This report will show a list of parts that are failing before they should. We show the most failures at the top and list the models. This will help identify items and/or models having extraordinary amounts of warranty alerts. For this report we used a term called shrinkage. It is defined by the loss of potential dollars of use because an item failed to fulfill its mfg yield amount. So, if a $100 drum was to go 100,000 copies and only ran 80,000 copies, the shrinkage amount was $-20.00. Then the report totals the shrinkage by item and stacks ranks from the greatest loss down from the top.
What to do with the information (thank you Jon Balter!):
Where to Start: Start with the top 3 times. These are your biggest savings.
So we are going to look at 40X641
Step one is to run the Item Transaction History (eAuto Report listed under Reports/Inventory).
- This will give you a good idea where the parts went and a quick overview.
- I like to run it for 6 months. The report goes back 6 quarters but I think 18 months is too much data to work with.
Now we want to look at the service calls. Again, I want to only look at a 6-month sample. I then print all the service calls. You can get all the calls by going to Service invoices and using this in quick search.
Then right click and print each one. I know this takes time but you need to dig deep to find a cause.
Items to look for:
- Are you billing this item? If so, stop looking. You are making money on it. You may use this to give to your sales person to upgrade the account to a better machine that makes yield on supplies or parts.
- Look for parts being replaced together. Technicians will often replace parts together. It becomes a habit. It may be unneeded. For example, I replace the drum unit with the fuser unit on small machine but they have different cycles. Drum is 60K and fuser 120K.
- Note: For my part I haven’t found anything abnormal yet.
Another report to run is the CEOJuice report: Item Yield By Make Report (ID136, ID90).
- I like to run this for a year.
- Here are the settings:
- You can exclude based on min yield or # of months but for this I want to see everything.
So what did we find? The manufacturer overstated the use of this part. The manufacturer put the part usage at 120,000 but that is not real world.
Actions I could take:
- Reduce the target yield to 60K and stop getting alerts
- Notify my parts and technicians to return 100% of these parts for warranty
- If my manufacture does not have a warranty program, I collect the parts with all the data and wait until I have 10 parts and request a special exception. I have all the technical data they will request in advance like service reports from machines and counter sheets.
- Notify the manufacture and have them lower the price for the part equal to the yield difference.
The result of these types of research is to get to a resolution for each part that is not making yield. Sometimes there is nothing that can be done. You will not win every one but those that you do win may save your company huge amounts of money.
Alert ID90 vs. ID405:
Alert ID90 lets you know when an item does not make yield. It will send an email with all the information needed for the return. It also lets the technician know when the part did not make yield. The one issue with this is that it is after the call is closed so the technician needs to hold on to the parts until the next call.
I believe ID90 should be used in conjunction with ID405. ID405 will give you the items that you as a manager need to research and ID90 will get you all the information to get them returned.
Setup both ID90 and ID405 together. Return all parts that don’t make yield and work on warranties. The issue is that most manufacturers, like Ricoh and Toshiba, have removed all warranties for a savings up front (BIG AIR QUOTES). ID405 will help you figure out what parts you want to push the issue with and which you are going to have to just take a loss on. You are not going to get Ricoh to give you an exception on every part, but if the cost is high enough and you can make a valid argument with real data, then you can at least win the big battles.