How a bike manufacturer is using Process Mining

Sam Werner

A Mountain Bike manufacturer successfully sells large quantities of customized mountain bikes to retailers. After receiving an order from a customer, the manufacturing department typically requests the purchase of materials such asseat tubes, fork crowns and handle bars to build the bikes. As customers often order on short notice, they need to maintain high standards in production planning. If required materials are not available at specified times, the whole production is delayed, customers wait and become unhappy.

Q: Departments constantly complain that purchase request are not fulfilled in time. Why is this? A: I think itโ€™s not the production department alone but other departments often give approvals far too late. Also my feeling is we need a lot of manual intervention in the process that slows us down. But we donโ€™t have any figures on that.


As I said, we had some assumptions of why things go wrong, but we could not support them with facts and numbers. In order to do this, we used a tool to visualize all purchases end-to-end. This gave us a systematic overview of where time was lost in the process and allowed us to drill down into relevant cases to find out why.

The good news is that most of our suppliers deliver on time. The bad news is that in many cases the purchase orders were indeed sent too late due to late approvals by the manufacturing departments. On average, approvals took 7 days, whereas the purchasing department needs them much sooner. We showed these numbers to the departments and, as a result, the approval process has been given a higher priority. We now monitor approval times as well as supplier performance to make sure that goods will be ordered and delivered on time.

p2p_cpm4Besides, we also found out that time was lost within the purchasing department itself because a lot of rework activities were delaying the process. Among these were mainly wrong master data entries and manual price changes, which required employees to put orders on hold and invest extra time. By looking at the rework cases systematically, we quickly found out that in the first quarter many suppliers had changed and others had not sent their current price lists. To improve the situation, we updated our master data records and price lists and reduced the average processing times by another 15%.

Additional Findings:

By looking at critical delays, we were surprised to discover that for 6 of our major suppliers, the automation rate was very poor. A root cause analysis revealed that in 12,300 cases, the goods receipt could not be scanned and processed automatically and had to be entered manually. We already talked to the suppliers and agreed to install a new barcoding system, which allows electronic processing.

Join one of our webinars to learn how process mining can help your business.

Or sign up for our trial cloud to try it for yourself.

Back to the blog