Let me start this post with a teaser: You have 20 light bulbs in your house and they are all the same kind, made by the same manufacturer. The bulbs are used for eight hours a day. The light bulb packaging indicates that each bulb has a burning life of 1000 hours and the fine print states that the burning life is 70 percent accurate. How many spare light bulbs should you keep for replacement of blown bulbs?
Assuming you go to local hardware store only once a year to buy the spare bulbs, and you are happy to have a 60 percent chance of spare bulbs available when needed. Now you find out there is another manufacturer who charges $2.00 extra, but the fine print on the box states 1000 burning hours with 97 percent accuracy rather than 70 percent. What happens to the calculation? Situation: Mom-in-law is visiting and now you want to have 95 percent chance of a spare bulb available rather than 60 percent. What happens to the calculation? Can you minimize your total spending by opting for higher priced manufacturer bulb and still have replacement 95 percent of the time? Lots of math :).
If you got this teaser you understand the very basic math behind the service parts planning. In the case of service parts, supply chain planning is vastly different from planning for manufacturing. I will try to put the key differences in the three functional areas:
Master Data, Demand Management, and Supply Management
A good system should be able to maintain key data elements and run analytics based on them. Some of the data elements important for service planning are:
- Service BOM Data: A manufacturing BOM could have 100’s of component and several levels, but a service BOM is much simpler. Think of copier machine; if the roller fails to pick up paper, the service part that is shipped to the end user is the entire roller assembly, which the end user can pull and replace. So the service BOM will typically stop at that level.
- Alternate Service Parts Data: This element is very interesting and if an organization is able to manage it well, it can reap huge rewards on the inventory metrics. It is fairly complex when compared to the typical alternate parts management in the manufacturing. Think of a failed hard drive in the end user’s laptop with specifications as 5400rpm/60GB. The service provider can ship equivalent or better replacement. Specifications permitting, the end user will gladly accept a replacement of 7200rpm/80GB, and it may also be easier and cheaper for the service provider to do so if the 5400rpm/60Gb is obsolete and hard to procure. This kind of alternate replacement typically does not happen in the manufacturing planning.
- Sourcing Data: Sourcing data needs to be maintained for repair partners apart from new buy partners. Repair lead times and repair yields should be maintained and considered in analytics.
- Multi-Echelon Data: Service organization set ups are more multi-echelon as compared to manufacturing. Customers use products everywhere in the world and may have service contracts of next day service or as quick as onsite four hours service. The data on the service level identified in these contracts should be available for analytics.
- Logistics Partners/Service Contractors Data: Several service organizations typically work very closely with logistics partners like Fedex/DHL to stock and deliver service parts. There may be field contractors/agents (the guy who came to fix my washing machine, which was still under warranty, when it broke down) who are tasked to fix the unit at the end user. The system should maintain information on these service partners.
Stay tuned for part two on Monday where I’ll be discussing Demand Management and Supply Management.