In the last decade, field service has evolved into a highly efficient machine, fueled by carefully scrutinized analytics, which provide the raw power for systems such as routing optimization, dynamic dispatching, workload balancing and dynamic parts replenishment. Gone (hopefully) are the days where technicians sit around the office waiting for a dispatcher to call their number, get lost driving to a new location and arrive without the proper part or training to fix the machine.
During my 30-plus years of field service employment, I’ve seen and benefitted from much of the increased productivity and efficiency that thoughtful analysis can facilitate. But there is one bit of data collection that is superfluous at best—and often completely inaccurate: interrupt time.
What Is ‘Interrupt Time’ — and Is It Accurate?
In a quest to balance workload with manpower, and get an accurate assessment of how long it takes to repair their equipment, some field service organizations require their service techs to document what is commonly referred to as ‘interrupt time,’ loosely defined as any interval during a service call that is not spent servicing the equipment.
Time spent doing anything beyond repair work—chatting with customers, phone calls, searching for spare parts, resolving billing problems, even bathroom breaks—are prime examples of interrupt time. With some companies, there’s an individual code to document each type of interrupt, with an expectation that each tech keep track of how long he or she spends in the washroom each day. It may be reasonable for a company to have the desire to know what its technicians do all day, but you might be able to see why this type of analysis can become extremely inaccurate. Consider a typical work day in the life of Bob, one of your star technicians.
Bob has been a tech for many years and has always tried to be accurate in reporting his interrupt time. However, Bob no longer wears a watch since he carries his smartphone everywhere he goes. His phone stays in its case on his hip until he needs it. Bob has been working for hours on a difficult repair. He needs a bathroom break, so he takes the long walk to the other side of the plant for the men’s room. It’s locked. Bob needs to get this repair done, so he grits his teeth and heads back to the machine, on his way, he meets up with the customer, who wants to talk. Halfway through explaining what he has done so far, Bob is interrupted by a phone call from another tech who needs a tool that only Bob carries. Bob agrees to go out to his van and set the tool on the windshield. On the way back from his van, Bob visits the men’s room again only to encounter a sign saying it’s being cleaned. It will take yet another long walk to get relief from his swelling bladder—and even more time spent not working on the machine.
Interrupt Time: Less Accurate Than the Weather Forecast?
The scenario with a fictional Bob is, unfortunately, not an exaggeration. If anything, it underplays the many distractions that today’s field service techs encounter during a service call. Add to this controlled chaos the stresses of compounding service calls and a difficult repair, and it is nearly impossible for even the best technicians to compile an accurate assessment of time spent away from a machine.
So, what gets reported? A somewhat educated guess, which is often far from correct. If a company is planning manpower, measuring productivity, or any other creative metric based on interrupt time reported by its technicians, they are using data that is about as accurate as next week’s weather report.
A Simpler, More Accurate Way to Track Down Time
Technicians do a lot more than repair equipment, and with every advance in technology they are tasked to do even more. Managers should be aware of every aspect of the job, but if they want to count the minutes of each task, they need a more reliable method. Until there’s an app that can automatically differentiate between repair time and interrupt time, I’d suggest a simple solution: Have a first-line manager ride along with each tech with a stopwatch and then average out the time. Techs will act differently when observed, but it will still be a truer representation of interrupt time than relying on estimates from multitasking techs.