May 13, 2020

Especially in the current crisis, it can be enormously important for companies to recognize supply chain bottlenecks early on in order to be able to act directly. The coronavirus crisis is putting the economy to the test. Increasingly, supply chain bottlenecks are occurring because suppliers may fail, and the prices for replacement deliveries are rising sharply. The question of how to maintain an overview and create transparency in procurement during uncertain supply situations is therefore of concern to many companies.

In addition, the demands on data analytics are increasing, evaluations on specific crisis topics are needed very quickly and there is no time for days of searching for the relevant key figures in different data sources. In the following blog, I would like to use practical examples to show how data analytics can be used to best master uncertainties in procurement and supply chain bottlenecks.

Avoiding supply chain bottlenecks in the current crisis – an example

First of all, it is paramount that there is transparency across departments about all relevant data in manufacturing companies. For example, it is important that it is clear in the departments involved which customer orders have to be prioritized and which ones have to be postponed if necessary, or whether there have been important changes in shifts or employee absences. With data analytics solutions, such information can be made available throughout the company so that the departments concerned can take action and initiate appropriate countermeasures.

If such data is then additionally linked – as is currently relevant in many companies – with freely available data on the spread of Covid-19 or certain municipalities of the disease, the effects on one’s own point of sales or delivery channels can be causally represented by combining the data. In a suitable data analytics solution, it is then possible to see which supplier of a company or a specific point of sales is located within a municipality and whose deliveries may be at risk. In addition, it is then possible, among other things, to track exactly which products you receive from these suppliers. As a result, it is possible to identify which of your own suppliers provides the same or similar products to the suppliers from the Covid-19 municipalities and could supply replacement products in the event of a failure.

With a corresponding write-back function integrated into the data analytics solution, it is even possible to modify the data directly in the interface and automatically forward it to the responsible department – in the case described above, the purchasing department. Internal communication is thus considerably simplified and countermeasures to supply chain bottlenecks can be proactively initiated.

Another example improved by the use of data analytics is the supplier evaluation. Using certain criteria such as capacities, price or delivery period, suppliers can already be quickly evaluated by data analytics systems. In times of crisis, the weighting of these criteria can be adjusted. The delivery period, for example, can become more important than the price and the weighting of the criteria can be changed accordingly. In this way, you can clearly evaluate the vendors according to the current situation. This makes it possible to recognize early on which supplier could create difficulties for procurement and where bottlenecks could occur.

Data science and replenishment times

In reality, moreover, estimates of suppliers’ replenishment lead times differ greatly from the real values. This means that the actual delivery duration varies greatly from the delivery duration stated on the master data. Here, it would be possible to make manual corrections, but it would be a process that would be very time-consuming and costly. By using machine learning, the replenishment lead time can be predicted more accurately using historical data, because the algorithm “learns” based on past deliveries. The accuracy of these forecasts has already been improved by between 15 and 42 percent at the companies where machine learning is used.

The coronavirus creates new challenges for companies in the area of procurement, as deliveries become uncertain. With the help of data analytics, it is possible to create transparency across different departments. Suppliers and logistics branches that are directly threatened by failures can be quickly identified by merging various data and, accordingly, a quick replacement can be found. Bottlenecks can thus be prevented in advance and production downtime can be avoided.

Furthermore, previous areas of analytics such as supplier evaluation can be adapted to the current situation. Data science also offers further solutions to make the analysis and planning of procurement more precise and better in the long term.

Which difficulties currently determine your supply chain? We support you in every data-driven challenge.


Jens Siebertz

Jens Siebertz ist Senior Vice President bei INFORM DataLab.