at all only a retail issue, it is a core supply chain and inventory challenge for virtually every company, though the problem mostly cast as "service levels" outside of retail.

As part of the research I did on OOS in the last few weeks, I found an article from 2010 in the International Review of Retail, Distribution, and Consumer Research from several European academics titled "Forty Years of Out-of-Stock Research - and the Shelves are Still Empty." That about says it all. The paper documents that despite study after study, and retail initiative after initiative, the needle has not really moved.

Case in point: as we reported in March, Walmart US CEO Bill Simon recently lamented to managers that stores sales were suffering because "We run out quickly and the new stuff doesn't come in."

It turns out that Walmart has been battling the problem since a least 2011, hired consultants, and placed a special executive to focus on the issue. But, Simon said, the problem lately has been "getting worse." (Walmart, we will note, didn't exactly deny these comments first reported by Bloomberg, but said they were taken out of context.)

So, mighty Walmart's supply chain is mightily challenged with OOS - no wonder the rest of us are too.

Here's why this issue is so complex: it is an equation that involves forecasting, "long tail" management, retail in-store execution, uncertain and/or difficult to calculate financial impacts, different impacts depending on product category, different impacts on retailers versus manufacturers, the Bullwhip Effect, the Perfect Order, vendor variability, store inventory accuracy, overstocks, collaboration, etc.

What have I left out?

It is a complex ecosystem for sure - yet I believe a "unified theory" (with a nod to Einstein) can be developed.

Here's a clue: In the videocast earlier this week, Joe Shamir, CEO of ToolsGroup, used a slide along these lines, which showed how less than 100% service levels at the vendor lead to more service degradation at downstream echelons all the way to the store shelf.


An important point from this graphic is that it isn't just an OOS problem - there is an equal and opposite reaction in terms of overstocks - too much inventory - either from the same issues that cause OOS or as a tactic to minimize the level of out-of-stocks.

In an upcoming videocast, MIT's Dr. David Simchi-Levi is going to among other excellent insights compare the differential impact on supplier fill rates and variability on required inventory levels. He will get to the core of "what is the purpose of inventory in the supply chain?" That is a question that more and more companies are really starting to ask, and it ties directly to OOS in every sector.

The first research on out-of-stocks is thought to be found in a 1968 article in Progressive Grocer magazine, and indeed much of the focus on the topic has been in the area of consumer packaged goods. That includes two famous studies, the last released in 2007, by Thomas Gruen of University of Colorado and Daniel Corsten of the IE Business School Madrid.

That latest Gruen study found out-of-stock rates in the CPG sector of 8.3%, slightly worse than the level found in 2002. The 2002 study also quantified consumer response to an OOS, as shown in the graphic below.

But of course, the response varies significantly by product/retailer type. If a can of soup is out of stock, a shopper might easily grab another flavor from the same vendor (good for the vendor and retailer), or the flavor they want from another brand (not good for the vendor, good for the retailer). If the size or color a shopper wants is missing at an apparel store, it is much more likely to result in no substitute purchase at all. As Shamir pointed out in this week's videocast, for long tail items, where demand is very intermittent, a stock out might result in no lost sales for some period, because there is no demand.

I am going to leave it there right now. But my research and listening to some of these experts convinces me that we can better tie the store side issues with the supply chain issues, and move the OOS ball further down the field in a more integrated and mathematical way. More soon.

Why haven't we done a better job improving out-of-stocks? Is there an opportunity to do a better job of connecting all the dots? Let us know your thoughts at the Feedback button (email) or section (web form) below.


Recent Feedback

I was just noticing yesterday how many gaps are on the shelves at my local well-known drugstore. One favorite item hasn't been there in months. Another not-good-for-customer is the trend toward "down-branding" -- reducing the number of brands in the store, while perhaps increasing the product versions.

Yes, there is an opportunity to do a better job connecting supply chain nodes. As the principles of lean manufacturing -- which already encompass supply chain relationships -- become well understood and embedded in daily practice, entire supply processes can be accelerated -- time as much as inventory is the factor. The principles have been codified as "lean supply chain" and written about in several strong books and in periodicals covering lean. So research is available now.

CSCMP is a sponsoring partner in one of the largest lean conferences held each year. APICS, which initially opposed pull systems to prolong its marriage to MRP, has come to support lean.

Yes, MRP and long range forecasting are part of supply chain execution. But the integration of forecasting and quick response operations up and down the chain could do much to fill shelves faster as well as increase cash flow.

(This is leaving out the employment over-downsizing at retailers and the resulting resentment and fear carried by folks on the job. More trust would increase caring about the job and whether stock makes it onto the truck or onto the shelf.)

I think your article is a brilliant and realistic view of the situation.

Karen Wilhelm
Lean Reflections
Apr, 19 2013

One of the consequences of supplier and retailer cost "optimization"/reduction, expanding product portfolios and increased channel differentiation is a reduction in the accuracy of inventory placement and timing. Each element has increased variability and our ability to mitigate that variability has not kept pace or has been constrained (longer runs due to cost reductions in manufacturing combined with a longer tail of SKU's to produce. Technology can only do so much; supplier and retailers must create a more effective planning "environment" to begin mitigating those elements that are increasing variability. 

Jeff Stites
VP Supply Chain and Logistics
Diamond Foods
Apr, 19 2013

I am a self-proclaimed retail junkie and unfortunately for my wife salivate at the topic of OOS. I firmly believe that some retailers can capitalize on the OOS with behavior modification, especially soft line retailers.

E.g. The Banana Republics or similar when faced with OOS situation can promise to ship the item for free to the consumer. Yes, they might incur some additional costs, but at least they will not lose the sale. Retail statistics indicate that the big box stores and malls in general are going to facing a decline in foot traffic due to increasing online sales, so in order exist they will have to begin to transition their customer and staff towards a greater online presence.

Raj Daniels
Strategy & Business Relationship Management
Not Provided
Apr, 20 2013

Really appreciate your article on this subject as we have been researching and working this problem at Clemson University in our model factory since the late 90’s.

Our original R&D work and  proof of concept solution was funded by the Department of Defense. A small group of us has  formed a spin-off company to continue the perfection of the solution and make it available beyond the DoD. The Army has successfully used the solution for all combat apparel since 2005 and three commercial companies have also had great success with it.

Our approach is a SAAS  tool that synchronizes consumer demand through distributors and manufacturers up as many tiers of the extended supply chain as we can get to work together (and getting two or more companies to work together in a different  manner  is extremely difficult). We achieve synchronization based on the simple cocept that everyone who orders replenishment, determines what to ship in VMI situations, or schedules  production should order,  ship, or make next that which is in shortest supply downstream of their location in the supply chain. We do this by creating standard structured “supply chain sections” throughout the supply chain as basic, but very flexible building blocks that use the same algorithms to compute what should be ordered, shipped, or manufactured next subject to existing constraints and always balanced in days-of-supply. These  computations are based on current inventory levels, lead-times, and strategic buffer stockage objectives  set in days-of-supply. We use forecasts, but minimize  the common problem of forecast error by running the tool frequently to re-balance all supply chain sections and thus replace most of the forecast with demand pull driven by downstream shortages.

Our results virtually eliminate stock-outs of replenishable products, shorten replenishment lead-times so that many non-replenishable products become replenishable during the selling season, reduce inventories by 50 to 90 percent, virtually eliminate expediting, and minimize resources required to generate replenishment orders as well as shipment and production schedules.

If you are interested in our work, we would like for you to visit us and learn more about our approach and our accomplishments at Clemson.

Bill Kernodle
Site Director
Clemson Apparel Research
Apr, 20 2013

The Out-of-Stocks problem in the CPG Industry has been around a very long time and will not go away until we completely change our thinking on how to manage the end to end retail supply chain.

First, we need to acknowledge some basic truths learned the hard way: 

1.       The retail store is the beginning & the end of the retail supply chain. It’s the beginning of information flow and the end of product delivery.

2.       Every CPG retail supply chain is actually two supply chains.

a.       The physical supply chain: the movement of materials & products from supplier to manufacturer to retailer.

b.      The information supply chain: the movement of consumer demand information from retailer to manufacturer to material supplier.

3.       Today, the information supply chain is broken as every retail node making up the retail supply chain is disconnected from the manufacturer’s  nodes and each node is buffered with inventory thus slowing down the ability to respond to true store demand changes.

Having wrapped our brains around these basic truths we then need to take advantage of major breakthroughs that have occurred recently.

 Today, it is possible for a retailer and a manufacturer to model and connect their  retail supply chain inside a single computer and drive it, end to end, from a unique store level sales forecast.

 Today, it is possible  to  resynchronize every node making up a  retail supply chain  every day based on changes in store level  consumer demand and thus completely eliminate the “Forrester Effect” (also called the Bullwhip Effect).

 Today, it is possible for a retailer and a manufacturer to use this model of their retail supply chain and manage it as if one company were managing it.

 Today, it is possible for a retailer and a manufacturer, using this “business model”, to create an Integrated Business Plan, identify gaps and take corrective action while there is still time.

Today, there are no secrets anymore on how to manage the end to end retail supply chain, from the store to the factory, and significantly lower store out-of-stocks as well as reduce retailer’s and supplier’s inventories and operating cost at the same time. 

Today, it is possible to achieve a 99.8%  On Shelf Availability (OSA) at the store while reducing overall supply chain inventory buffers at the same time.

 Today, there is a proven and well documented business process, and enabling technology, to actualise all of the above and it even has a name: It’s called “Flowcasting the Retail Supply Chain”

Andre Martin
Not Provided
JDA Software
Apr, 20 2013

I agree with a lot of what Andre Martin says. The current problem is that the information supply chain is lacking, which causes out-of-stock on a regular basis. A possible resolution to this problem is have retail supply chain implement a system where it constantly checks what products have been bought off the shelves and updates the company digital inventory so that at a certain threshold, it will automatically put in a re-order depending on the re-order point. Companies have always focused on become more efficient at the physical supply chain. The importance of cutting transportation cost can't be discounted, but companies won't be able to take advantage of a good physical supply chain if their information supply chain is lacking. If Wal-Mart, who has historically been known to run a successful supply chain, struggle with informational supply chain as well, then this is definitely an issue that needs to be addressed. Loss sales due to OOS have been plaguing retailers for quite a while now and if they can make a change in their supply chain, there will definitely be a huge change in profits. 

James Du
Not Provided
Not Provided
May, 01 2013