Laboratory KPIs

[bpillar]

Dears,
we are revising the KPIs of our testing laboratory that makes tests to qualify products destinated to OEMs, due to the fact the actual ones are not so significant and effective to monitor and evaluate the performances.
Can someone suggest, some process indicators to monitor the performance in terms of effectiveness and efficiency, and how calculate them?

Thank you in advance!

 

[Steve Prevette]

I do have a lot of metrics (KPI) materials stored here on the Cove. See Index of /Cove_Members/Steve's stuff (elsmar.com)

I would start with
https://elsmar.com/Cove_Members/Steve's stuff/Choosing_performance_indicators_Day_One.pdf
https://elsmar.com/Cove_Members/Steve's stuff/Generating_and_Using_Control_Charts.pdf
https://elsmar.com/Cove_Members/Steve's stuff/Implementing_Performance_Measures.pdf

 

[bpillar]

Hi Steve,
Thanks a lot for your reply. The material you proposed it's interesting, but we were looking to speific examples of KPIs applied to a testing laboratory. We have some ideas, but we are not convinced that they are so effective as happened in the past. So if it's possible to have some suggestion we will appreciate.
Thank you

 

[Mike S.]

Ask you internal and/or external customers what is important to them.

Once when I ran a testing lab, quick TAT was important to my internal customers, so I had a KPI of % of samples completed within 24 hours of receipt.

 

[Tallinec]

Dears,
we are revising the KPIs of our testing laboratory that makes tests to qualify products destinated to OEMs, due to the fact the actual ones are not so significant and effective to monitor and evaluate the performances.
Can someone suggest, some process indicators to monitor the performance in terms of effectiveness and efficiency, and how calculate them?

Thank you in advance!

If your lab is not a non-profit organization the important KPIs are:

1)profit in areas of lab activities.
2)% of last year customers which are using your lab services in this year.
3)number of foreign customers which are using your lab services in this year.
4)number of private labs which are using your lab as a subcontractor in this year.
5)number of govermental organization which are using your lab as a subcontractor in this year.
6)number of cases when your lab (or divisions of your lab) works had to be reworked by stuff of your lab.

№ 6 is very difficult to register and monitor because the stuff tend to rework their bad done works quietly and not to report about their small and medium mistakes. Laboratory top management shall maintain very positive work environment to motivate stuff to report their reworks. So, № 6 is a fantasy.

 

[Tallinec]

Ask you internal and/or external customers what is important to them.

Once when I ran a testing lab, quick TAT was important to my internal customers, so I had a KPI of % of samples completed within 24 hours of receipt.

But if a sample came in Friday or in one day before public holidays? Or when in testing it was found that the sample has a hidden defect?

 

[Mike S.]

But if a sample came in Friday or in one day before public holidays? Or when in testing it was found that the sample has a hidden defect?

We had rules to cover these kinds of situations. The intent of my post wasn't to go into great detail, but rather to pass along some general ideas for selecting KPI's.

But since you asked, as I recall the goal was completion of testing within 1 working day, weekends and holidays not included. If the sample was received at noon on Friday, it was "due" by noon on Monday. If the sample was flawed, a rare but possible situation, the clock stopped as we were not responsible for sample preparation, only testing.

What's important or applicable for one lab isn't necessarily important or applicable for another. IMO, the key is satisfying the needs of your customers, internal and/or external.

 

[Jim Wynne]

Dears,
we are revising the KPIs of our testing laboratory that makes tests to qualify products destinated to OEMs, due to the fact the actual ones are not so significant and effective to monitor and evaluate the performances.

(My emphasis)

I'm a little confused. You say that current KPIs aren't doing what they were intended to do, so you must have some idea of what the intentions are. Only you and your organization can determine what's expected of lab processing in general. What one person might claim as efficacious doesn't mean (a) that it actually is, and (b) will be suitable for a lab operating under significantly different conditions. Your organization's management should determine what they want the lab to do, how they are to do it and how to evaluate performance.

 

[Tallinec]

But since you asked, as I recall the goal was completion of testing within 1 working day, weekends and holidays not included. If the sample was received at noon on Friday, it was "due" by noon on Monday. If the sample was flawed, a rare but possible situation, the clock stopped as we were not responsible for sample preparation, only testing.

Please, share your experience.
1)Do you have in your lab the person who is involved in monitoring and calculating "% of samples completed within 24 hours of receipt"?
2)Where from this person/persons receives info about samples completed within 24 hours of receipt and samples not completed within 24 hours of receipt?
3)Or may be your lab is using software where a lab technicians or a serior technicians regulary put info about date and time of sample receipt and date and time of sample completion?

 

[Mike S.]

These days I would use a barcode wand-in wand-out type system and let the computer keep complete track of it.

Back when I did this such capability was not ubiquitous. As I recall we started monitoring 100% of samples manually, the sample prep person would stamp the sample with a timestamp when it was brought to us, the techs would record when the last measurement was made and enter this into a simple computer program that calculated TAT and recorded it. We eventually went to sampling a random 20% of samples for TAT, which still told us what we wanted to know about the health of the system.

No one got beat-up if a sample went over 24 hours. We looked at the system.