Day 141: My KPI is Your KPI (You Just Didn't Know It)

It's long been said that you can't manage what you don't measure. I'd say there is a lot of truth to that statement, both for managing current state, and for monitoring your progress towards a future state. If you aren't tracking your progress towards some goal, you are not likely to get there any time soon.

Organizations use KPIs, at different levels, to align employees, suppliers, processes. The right KPIs can yield great results.

But if you have been around a while, as I have, you probably have seen the scourge that is Death by KPI... too many KPIs for employees to understand what is important, to help them make good decisions, to help them identify when a process is not in control... The noise can be overwhelming. Equally dangerous are KPIs that result in unintended consequences or behaviours. I call this My KPI is your KPI (you just didn't know it)...

This past week I had a situation where my department was affecting a KPI for another department. We run two production shifts here, and my staff work on the first shift only. One of our tasks is to audit production, and sign off job travelers as we go. Many jobs cannot close until we sign off. So, in some instances, we sign a job off the morning after the work is completed by the afternoon shift. The job is then closed in our ERP system that morning. 

If the product is intended to ship to a customer immediately, we tackle those audits first, to make sure they can meet any shipping cut-off times. The rest, which may be for upstream production jobs or for inventory, are completed afterwards. Our department has not been responsible for any delay in shipments to our customers by prioritizing our work in this manner. 

Recently, the Production group stated to me that my team was causing "late jobs", that is jobs were closing after the planned job end date in our ERP system. As there had been no reported delays of shipments due to my team, I was surprised by this statement, and so decided to investigate further. It turned out Production was being measured on total job completion, regardless if the next operation was an internal one or a customer shipment. So a job that closed, say, less than 12 hours after its planned end date was showing up as late, and Production's "jobs closed on time" KPI was showing less than 100%. Even if it wasn't shipping to a customer for days.

Does this matter? Is it important that less than 100% of jobs are closing on time if the product is not intended to be shipped immediately? 

I struggle with this because my top priority is to ensure that no shipment to a customer is late. One of the top level corporate KPIs here is On-Time Delivery (OTD) - every one in the building knows what it is and how important it is. Our customers do not see value in the internal processes, the internal jobs and their planned movements or transactions within our factory walls. Well, it turns out that many of our planned job end dates leave plenty of room for picking, assembly, test and inspection - typical jobs have 5-7 days of time allocated whereas the actual time required is far less - there is a lot of buffer in these schedules (that is a story for another day).

So I come back to my original question: does this internal KPI provide value to anyone outside of the affected department? 

And if the answer is "Yes", what action should be taken: should they adjust their measurement, should they finish the work a day earlier, or should I adjust my staff's working hours?

The answers, I think, are "No" and "None of the Above".

Here's why I think this: We plan our production jobs on a fixed lead time of 7 days. Lead time is made up of 5 key steps: Queue, Wait, Set-up Time, Run-Time, Move. Our Production lead time includes all 5 steps. Child jobs that feed parent jobs all have the same 7 days lead time, we do not pull out the Queue or Wait times in between these jobs. Add to this that our ERP system plans Start-to-End, in a Push method, and our Production team Pulls work based on End date, no wonder there is this level of conflict. Production has the freedom to pull jobs to start, based on the end date, not on the start date. So a job scheduled to start Monday and finish Friday might not actually start until Thursday. My team has no say in when job actually start in Production, yet Planning has allocated plenty of time to complete and close jobs.

So until we address the lead time for our production jobs, to increase their accuracy for the true Set-up and Run times, then what does an artificial ~12hr difference really make? None - in the eyes of the Customer. Some - in the eyes of the internal customer, particularly if some internal scorecards are affected.... But our schedules are so padded for Queue, Wait and Move, that we should have bucketloads of time to get jobs completed and closed on time.

Another reason I think no action should be taken is that this delay is one of the smallest contributors to late job closing.

Take Production data for May (to date):

- 1577 Jobs of which;

- 1503 Jobs closed on time or early therefore;

- 64 Jobs were delayed in closing (4.06%) of which;

- 21 Jobs were delayed due to any and all production constraints including delayed sign off and job closing (1.33%).

I'll be generous and attribute halfish of the 21 delays to my team, as there are other factors that delay job closing. So 10 jobs delayed by inspecting the next morning makes for 0.63% of all jobs. So for less than 1% difference in job completion, I should add another whole person? No. I want to understand what other production capacity constraints truly exist - like labour shortage, equipment downtime, lack of parts, vendor quality, and so forth. Is a goal of 100% a good thing? Yes! But I haven't seen enough information to suggest that my team's delayed inspection and sign-off is the biggest challenge in achieving that goal.

I guess I'd suggesting that KPIs specific to a group, department or process are useful within that group, department or process if it provides value to those who use it. When KPIs start to creep across processes, departments, groups, it is probably worth engaging the upstream and downstream customers to make sure we aren't driving unintended activity or consequences - most of which are of little to no value to our Customers. The paying kind.