Most people, when asked about predicting the future, might joke about using the ability to get rich: bet on a winning team, invest in the right startup, pick lucky lottery numbers. This is fantasy, of course, but the idea that anticipating what’s next can be good for the wallet is very sound. Service organizations with workers in the field are constantly trying to improve efficiency and reduce costs. When they can’t accurately forecast demand, they might end up with too few workers to address an emergency job, or too many technicians loafing around with not enough work to keep them busy. The lost productivity across a service organization without good forecasting, or missed SLAs due to restricted capacity can easily be measured in dollars, sometimes in millions of them. Wasted travel time is just as expensive.
Success metrics for service organizations typically include cost containment, productivity, SLA compliance, and—importantly—customer satisfaction. Travel time has direct impact on all of the above, which means routing is absolutely fundamental to field service. It also makes the impact of predictive routing that much more powerful.
The Challenges of Estimating Travel Time
Estimating accurate travel time is not just an aspiration for field service, it’s an essential capability for organizations looking to maximize efficiency. But it’s difficult to deliver. You need to take into account the location of your technicians and the job site, have information on at least one accurate route, whether other routes are available, the likely travel time along each of those routes, and how and when it is best for the technician to reach the job site in the context of their entire working day. Sounds complex but manageable so far.
Now introduce typical service coverage areas for each available technician: are they simply defined by radius? City, county, or state lines? Is it by proximity to the nearest hub for your business or distribution center for parts and equipment?
Then factor in time of day. Is the shortest route to the job site also the busiest in the time preceding the appointment? Does the level of traffic vary dramatically by time of day, enough to increase travel time by several minutes? Is there construction happening along any of the available routes? Are the weather conditions likely to impact travel time?
Once your dispatcher or field service management solution has accounted for all of the above, can you also rely on historical travel data to verify accuracy? Can these optimized routes be adjusted on the fly as new information becomes available or if emergency work comes up?
As you can see, while plenty of organizations have managed to plan for travel, the level of detail and visibility available for decision making can be daunting, but deliver significant benefits.
The Value of Getting Predictive
The cost of fuel, lost productivity, and SLA penalties regularly eat into the margins of the world’s largest service companies. Low customer satisfaction leads to customer churn and losing business to competitors. It seems as though the level of detail needed to make travel time predictions more accurate is excessive, besides being difficult to deliver. But savings in hundreds of thousands of dollars on fuel costs, or millions in labor costs, are not insignificant. Minutes or hours returned to the days of each happy customer who remains for your loyal brand are nearly priceless.
Field service and mobile workforce management software with machine learning capabilities is making these predictions possible already. Combined with robust mapping data sources such as HERE or Google Maps, and historical data, service organizations are getting smarter and more accurate about travel times each day. When predictive traffic information can be fed into the system (as is the case with Google Maps) the accuracy gains are immense, and the results indisputable.
Some challenges remain. Network availability and latency in different service areas can impact routing information. New locations and addresses not available in your existing mapping source will inhibit predictive accuracy. But until the future is truly transparent to us, at least in this respect we can get very close to predicting it.
Reducing the number of missed appointments and SLA violations while increasing customer satisfaction and accommodating more jobs per technician each day are top priorities for every service organization. The willingness to go after the extra seconds and minutes saved by predictive routing using the best available solutions will be the difference between achieving operational excellence and maintaining operational status quo while competitors lure away your customers. Can you afford not to do it?