Optimize the digital order pipeline with Supply Chain analytics
From data via insights to actions
Organizations often struggle to guarantee an agile, flexible and transparent supply chain that is capable of coping with a continuously shifting playing field. An increase in complexity comes along with an increased amount of available data. Using this data to maximize supply chain efficiency is not straightforward – however necessary to make informed decisions. It requires selecting the relevant data, assuring a steady flow of accurate data, and translating the data into useful interpretations. In this article, we explain a case in which Deloitte supported a world famous designer of seasonal consumer goods to improve their digital order pipeline process by advanced analytics.
Over the last years, the development in online sales has exploded and this particular sales channel has become increasingly important for retail businesses. As we enter into a new era of online purchases, the competitive edge lies in the efficiency of the overall service, which puts an emphasis on transactional ease and short delivery times. The type of supply chain is described on the right hand side of Figure 1. Although many customers might perceive delivery time as the physical movement of the purchase to their doorstep, the total delivery time includes all internal processes from the moment a purchase is made, and therefore just as important to the customer experience and satisfaction as the physical delivery itself.
In order to guarantee good service levels, it is necessary that all steps of the online process are transparent and can be measured. As such, the client asked Deloitte to analyze their online customer order process with the goal to increase the reliability and speed. In the online customer process, from the moment of the check out on the website to the order being dropped to the distribution center (for picking), the order goes through at least ten different steps. This process is called the digital order pipeline - and each change between stages in this process is logged in the system. This process log, containing more than 100.000 orders, served as a valuable starting point to analyze the digital order pipeline and assess where and how delivery reliability and speed could be improved.
Of the total amount of 100.000 orders, roughly 70% were completed within the allowed time. This group will be referred to as the good group. Another 22% were completed at roughly twice the maximum amount of time, 7% within 6 times the maximum and 1% which were completed in an even longer time. The decision was made to focus the analysis on the 22%, as the group represented the largest opportunity to identify structural and recurring delay drivers. In order to identify the structural drivers in more than 20.000 orders, clustering techniques were applied.
Figure 2 shows a simplified version of one of the resulting clusters. The x-axis shows the different steps in the online order process, while the y-axis shows the deviation from the average time it took an order to complete this step in the good group. In this particular cluster, it is clear that the hold period step is the source of the delay for this group of orders. The graph indicates that in some cases the time spent in the hold period was over 200 minutes longer than the time spent by orders in the good group. The other steps of the total order process take a similar amount of time as used by orders in the good group. Understanding that the delay occurs in the hold period allows the client to apply standard lean six sigma techniques to reduce process variability.
By using process analytics, the root causes of delay in the digital order pipeline could be identified within a massive quantity of data in an intelligent, structured, and efficient manner. In this particular case, the visibility provided by the advanced analytics led to more than 20 structured improvements in the 10 step order process (e.g. increasing frequency of batch-runs for some of the steps or inventory allocation rules) which resulted in increased delivery reliability and an average processing time reduction of 50% per order.
The example presented in this article shows the power of applying advanced analytics in a supply chain. In this situation, the available data contained the information needed to improve the different processes and achieve the supply chain goals, but the sheer volume of data and number of variables were simply too large to analyze. The increasing complexity of the supply chain resulted in a direct need for solutions that could support fact-based decision making. Deloitte supported the client with both the initial supply chain analysis as well as the development of several decision support tools. This helped the client to propose shorter delivery times to the end consumer and meet these promised delivery times, making this a very successful example of how the application of advanced analytics can be leveraged as a key differentiator in supply chain improvement initiatives.