Katelyn Burrill | 12.06.17
Summary >

While field service delivery always starts with a plan, it’s never set in stone. In reality, there will be factors beyond anyone’s control that could spoil the best of plans. But what if there was a way to forecast any disruptions or delays before they happen? With the right data and intelligent technology, it’s possible to prepare for the unexpected, and avoid letting disruptions threaten productivity.

Predictive field service uses artificial intelligence (AI), machine learning, and data science to increase schedule accuracy and ultimately deliver better service to customers. With artificial intelligence learning and growing smarter every day, the possibilities for predictive field service are endless. Let’s look at some examples of predictive technology that can benefit service organizations today.

Predictive job duration

Though you may schedule an hour for a technician to complete a repair, odds are the job will take longer or shorter, depending on the job type or skill level of the technician. For example, replacing a utility meter might average an hour, but a more experienced tech might be able to finish the job in half the time. This leads to unnecessary idle time even though that same tech could be completing twice as many jobs that day. Likewise, if the assigned tech takes two hours to complete the job, that means schedule delays and unhappy customers.

Predictive job duration calculates the most accurate time it will take for a technician to complete a job, based on all relevant job details and technician information. In other words, it forecasts how long it will take a particular technician to successfully complete a specific job type. Having this information while scheduling can help you maximize the productivity of your workforce, without over or underutilizing resources.

While many solutions only consider average technician skills and job durations, predictive solutions factor in individual technician skills and historical job data to automate scheduling decisions for you. This includes a particular tech’s past performance during a particular job type or time of day, and when with a specific customer. Averages don’t reflect reality so you can’t count on precision. But using actual and individual performance data, based on multiple scenarios is much more accurate and leads to optimal decisions for your business.

Predictive customer cancellation

No matter how well you plan your schedule for the day, it’s always possible that a customer could cancel and disrupt the entire day. A long-time technician or dispatcher might inherently know when a customer is going to cancel. For instance, on a sunny spring day, you might anticipate more cancellations because customers would rather be out enjoying the weather. With the right data, your scheduling solution can also predict (likely more accurately) the likelihood that a customer will cancel or not be home when a tech shows up, and then plan around it. This way you can mitigate the potential impact of a cancellation, which includes wasted resources, unnecessary travel and fuel costs, and productivity loss.

Predictive customer cancellation considers both structured and unstructured data in its calculations. Structured data includes weather patterns, time of day, and customer demographics, while unstructured data includes dispatcher notes and customer history.

Using this data, a smart field service management solution can automatically avoid scheduling certain jobs when a customer is most likely to cancel. For instance, let’s say a meter replacement is scheduled at an office during business hours. The business owner might not have been aware that power needs to be shut off during the replacement, and cancel as soon as the tech arrives and informs him. On the other hand, predictive technology would consider this factor and only offer timeslots outside of office hours.

Predictive first-time fix

No customer wants to hear “Sorry, we can’t fix your problem today; you’ll have to make another appointment” – especially if they already called out of work or cancelled their plans for one appointment.  Of course, sometimes the problem ends up being more complicated then you originally thought, or the assigned tech didn’t have the right skills to do the job.

With predictive first-time fix, it’s possible to calculate the probability that a particular job will be fixed the first time, based on the technician assigned. In other words, by factoring in individual technician skills and job data, you can determine the likelihood that a particular tech will be able to carry out a first-time fix. Reducing repeat visits is crucial in increasing customer lifetime value. It validates to your customers that they can always count on you for a quick, seamless solution to their problems.

You can assign each available technician a first-time fix rate, and the AI can automate assignments in a way that maintains customer satisfaction and decreases the risk of missing an SLA. For instance, the solution can automatically assign techs with the highest first-time fix rates to higher priority jobs, such as for VIP clients with aggressive SLAs.

Predictive parts management

No matter how skilled a technician is, without the right parts or tools, a first-time fix is impossible. Being said it’s important for field service organization to keep track of parts data and ensure that a tech is always fully prepared before heading to a job site.

Sometimes a customer might describe a problem in one way, and it turns out to be a completely different issue when the tech arrives onsite. The tech might have the parts to fix the problem described, but not the actual problem and have to come back another time. This is neither the customer’s nor the tech’s fault—customers might not have the insight to diagnose a problem, just as a tech can’t help that the problem was misdiagnosed.

However, by tracking both parts data and job history, predictive technology can ensure you are prepared for anything. For instance, if there’s a history of customers calling in with a particular problem and misdiagnosing it, machine learning solutions can recognize this and make sure that the tech has the right tools either way. This lessens the potential consequences of disruptions due to poor parts management, including repeat visits and poor customer satisfaction scores.

With the power of predictive field service on your side, it is possible to keep your business running smoothly, despite any obstacles thrown your way. The intelligent technology can automate and optimize business and scheduling decisions to minimize the efforts of your dispatch team, and free them up for more important tasks. As long as you consistently track relevant metrics and take advantage machine learning technology you can meet customer expectations without overspending or sacrificing your business goals.

In a future post, we will discuss some of the data you should be measuring. In the meantime, for more field service and technology trends, visit the ClickSoftware blog today.