What is the most common problem in warehouses?
It is not poor picking performance. It’s not the labor shortage. It’s too much inventory. They are full.
I have had the opportunity to visit many warehouses over the last few years. And with very, very few exceptions, they all had one problem in common: they were full.
And that’s a big problem. Just in case you were wondering. I will say it is the biggest problem most warehouses have, because a full warehouse brings with it a whole host of operational issues downstream in the chain of cause and effect. But let’s examine this step by step.
The Most Common Problem in Warehouses: They are Full
It is perhaps the most frequent observation I have made in my ten years or so of working with customers in intralogistics. Most warehouses are full. Companies undertake warehouse automation projects because they need a way to store a growing amount of inventory in a limited amount of space. (This is one of AutoStore’s strong selling points). Companies are acquiring additional warehouse space, or they contract 3PLs because they can’t find a place to store their products. Pallets sit in driveways and receiving areas. Warehouse employees struggle to find storage locations for incoming goods.
Let’s talk about why this is such a big deal.
Upstream and Downstream, Causes and Effects
Two things are important to understand when it comes to high filling levels of warehouses:
First, it is the result of poor decisions upstream. From this perspective, a high fill rate is merely a symptom.
Secondly, it leads to various problems and inefficiencies downstream. From this perspective, it is a cause.
The fill level of the warehouse is one of the most important determinants of its performance. And yet warehouse operators often have little to no influence on it. The fill level of the warehouse is determined by five variables:
- How much space is available (yes, you knew that)
- How many SKUs the warehouse has to store
- How much inventory of each SKU the warehouse has to store
- How the inventory is kept
- Peak load due to promotions, seasons, etc.
If anything, a warehouse manager might have a say in the first and the fourth question, assuming he has been around long enough to influence the warehouse design or investment decisions. But as far as the other three points (and sometimes all five) are concerned, these decisions are made outside the realm of warehouse management:
- The number of SKUs offered is a strategic decision made by executive management with support from sales and marketing (well, hopefully).
- How much inventory is held for each SKU is often the result of decisions made in the purchasing department.
- The variability of demand is usually influenced by marketing and sales through promotions, if not by external market forces.
- Purchasing, in turn, increases inventory before the peak season to ensure the availability of the goods on offer.
Warehouse management then has to deal with the results of all these factors. It finds itself in a sandwich position between purchasing and sales and nobody asks for its opinion.
Downstream Effects of High SKU Count and High Inventory Levels
Of course, the warehouse manager must take the blame if the productivity of the warehouse workers is deemed insufficient or if the storage costs are too high. And certainly there are better and worse ways to organize and manage a warehouse. But let’s be clear about the implications of SKU count and inventory levels in a manually operated warehouse.
One of the problems with a high SKU count is that each SKU also brings with it a certain amount of inventory. At least most of the time this is the case. Even manufacturing companies that embrace Lean and source some of their components just-in-time (JiT) from their suppliers can only benefit from no (or low) inventory for a small number of SKUs, as JiT requires a significant amount of supplier training and administration and is therefore only feasible for a small number of suppliers. So with a high number of SKUs, it is likely that there will be inventory for most, if not all, SKUs.
The obvious consequence is that with more SKUs, there will be more inventory. But you already knew that. However, what I find in most warehouses I visit is that inventory levels are higher than necessary and warehouses are filled to the brim.
Inventory has both direct and easily measurable costs and indirect, hidden costs. Let’s discuss some of the direct and indirect costs. This list is not exhaustive.
Direct Cost: Obsolescence
Obsolescence can be a big problem or not a problem at all, depending on the industry. There’s a fantastic HBR article from 2005 about inventory costs at Hewlett-Packard1. The summary is that HP couldn’t make money in its PC business in the 1990s because of high inventory costs. A major factor in inventory cost (which the company did not fully understand and didn’t manage well) was obsolescence. Computer hardware depreciates quickly due to advances in technology. That’s not your problem if you’re in the toilet paper business.
Direct Cost: Perishability
Inventory can become worthless not only due to obsolescence. It can also degrade in quality or spoil altogether. Again, that’s a bigger problem for some industries than others. In food production and food retail, spoilage is a major cost driver across multiple stages of the supply network. It is also a problem in the storage of pharmaceutical products. These are the obvious cases. But the quality of many other products can also deteriorate over time. Plastics or rubber, for example, can change some of their properties over time. The coating on glass can deteriorate. Metal parts can corrode. And so on.
Direct Cost: Warehousing
More SKUs and more inventory require more storage space and more racking and shelving. More warehouse space requires more energy for climate control. Larger warehouses come with higher costs for fire protection and insurance. They are likely to be taxed more. Maintenance and inspections are more expensive.
Direct Cost: Lost Interest
One aspect that can easily be forgotten after a long period of low interest rates is the fact that capital tied up in inventory doesn’t earn interest. As the value of inventory is often in the millions, at current interest rates (December 2023) the lost interest will easily run into the tens of thousands of Euros for many companies.
Hidden Cost: Opportunity Cost
Somewhat more subtle and less visible than the loss of interest is the larger problem of opportunity costs. Opportunity cost describes the lost profit from the best investment alternative that you did not choose, and it is perhaps one of the most underappreciated concepts in all of economic theory. The capital invested in inventory cannot be invested in another way that would yield a greater benefit. With inventory, the case is particularly clear. Since you only need a certain amount of inventory to maintain your service level, any additional amount of inventory is simply wasted money. If you were to hire additional sales staff or invest the money in revamping your website or in a better warehouse management system, you would be better off.
Hidden Cost: Higher Process Cost
One of the biggest problems is also one that, except in extreme cases, is the least visible. High SKU counts and high inventory levels reduce employee productivity:
- Each additional SKU brings with it a certain amount of inventory that requires storage space.
- Therefore, each additional SKU increases the walking distances in the warehouse.
- Therefore, each additional SKU reduces labor productivity. Labor productivity is defined as the amount of output you get for a certain number of labor hours of input.
- In fact, there is a clear linear relationship between an increase in SKU count and a decrease in labor productivity.
- Lower labor productivity, of course, directly translates into higher labor costs, as you need more employees to get the work done.
When the number of SKUs increases, labor costs in the warehouse are not the only cost increases:
- You need to conduct more contract negotiations with suppliers and therefore need more purchasing staff;
- Your purchasing conditions deteriorate as your sales (and purchasing quantities) are spread over a larger number of products and suppliers.
- you increase your risk of product obsolescence, spoilage and other reasons for write-offs;
- taking stock takes longer.
And let’s not forget that most of the SKUs we add to a portfolio are slow movers. So the significant increase in costs we can expect with each additional SKU will not be offset by significantly higher revenues in many cases.
By the way, in an automated warehouse, an increase in SKUs does not lead to higher labor costs for picking. When warehouse workers pick at goods-to-person stations, it makes no difference whether they pick a portfolio of five thousand, ten thousand or one hundred thousand SKUs. However, it does lead to higher CAPEX. More SKUs require larger systems and larger systems are more expensive. Some systems, such as zone picking systems, only make sense for a smaller number of SKUs, and with a larger number of SKUs, you will need to resort to much more expensive goods-to-person systems. Higher capital expenditure means you spend more money for a given amount of output, which in turn means you lower your productivity. Instead of lower labor productivity, you now have lower productivity of capital. Since the only thing that matters is total factor productivity, which includes both capital and labor, a higher number of SKUs results in lower productivity, whether in a manual or an automated system.
Examples of High Inventory Situations
Let me illustrate my points with three cases. All three are recent projects. They are not even extreme cases. I am confident that if we pick ten random warehouses, we will find more serious cases than those described below.
During a warehouse visit, I observed a warehouse worker driving around with a pallet that had just been registered in goods-in. He was looking for a storage position in the pallet rack. There was none (which he knew because the WMS told him so). But he drove around to find a pallet position that was empty enough for him to consolidate the goods from the two pallets on one. When he found one, he began transferring inventory from one pallet to the other, and the pallet position in the rack now contained inventory from both SKUs. The entire procedure took more than ten minutes. More than ten minutes to put away a pallet in a (rather small) warehouse! And it was certainly not the only pallet that had to be put away that week.
So, storage was extremely slow. But as you can certainly imagine, the division of storage locations described above can also have an impact on picking performance: Because there is more than one product per storage bin, you have to correctly identify the products at the pallet location (which takes time) and there is an increased risk of picking errors. And, of course, there is no such thing as ABC zoning in a warehouse that full.
In one project, the company was investigating the use of warehouse automation (at least €2.5M investment) as they were running out of space. In addition, the company had already purchased additional warehouse space that would soon be operational and wanted to better understand how to best connect the new warehouse to the existing one.
When looking at the data, it turned out that it was not just the (< 500) number of lines picked each day that would make it difficult to justify an investment in automation. It also turned out that with an average inventory reach of more than 100 days and around 10% of SKUs not sold in the past year, the problem was not a lack of space, but excess inventory. They could virtually halve the warehouse space immediately (and reduce it even more after reviewing the product portfolio and supply agreements) without customers noticing. The company did not need automation or the extra warehouse space it had just acquired.
A 3PL had just spent more than €5M on warehouse automation (some goods-to-person picking and zone picking). Touring the warehouse, I was surprised by some odd design and process decisions (five GtP stations for three miniload aisles, manual scanners at the GtP stations, zone picking for 60,000 SKUs, single-order picking in a huge, manually operated area for small parts…), but also by the fact that the warehouse manager told me that they found out right after going live that they needed to expand the automated part of warehouse very soon, both because of performance and lack of space.
During our tour of the manually operated part of the warehouse, we also saw order pickers picking from pallet racks ten meters up. The warehouse manager told me that the pickers were complaining about having to pick so often from high racking positions (it looked quite unergonomic and dangerous) and that this was clearly not contributing to their performance (which management were concerned about). This was despite them saying that they had introduced ABC zoning years ago and that the low shelf positions should be approached much more frequently.
Again, a look at the data revealed quite a few issues: a number of SKUs had not been sold for five years; in fact, several of them had never been sold. The GtP system was at capacity because the fast-moving products were being fed into the automatic system instead of the slow-moving products that were clogging up the manual area (and the fill rate of the double-deep miniload was so high that the systems sometimes came to a halt in a deadlock). And the pickers had to pick so frequently from high rack positions because the ABC zoning that was set up years ago could not be maintained because the fill level of the pallet rack was so high that every new pallet that came in just had to be put in any location that became available, regardless of which zone it would belong to. They simply had too much inventory, and to make matters worse, the inventory was poorly organized.
Conclusion and Recommendations
In this article, we have examined high inventory levels as a major problem in warehouses. We have found that high inventory levels have some direct costs that are easily visible and measurable. Worse, however, is that they also cause indirect costs that manifest themselves through inefficient operations downstream. These downstream effects on performance cannot be fully understood without reference to benchmark values, which no one has, but we can say with certainty that they are significant. Not least because they trigger important investment decisions, such as investment in automation, which often enough would not be necessary if the manual system was properly managed and inventory levels were lower.
However, we have also found that the causes of high stock levels are usually not to be found in the warehouse – but in management, sales, marketing and purchasing. These are the people who decide on the number of SKUs, purchasing quantities, and promotions.
Since logistics is not directly involved in the inventory decisions that affect them, my recommendation is as important as it is cliché: you need to break down silos and create reliable lines of communication between logistics, purchasing and sales. All three functional areas are tightly interlinked but often act as if they have nothing to do with each other. The incentives offered in many companies are clearly not aligned between these functional areas (more on this in a separate article). Solving high inventory levels and all the associated downstream problems will simply not be possible without communication and a degree of objective alignment, as well as management’s understanding of the causal relationship between their decisions and the performance issues in logistics.