Katelyn Burrill | 12.20.17
Summary >

In a fantasy novel or movie, it’s magic or an innate “gift” that allows people to see the future. In reality, it’s data (and lots of it) that powers predictions. Though it might not be possible to see the future with a mystic vision, it is possible for field service providers to predict the likelihood of a disruption in seconds. With the help of machine learning to quickly process and learn from vast amounts of data, businesses can prepare for anything that could get in the way of flawless service delivery.

But before you can get predictive with field service, it’s important to collect the right data. In this post, we’ll discuss some examples of data you should be tracking, and how you can use each in your field service strategy.

Historical job data

Historical job data includes any information that was relevant and significant to a service visit completed in the past. This includes the task type and duration, success rate, and parts and tools used. It also means looking at both the successes and failures of a particular job. Here are some things you should consider:

  • Was it a first-time fix? Or did it require repeat visits?
  • What kinds of problems came up during the job? (Missing parts, misdiagnosis, etc.)
  • What factors were consistent in the most/least successful visits?

Collecting this data is key in helping you (and your intelligent field service management solution) create the most accurate and optimized schedule. A predictive field service solution can analyze what worked well in past visits, and what caused problems. It can uncover what’s needed to ensure the most efficient service day possible—whether that means moving around the schedule and appointment slots, or assigning certain parts or workers to a job.

For example, you might find that there’s a history of customers misdiagnosing a particular problem and, as a result, technicians showing up with the wrong parts for the call. With historical job data, a machine learning solution will know to automatically assign multiple parts for the visit, so your tech is prepared no matter what.

Technician data

Like snowflakes, no two technicians are exactly alike. Each has their own skills, specialties, and personalities that affect how they perform during a particular task, or even with a particular customer. So when assigning jobs, it’s important to consider everything about the tech. This includes everything from their role and experience level, to their skills and abilities, and their past performance history. This also includes:

  • How long they’ve been with your organization
  • How long they’ve been in a particular role
  • How many times they’ve completed a particular job
  • How they performed with a specific customer (consider customer complaints and praises, or if a customer requests them specifically)

The main idea behind collecting this data is to ensure you assign the tech that is best-suited for that particular job. You want to assign the person who can fix the problem quickly, on the first visit, and make the customer happy. Imagine that a customer’s heating system breaks down in the middle of winter. With an urgent problem like this, you don’t want to send someone who’s never worked on a heater before.

Being strategic about who you assign to a job helps your organization maximize efficiency and complete more jobs per day, without exhausting your resources. It also means more first-time fixes and in turn better customer satisfaction rates.

Customer data

Your customers are just as diverse as your technicians, and each must be treated as individuals when scheduling. Therefore, it’s important to track individual customer data, which includes anything from service history to demographics like age, occupation, and location. It means noting things like:

  • Type of service calls the customer made in the past
  • Who their tech was and how their experience was
  • When (and why) they’ve cancelled or rescheduled an appointment (or were a no show)

Customer data can be especially useful in predicting the likelihood that a customer will cancel or fail to show when a tech arrives. With machine learning to make the right assumptions from demographic and historical data, it’s possible to avoid the added costs and productivity loss that comes with cancellations and no shows.

A predictive field service solution automates appointment booking and creates a schedule that mitigates the risk of a cancellation. For instance, if a customer is a nurse who works overnight shifts, your solution could automatically avoid early morning time slots when they will likely be sleeping. Or if a particular customer has a history of cancelling late in the afternoon, the solution will know to avoid that time slot with that customer.

Environmental data

In addition to data directly tied to your field service organization or customers, there are several external factors that have a significant impact on your service delivery. Insight into things like historical traffic and weather patterns can help you plan your schedule far in advance, while cutting the cost of fuel, maximizing productivity, and reducing the risk of missing an important SLA. Let’s take a closer look at each:


It’s no secret that traffic impacts field service delivery. Morning rush hour traffic can make a tech late for an appointment, and put them behind schedule for the rest of the day. Predictive travel technology uses historical time-of-day and day-of-week traffic data to estimate future travel times, allowing you to plan ahead for it.

Likewise, historical traffic data combined with the location of the tech and job site, allows an intelligent field service management solution to determine in advance the most optimized route on a particular day. Finding the route with the least amount of traffic and quickest travel times leads to more job completion, happier customers, and reduced travel costs. It also minimizes the risk of SLA penalty for missing a customer appointment.


Extreme weather like snowstorms or hurricanes can be detrimental to service delivery. But this weather also means an abundance of emergency jobs that need to be addressed. With machine learning to process historical weather data, a predictive field service solution can proactively add more flexibility in the schedule when there’s a higher chance of extreme weather.  This means scheduling the right mix of low and high priority jobs to you can easily reshuffle the schedule to meet emergency jobs and SLA requirements.

With a powerhouse of data and machine learning technology, you can unleash the power of predictive field service and deliver near-flawless service delivery, every time.

For more on predictive field service and other ways to achieve service optimization, visit the ClickSoftware blog.