Why Adoption is Slower Than you Might Expect
We are moving more goods from warehouses to consumers than ever before. And the more e-com replaces brick-and-mortar, physical commerce, the more business is done in warehouses. Accordingly, logistics service providers (LSPs, or third-party logistics providers, 3PLs), which operate warehouses on behalf of retailers, have enjoyed healthy growth for several years, and show no signs of slowing down.
As you might imagine, companies that make their living from warehouses have a strong incentive to improve operations, i.e., increase productivity or otherwise reduce costs. With half the world excited about warehouse automation these days, you might expect 3PLs to be at the forefront of technology adoption. However, while some of them are successfully drawing media attention to the few cases where warehouse automation is being used (I’m looking at you, DHL), it’s pretty obvious that 3PLs, by and large, continue to stick to manual processes. This is despite rising wages and, more importantly, increasing labor shortages. So what’s going on? Why are 3PLs not among the top customers for warehouse automation technology? This is the question I want to explore in this article.
They would like to!
The reason 3PLs are comparatively slow to adopt warehouse automation technology is not that they are not interested. In fact, they are very interested and understand that they have many reasons to embrace warehouse automation. In addition, all of the larger 3PLs have some sort of full-time innovation scout, at least one, sometimes an entire department, responsible for identifying and researching new technologies, including new automation technologies. They attend trade shows and conferences, constantly run pilot projects with startups, and always have an open ear when someone calls them with a new idea. Their slow adoption of automation technology is definitely not for lack of interest.
Obstacles for Adoption of Automation Technology
In my work with 3PLs, I have identified seven main obstacles to the adoption of warehouse automation at 3PL warehouses:
- Short contract periods with customers
- Lead time for systems engineering
- Implementation and ramp-up time
- Strong competition among 3PLs – and its side-effects
- Vastly differing customer requirements
- Ownership of warehouse buildings
- IT integration of automated solutions
Let’s discuss these obstacles one by one.
Short Contract Period with Customers
3PLs‘ customers, whether retailers or manufacturers, rarely commit to contract terms of more than three years. Often enough, they are willing to commit for only two years. While there is a chance that the business relationship will continue and the contracts will be renewed, this „maybe“ is not enough to base a business plan on.
Automation projects, on the other hand, almost always have a payback period of more than three years, often more than five years, sometimes more than seven years. As a tendency, the higher the degree of automation of the warehouse, the longer it takes for the investment to pay for itself.
It is clear that these two tendencies do not go well together. Of course, 3PLs cannot commit to significant investments that take longer to return than customers are willing to work with them. The economic risk would be unacceptable.
Lead Time for Systems Engineering
Automation projects often have a significant lead time. In the larger1 projects I was involved in, it always took about two years from the first customer contact to the signing of the contract. The engineering and sales phase was considered short if it took less than six months.
The tender phase for 3PLs – that is, the time 3PLs have to respond to an RFP that their potential customer has posted – is often about three weeks. That’s not very long, and 3PLs have standardized their planning and costing of manually operated systems to the point where they can handle such short time frames.
However, the chances of a warehouse automation company being able to plan and quote something in three weeks are very slim. Often enough, it takes two weeks or more to get an NDA signed because of the need to remove all the silly ideas that some lawyer with too much time on his hands had written into the document. During the recent warehouse automation boom between 2017 and 2022, three weeks was often not enough to get a callback from the warehouse automation company because they had a huge backlog of requests.
And it’s not that engineers in warehouse automation companies are slow. There are many excellent people in the industry with great expertise and they work fast and long hours. But it is this very expertise that allows them to recognize that many projects have a whole range of degrees of freedom and the problem needs to be better understood before a solution can be presented. Often, proper data analysis is required before a system can be designed and quoted with confidence. That alone can easily take a week – or more if there are questions about the raw data or specific process requirements that are not apparent from the raw data but need to be considered for the analysis.
Aside from the inherent complexity of developing an automated system, warehouse automation companies have largely done a terrible job of simplifying and standardizing their solutions. Even though things have improved, there is still a lot of manual work involved in sizing and drawing even the simplest and most unremarkable systems.
I have mentioned in other articles how quickly AutoStore people can sketch a system and provide a quote. That’s true – but the storage cube is often not a complete system, it’s just one part of the system. A more comprehensive automation solution that includes more than just picking and decanting requires more effort and more time. That said, the very short lead time for AutoStore-only offerings is probably one of the main reasons AutoStore is the only automated system that has gained a foothold in the 3PL industry.
Long story short: 3PLs have about three weeks to respond to their clients‘ RFPs. But they will not get a quote from an automation company (other than AutoStore integrators) in that time. Of course, then, they cannot consider most automated storage and retrieval systems in the RFP phase.
Implementation and Ramp-up Time
When the contract with customers lasts two or three years, any solution on which the system is based must be implemented and put into operation in the shortest possible time. Most automated systems take more than a year to build (counting from the contract signing). After the contract is signed, the automation company begins the time-consuming process of detailed design, i.e. the system outlined in the sales process is converted into construction-ready drawings which will form the basis for the bill of materials and scheduling. At the same time, the IT people start drawing up the detailed specifications for the IT system. From the point of view of the automation company, it makes a lot of sense to carry out these time-consuming activities after the contract has been signed and not during the sales phase, simply because most sales projects are not won and thus too many man-hours would be wasted.
Once the detailed planning is finished, material can be ordered and production be started – provided there isn’t a huuuge backlog of orders that need to be finished before the new project can be started, which has been exactly the situation for the past couple of years. For the incumbents in the warehouse automation industry – Dematic, SSI Schäfer, Swisslog, TGW… – there is huge variation in the solutions and they have a broad portfolio of technical components, so that for economic reasons they can start production of most parts only upon contract signature. The order penetration point is early in their value chain.
In addition to resources for mechanical parts, resources for IT systems were and are in short supply. The legacy IT systems in the warehouse automation industry are often complex and pretty crappy. For whatever reason, the IT cost of even the simplest automated systems, built hundreds of times before, is creeping up at ten times the rate of inflation. If you look at the details of the IT cost calculations, you’ll find gems like the estimated implementation cost for a bloody label printer or a carton erector at 400 man-hours. Despite the fairly generous estimates for programming and implementation effort, IT is the primary reason projects are completed late or never. Which is no surprise to anyone who has ever worked with IT people.
For 3PLs, however, this is a problem because it lengthens the implementation phase and shortens the economic life of the system in operation. They simply cannot afford to wait more than a year to get the system up and running (and have another six months of ramp-up time) on a three-year contract term.
Competition Among 3PLs
To make things even more complicated, the 3PL industry is quite competitive. Clients send their RFPs to not one, but a handful of 3PL companies. In such a situation, each bidder tries to be competitive (a) on price and (b) on implementation time.
To be competitive on price, they have to be very accurate – and very confident – in their costing. Most of the time, this works out well. But sometimes it does not work out, and they lose money on every order they pick for their customers throughout the contract period. But it becomes an absolute gamble when you consider the price volatility of automated systems today. The steel price index alone can tip a cost calculation, and it is not easily possible to pass the cost on to 3PL customers.
To be competitive on implementation time, every 3PL wants to offer the shortest possible implementation time. For the reasons outlined above, this does not mesh well with most automated systems.
The competitive situation makes it very difficult for any 3PL to break out of the cycle and offer an automated solution that should ultimately lower costs for themselves and the customer.
Variability in Customer Requirements
Many 3PL customers have very different logistical needs. Some have large, bulky products and are simply looking for additional off-site storage to free up some space in their own facilities. Others want you to pick 10.000 orderlines per day from 100.000 different SKUs. Some have flammable or toxic products, some have baby cosmetics, some have food. It’s impossible to anticipate the next customer’s needs.
This is important for several reasons. First, a solution designed specifically to meet the needs of one customer is unlikely to meet the needs of another. Therefore, the solution design must be as generic as possible. But this is easier said than done. Imagine a very simple goods- to-person picking system based on single-level shuttles. Two customers, Customer A and Customer B, offer small products that can be easily stored in standard plastic totes and ergonomically picked at goods-to-person picking stations. There is only one difference: Customer A has an average of 1,2 orderlines per order, while Customer B has an average of 25 orderlines per order. The shuttle system that works well for Customer A is unlikely to perform as expected for Customer B (read this article if you want to understand why).
And since it is time consuming and unlikely that you can take apart a custom automated system and sell it in pieces, but it is very likely that you can either sell your traditional pallet racking or use it in another warehouse, there is a clear tendency to stick with standard manual equipment.
In many cases, the warehouses operated by 3PLs are not owned by them, but leased from commercial real estate developers. And that makes sense, given the often short contract terms combined with the high variability of customer requirements. You can not know in advance how much space you’ll need or which location is best for your next customer. And, of course, you are not going to install automation technology in a building you are only renting for a short term, just as you would not do an expensive hardware tune-up on a car you are leasing for three years. At least that’s true for technology that is virtually immovable, such as high-bay racking, stacker cranes and conveyors. It’s a different story for AGVs, and it is one of the reasons AGVs are attractive to 3PLs.
Even in cases where buildings are owned rather than leased, 3PLs often prefer not to permanently modify them with automation technology, or to increase the building height such that they only make sense for automation. There is a well-established market for second (or third, or fourth…) hand warehouse buildings. The closer a building stays to standard specifications in terms of building height, floor slab, power rating, etc., the easier it is to sell.
Earlier, I briefly and cautiously suggested that IT… well, can be a problem. Sometimes. IT can be a problem for warehouse automation vendors and their execution time, because the effort to customize a control system for the automated system or even a complete warehouse management system (WMS) for the customer is high, and so is the risk of screwing up.
As for 3PLs, they usually already have a WMS and a fairly large IT team to take care of it. And like every company on the planet, the IT team is already working at capacity and barely able to keep up with all the things the regular folks want them to do. Any bit of automation technology that requires integration into the WMS is going to face some resistance because the team at IT is already drowning in work. So if you want them to work with the automation vendor’s team IT to develop requirements specifications and program interfaces… that’s another barrier to using automation technology.
What 3PLs can do About it
As it becomes clear from the text above, there is only so much 3PLs can do about the situation. Some of the most important reasons are external and thus lie outside their sphere of control.
However, there are a couple of things they can and should work on.
Building IT Competence and Resources
It seems to go largely unnoticed that the most successful logistics companies are, in fact, IT companies. And yet, if you talk to 3PLs, it seems like they consider IT something that’s a necessary evil rather than something that is a very central strategic topic. It is something a designated department does and the rest of the company doesn’t want to have anything to do with it. While I certainly have sympathy for emotional resistance against an IT focus, I do believe this is problematic and will haunt them.
The reality in most warehouses is that every regular process needs to be foreseen in the WMS. But what if it turns out the process is inefficient and should be different or improved? What if you want to run an efficient continuous improvement process which (obviously!) needs to include IT? What if you want to quickly integrate a new technology and test out its productivity benefits? Believe it or not, but some of the largest 3PLs in the world (!) are unable to tell in a multi-client warehouse which clients make them money and which do not. That’s an IT problem, because tracking productivity individually per client is not foreseen in their WMS.
Also, 3PLs sit on a treasure of data, yet they aren’t able to harvest it. They spend a fortune on motion tracking systems or digital twins when they could get the very same information these systems deliver from some simple queries on the WMS data. Do you want to know how often employees need to pick heavy items from unergonomic positions? Look into your WMS data, it’s all there. You don’t need motion tracking systems to find out. I will be happy to help you if you cannot figure it out by yourself.
Developing a Customer Acquisition Strategy
Often 3PLs have some sort of industry focus. Some specialize in industrial customers and spare parts, others in retail. That’s a good start, but it’s not enough.
One sales approach is to respond to RFPs and offer customized solutions to customers. That is what is being done in most cases today. It is a very different approach to develop a solution that fits many different customer requirements and then approach customers that fit in. The latter approach not only lends itself to multi-client warehouses (see below), but also potentially enables lower customer acquisition costs, lower IT integration costs, and lower operating costs due to the standardized nature of such solutions. There are a number of 3PL startups, such as 360 Logistics from Norway, that are taking this approach, and I believe it makes sense. Look for customers that fit your solutions, rather than tailoring your solutions to each individual customer.
Building Multi-Client Warehouses
The corollary to the previous point is that multi-client warehouses should be emphasized. If you are building solutions that are potentially suitable for different clients, you may as well bundle clients in the same warehouse. With a multi-client warehouse, you do not have to rely on a single customer for the payback period of your automation equipment. It also hedges the risk of a single customer not producing the throughput (and therefore the revenue).
What 3PL Customers can do About it
Customers are a curious breed. They complain about 3PLs not being innovative enough, being too expensive, and so on, all the while making it very difficult for 3PLs to deploy automation technology that would be “innovative” as well as lower cost for them. As long as 3PL capacity is treated like a commodity, with different 3PLs offering perfectly substitutable fulfillment capacity, the adoption of automation technology will increase only slowly.
In order to enable 3PLs to take the risk of high CAPEX investments in automation technology, customers either (a) need to commit to longer contract periods (not the popular two-year and “then perhaps another two-year” model) or (b) they need to simplify their requirements so they fit into more standardized multi-client sites. Preferably both. It will make a big difference to their logistics cost.
What Automation Companies can do About it
Standard Solutions and Productification
Many warehouse automation companies have completely missed the market need for standardized solutions. The “every customer is different” mindset has led to hundreds of thousands of man-hours wasted on the whims of consultants and customers who think their requirements are special when in reality they are not. Every custom-made solution requires software customization, making the entire solution even more expensive. The result is that warehouse automation for most warehouse automation vendors is a low-margin business, which is almost funny. If you look into the annual report of the KION Group, you will find that Linde Material Handling makes more profit selling forklift trucks than most warehouse automation companies. Linde is a fantastic company, but the reason for the difference in profitability is made by many of the warehouse automation companies not being so fantastic and operating basically without any strategy.
The advent of Micro Fulfillment Centers (MFCs), which predictably turned out to be mostly useless, had one positive aspect: more warehouse automation companies were nudged to think about more standardized solutions. Because most customers don’t need to pick 10.000 orderlines per day and require highly specialized solutions.
AutoStore have demonstrated how to do it. They have effectively and successfully “productified” warehouse automation. The process of planning, quoting and installing a system has been streamlined to an extent that cost has come down significantly, both for themselves as well as for their customers. And before you say “Well, AutoStore is simple and cannot be compared to our broad portfolio of technology and customized solutions”: look at Witron, they have standardized the highly complex process of automated case picking to an extent that allows them to present a complete solution with a budget price in the second meeting – while all their competitors still have questions about the planning data.
Help IT Get Their Act Together
Earlier in the text, I pointed out that 3PLs need to understand the importance of IT to their future success. It is not helpful to view IT as merely a necessary evil. The way IT is integrated into the company’s strategy will most likely be a competitive differentiator not too far in the future.
The same is true for warehouse automation companies. One of the many funny things about the warehouse automation industry is that every time an engineer or sales rep changes companies, they tell you how terrible the IT was at their previous company and that it was impossible to sell anything because of it. When IT ’s estimate for a simple zone picking system is €500k or IT estimates the need for 150k man-hours for a case picking system (convert that into human lives), you know the company is doomed and you’d better run. IT in the warehouse automation industry is an absolute disaster and there’s no point in sugarcoating it. Most companies in the warehouse automation industry would probably be better off throwing away what they have today, hiring some of the employees from the recent layoffs at Google, Meta & Co. and starting from scratch.
The fact that several warehouse companies have allowed IT to degenerate to an obstacle to business success speaks to (a) the quality of management (or lack thereof) and (b) the lack of understanding of the role that IT will play as a strategic advantage for warehouse logistics (related to point (a)). Warehouse automation companies desperately need some experienced IT executives who happen to know something about logistics. Right now, they have executives who know something about logistics (if they are lucky) and can use Outlook to write emails. That’s not enough. Once the first warehouse automation company develops really good software, their competitors can pack up.
If warehouse automation companies want to work with 3PLs, which generally seems like a good idea, they need to simplify IT customization and integration. And IT needs to become part of the business strategy, just as standardization needs to become part of the product strategy.
I mentioned earlier that AGVs are attractive to 3PLs because they are movable. If a warehouse location closes, they do not become useless, but can be moved to another warehouse location and put into service there. Now, unfortunately, this is an almost hypothetical argument. The state of the art of many incumbent AGV vendors‘ software is such that extensive modifications would be required if the vehicles were to be deployed at a new location. Or even if anything changes at the same site. In an AGV project at a 3PL I supported some time ago, we interviewed shortlisted vendors as part of the selection process. One question I asked the vendors was how much effort and cost it would take to move a pallet handover location, say, five meters to the right. The (established and well-known) AGV vendor’s answer was (and please trust me that I am not exaggerating): „I don’t know. We need to send someone on site to assess the situation, and then the IT department will have to make the changes in the software. That may take a couple of days“. That’s the state of software in the warehouse automation industry. Fortunately, one of the startups that participated in the RFP (and eventually won the bid) showed that the same problem could be solved in their software within ten seconds using drag and drop.
In this article, I have highlighted the barriers to adopting warehouse automation technology at 3PLs. Some of these barriers are homegrown, while others can be found among both 3PL customers and warehouse automation vendors. Fig. 1 summarized the obstacles outlined above. I concluded the article with some specific recommendations on how stakeholders can help break down the barriers and enable 3PLs to better leverage automation technology than they do today. With a growing warehousing market and an increasing shortage of labor for warehouse work, 3PLs need to take advantage of the tremendous potential for productivity gains that automation offers.
1 Larger: above €50M investment volume for logistics hardware and software