How frustrating to walk off the production floor after a grueling variance meeting and see office workers surfing the web or standing around having a cup of coffee. Even worse, urgent engineering change or variance requests disappear into the ether unless they are constantly expedited. How about customers waiting for weeks to obtain an answer to why a product failed in the field? Should they change their installation practices? Was it a material problem?
Increasingly I’m being asked about back office processes. If it was just a few people causing inefficiency in a department, that’s an easy fix. But it is never quite that simple. It seems like there are waves of busy times and then periods of slow times. Depending on who does the work (we all have our goto people at corporate), the outcomes and response times can be very different.
These are typical scenarios:
Recently I was taking a plant tour with a large Auto OEM and the managing director felt that their production process was in pretty good shape. He then opened a door to a large room full of QC engineers and stated that he wasn’t sure if he had too many, too few or just the right number of engineers working.
A couple of weeks ago I was speaking at a workforce management seminar in Belgium and a manager from a large pharma manufacturer asked about improving the productivity of their QC department whose employees work in a lab environment.
I always ask the same questions. How volatile is their work? Is it slow sometimes and busy others? Are there small jobs and big jobs that flow through the department? Does work get re-prioritized based on production or customer demand? Are the people working in the department reasonably happy and skilled employees?
The answer is generally yes to all four. What I then draw is the curve described by the Kingman formula. What it shows is the relation between utilization and wait time. The drivers of the function are the variability in arrival times and variability in the cycle time.
A good introduction to the formula is available on Wikipedia.
An example of the curve is shown below. What is immediately obvious is that wait times increase dramatically as utilization approaches 100%. The result of this is that departments with little control over the variability of demand and cycle times must run at lower utilization rates in order to maintain acceptable service levels.
And just because a process is documented and looks efficient doesn’t mean that the variability has been driven out of it. Just like the routing on a production floor, an office process is generally documented assuming perfect conditions…open capacity, consistent workload, fully skilled employees and no interruptions. In other words, it assumes perfect standardization.
We are all familiar where high variability in the process and the financial pressure for high utilization causes long wait times. The doctor’s office is a familiar one. For those with appointments at the end of the day, patients can be waiting for 30 minutes or more after their scheduled appointment. It’s tough for the office to schedule the right amount of time for each patient because it’s difficult to know what care each patient will require. It also has to deal with patients arriving late. But if the office doesn’t book the schedule pretty full, the office can’t be run profitably. A doctor’s office is a relatively simple example and many have implemented fixes to ensure higher levels of service while maintaining high levels of utilization. Canceling appointments for patients who arrive late, increasing flexible capacity by adding Nurse Practitioners and scheduling different amounts of time based on the predicted effort for a scheduled patient are a few examples that increase utilization and maintain service levels.
Knowing that standardizing the process and shaping the arrival times of the work will help maintain service levels while allowing for the increase in utilization, the next challenge is identifying where the variability is the greatest. I’ll assume that the first place to look for that answer is with the employees themselves. This is a good start and take what improvements can be identified. The next level gets a little harder. Workflows often cross departments where priorities change and individuals don’t have knowledge of the entire process. This is where some data is going to be required. Often this is also where the improvement efforts slow down. Collecting data around office processes can be challenging. From employee resistance to complex flows it becomes difficult to know what to track. One change in environment that seems to be going underutilized is that office employees are moving to electronic records. From Engineering Changes, to electronic lab notebooks to CAD systems to document management, employees are logging on to applications, doing their work and logging off or checking into another piece of work. As compliance around every aspect of our lives continues to increase, technology producers are tracking our every move to produce a record of what we have done and when. Often this information goes unused unless it is needed for some type of inquiry.
This information is an untapped goldmine for understanding workflow in the office. This electronic trail reconciles work and employee and time. With this you can generate metrics to understand when it is busy, when it is slow, and how long do different types of work take to traverse through a process. When this information is connected to a workforce management system you go from understanding what happened to predicting issues and being able to address them immediately by shifting capacity to where it’s needed and prioritizing work before it is late.
As with production, the idea is not to work the office staff harder. It’s to improve and standardize the workflow through the office so service levels improve without increasing capacity. Imagine lab results returning faster. Engineering requests approved on time without follow-up. Customer inquiries and complaints responded to more quickly. How would this impact your production lead times and competitive stance in the market? All with no increase in labor cost. Kingman has done the hard part by showing you how to improve utilization and service times and what the ROI will be. Now it’s your turn to drive variability out of the office.