Victoria Herzog, a student at Kühne Logistics University in Hamburg, took on the topic of the optimal size of mail order packaging for her master’s thesis.* She demonstrates how the average fill level can be boosted by over 20 percent. In collaboration with Tchibo GmbH, she applied the data from 200,000 orders to a model for calculating the package size that Tchibo should use for its online shipping in order to improve the overall fill level. Her results could be significant for the entire industry. For her master’s thesis, Herzog received the Best Thesis Award 2014 from KLU.
The company wanted to optimize the utilization of packaging contents because a higher fill level results in a number of benefits. The smaller the package, the less material the box requires. The logistics costs also drop and since the total volume to be transported is reduced, the company’s carbon footprint becomes smaller. Another benefit: fuller packages are less likely to be damaged by other, heavier packages. “Considering the multitude of packages we ship every day, an optimized package size is a key factor. Not only for economic reasons, but from an ecological viewpoint as well,” said Marc-Stephan Heinsen, director of supply chain management & logistics at Tchibo.
It isn’t easy to solve the problem of determining the correct packaging size, according to Herzog: “Unfortunately, you can’t adjust the size of the package to each individual order. Tchibo’s product range is too diverse and it would take too much time. Booksellers have an easier time of it.” Imagine you receive an order for a large cutting board and a pound of coffee… “We must use rectangular packaging, and regardless of how you pack the products, the package is always only half full in this case. Separating the order into two individual packages is not an option, since the customer would have to bear the increased costs,” said Herzog.
Currently, Tchibo uses over 20 different box sizes. “It isn’t possible to use more different-sized boxes since they would all have to be at each workstation in the Shipping department and the employees would have to put too much extra effort into their work,” explained the KLU graduate. The problem is finding the right boxes to have at hand. To determine the ideal selection, she programmed different algorithms to test a range of different packing options. “For an average order consisting of six products, there are many ways to put the items into a box,” said Herzog. “You have to test the options to arrive at the box with the smallest volume.” Another important step is inputting the correct boxes into the system. If the optional boxes are all too small, the products will not fit into them and if they are too large, too much unused space is the result – with the familiar consequences.
To determine the box sizes in the optimal selection, Herzog relied on the data from over 200,000 anonymized orders from 2013. “It takes all night for the computer to finish the calculation – you can’t afford to get anything wrong!” she said. Her calculations led the young academic to identify a new range of box sizes that improved the average fill level by over 20 percent.
Prof. Kai Hoberg, the Kühne Logistics University professor for supply chain strategy who supervised her master’s thesis, takes the scenario further: “You could also imagine supplying packaging for stationary retailing that is different from the packaging for the online shop. For online trade, packaging properties such as customer appeal and anti-theft features are not top priority – instead, smaller, simpler packaging is the key here.”
Academia and online resellers hand in hand. Everyone benefits from collaborations like this.
*Title of the master’s thesis: “Filling degree optimization of shipping packages in the B2C-market at Tchibo GmbH. An application of supply chain analytics”