You have your logistics operation running smoothly. Delivery to customers is as promised and distribution costs are pretty predictable. The question is: are they optimal?

### The Sooner the Better

There are two factors that have a big impact on transportation cost. One is the number of loading meters that is ordered. The number of meters is simply calculated based on the number of pallets that need to be shipped.

The other is timing of that order. The earlier you order transportation capacity, the cheaper it is. And it is between these two factors that the tension arises. You need to order early, but often you don’t know the exact number of loading meters needed before the last item is picked and palletized.

### I Guess…

The way this is typically solved is by doing an estimation of the needed transportation capacity based on the ordered volumes. For full standard pallets, this process works reasonably well. If there is a mix of products or pallet sizes or even mixed pallets, this gets more complicated. The IT department is asked to develop a tool to estimate loading meters. The total volume of all products are added up and divided by the volume that fits on a pallet. Now you have a rough indication.

### …We Will Add a Safety Margin

As products are picked and palletized it is found that more pallets were used than initially estimated. Now the pallets don’t fit in the truck, and additional ad hoc and expensive capacity needs to be arranged to get everything to the customer in time. To make sure this doesn’t happen often IT now adds a fudge factor, so there is a safety margin. Everybody happy!

### Money on the Table

While this solves part of the problem, it still leaves a lot of money on the table. When investigating these processes we typically find that the KPIs created for filling pallets and trucks are overestimated by 5 to 10 percent. That means that for every 10-20 trucks ordered, you pay for a truck you don’t need!

There is a simple solution to this problem. Well, simple, there are complex algorithms at work… But for you the solution is simple. 3D load building software can calculate the exact capacity needed. Many variables come into play, and they need to be known and correct. To name a few: product dimensions, weight, stackability constraints, compatibility constraints, picking strategies, loading strategies, and customer requirements (e.g. pallet types allowed, maximum pallet height).

This is where data-driven decision making can be used to improve your efficiency. As described above, this will lower your transportation spend. But you are also only looking at physical constraints. When Big Data and Analytics are properly used there are many more ways to improve processes and save costs. A data-approach will allow you to get answers to questions like: do we have the right arrangements with our customers? Are you we using the right pallets? Can the carton dimensions be optimized in a clever way?

### Ok. But where do I start?

Groundwork – Make sure you have good data

Garbage in is garbage out, so data quality is key. Analyzing and optimizing is done based on data. This extends to your suppliers and your customers as well. They need to provide correct data and in full. Agree on standards, and make sure they are adhered to. For example: Agree on how length, width, and height are measured and reported. One person’s width is another person’s height. Data quality is the foundation. It’s a prerequisite. Actually, it’s a continuous improvement process in itself. Data is never perfect, and will always need improving.

Step 1 – Measure: Choose suitable KPI’s

Define KPI’s that help you find outliers. Use the outliers to find improvements in your processes. For example: when you ship out cartons of various sizes on pallets you want to know if these cartons are loaded on pallets in the most efficient way. The maximum number of cartons on a minimum number of pallets.

At one of my customers, a manufacturer of electronics, we use a KPI we have dubbed the Liquified Fill Rate. Add the volume of each box on a pallet and divide by the surface of that pallet. This gives you the average height of the pallet, which can be used as an indicator of how efficiently the pallet has been loaded.

Step 2 – Analyze: Find the outliers

The next step is analyzing the collected data. Do a weekly root cause analyses on the outliers. Look for unexpected results, both positive and negative. The positive results can be used as best practices, and the negative results indicate where improvements are needed.

Step 3 – Improve: Solve issues and implement improvements.

When you have found potential improvements, you need to work out how issues will be resolved, and what improvements are needed. These can be procedural improvements, like updating work instructions, but it can also be (re-)training staff. Other possibilities are for example the use of a different size pallet, or changes in packaging to make cartons stackable.

Once improvements are implemented you go back to Step 1.

### Money in the Bank

By following (and repeating) these 3 steps there will be less money left on the table, and more money in the bank! 3D Load Building software will not only allow you to analyze larger amounts of data, and complex problems. It will also help you to collect and combine the load building knowledge and experience of the people in your organization. This makes it easier to find and roll out improvements, and also prevent the loss of valuable information if people change jobs and leave your company.

Suggested reading for those that want to save even more: 13 Ways to Save Money With 3D Load Building