Field Service in the Age of AI


Share on LinkedIn

How often is it that a call to a service hotline ends up with the need to have a service technician come to your place? Something like, for example, your washing machine does not work but shows an error code, so you start your research.

The stark reality

The research is easy enough, supported by a chatbot that apparently has access to a good number of useful documents but as none of them is helpful, the underlying AI escalates your call to a human agent.

So far you are having a good experience, considering that in the first instance you suffer a broken gadget in your house. The digital customer journey seems to be good enough.

The agent is friendly enough, too. But hey, the first thing he asks is what your problem is. Doesn’t he have all the information that the chatbot inquired, the search history relating to the current incident and the documents you had a look at, and discarded as not helpful?

From here the journey and the experience deteriorate. He asks for the type of device, address, and contact information; surely, everything pretty important but didn’t you register the device, providing all this information – and didn’t the chatbot know this?

At the end he cannot schedule an appointment but needs to come back to you. Well, at least things are proceeding, or so you think.

A couple of days later you do a follow-up call as you didn’t hear anything from your friendly service agent, who profusely apologizes and goes to work right away. This time you hold the line…

A few minutes later you get the good news. Sincerely the next possible service appointment is 10 days out and the technician will be able to come in the morning between 8 am and 12 pm or in the afternoon, from 12 pm to 4 pm. “No, sorry, I cannot give more precise time slots, the technicians are planning their tours themselves the working day before and they do not really know how long each job takes. I am so sorry. But I can take a note asking him to give you a heads up about 30 minutes before he arrives”.

How good that you have the chance to work from your home office.

Ten days later

The technician who arrives on the announced day is a junior. He starts disassembling the washing machine, looks a bit confused at one part, “uhm, what is this?”, but finds the broken part easily enough. But then: “oh, I do not have this part with me, so I need to return on another day”.

This is admittedly an extreme scenario but sadly still not one that is unlikely to be experienced. Even worse: All of us are likely to have encountered one or the other part of it. Even in B2B settings. Field service scenarios are notoriously difficult to handle efficiently and they face challenges throughout the process.

On the other hand, the most valuable asset that people have is their time. So spending little time in coordinating a repair, having an accurate arrival time of the service technician, and a fast and efficient repair in one visit are paramount traits of good customer service. This is something that e.g. Forrester’s Future of Customer Service Study found. And this is true for businesses, too. Doubly so, as machine downtime equals loss of income and perhaps even penalties.

Service businesses, on the other hand, are asking themselves quite some questions, too, or at least they should.

These are questions like: Who takes how long on repair jobs? Who of my technicians spends lots of time on the road? Why are there so many repeat visits necessary? Are the right technicians assigned to a job? Do I need to upskill my personnel? Where can I optimize my processes and my systems to become more effective and efficient? Am I making the process as frictionless and short as possible for my customers?

There is a better way

All technology needed to avoid the customer nightmare that is described above is already available.

The washing machine is fully equipped with sensors that provide ongoing information on its status. It is registered during setting it up and connected to the home wi-fi. The corresponding app provides the owner with a constant ‘health’ status of his washing machine. Predictive algorithms make sense of the flurry of sensor data and create alerts if parameters indicate a looming failure. These alerts are sent to the customer app so that maintenance can get initiated before the machine breaks. With a consent given this alert is also sent to the registered service company that can already identify the right service technician, initiate a tentative dispatching plan and, most importantly, can proactively call the customer, suggesting a number of possible appointments. All this can even happen without human intervention.

Knowing the most likely cause of the issue and the address of this case as well as for the other cases it is easy to give a fairly precise appointment already early, as work, as well as travel times, can get calculated with fair accuracy. Knowing the cause of the issue also makes sure that the technician has the right parts at hand to get the job done to the customer’s satisfaction.

The customer could be able to reschedule the appointment using the app or a chat tool of his choice, be it WhatsApp, Messenger or iMessage, as the chatbot is connected to various services and the conversational AI with its natural language understanding subsystem recognizes the intent. It knows about the issue, the necessary skills, and thus can offer to reschedule the service.

Adding Uber-like location services to the service technician’s car could provide the customer with a real-time update of his arrival time via the app. That builds even more trust with the customer and makes sure that there is the possibility for ongoing communication as per the customer’s individual needs and wishes.

How to get there

Such a scenario, albeit with connected kitchen devices, is currently being rolled out by a Swiss kitchen appliance maker, based upon Microsoft’s Azure infrastructure. As a caveat, the company does not (yet) look into the field service part.

As said, the ingredients are there. The necessary sensors are getting cheaper by the day, and the software is constantly improved by the software vendors. All of them are offering IoT infrastructures and the AI services that are necessary to operate the described scenario. Salesforce’s Service Cloud just improved its Einstein capabilities for better service interactions. SAP has a leading field service management approach, that even includes crowdsourcing of service technicians, Microsoft has probably the leading infrastructure and is strong in combining this infrastructure with the service business process.

And this very short list is not even mentioning the many specialized partners in their respective ecosystems.

With all the improvements that came in the past year it is now time to address this topic strategically and not as tactically as I described it about two years ago.

Sooner or later the economy will turn again. When this happens, being able to distinguish oneself with the ability to consistently (and efficiently) engage with the goal of a great customer experience will be key.

And the time to prepare for this is now.

Customer experience is a platform play! Consequently, as this topic needs to get addressed strategically, the first important decision is the one for the right platform, which is not just a technology platform, but also includes an ecosystem, insight, and the necessary productivity tools and services to make the difference.

The steps to take

Regardless of the detail, strategy it is important to take some steps early.

  • The most important one is to decide upon a platform. It will enable your business – or not, if you do not choose wisely, based on current needs and the anticipated future. There are not many, but this decision has a long term impact; and this decision is determining what you will be able to do now, and later. And what it will cost.
  • Develop a solid skills matrix of your service personnel and keep it up-to-date. This is the foundation for automatically finding the right technician for any given problem.
  • Identify the process breaks, prioritise, and close them. Some of these will be low-hanging fruit. Use a data-driven approach for this.
  • Data itself is key. It will show you where problems are and will help in training the AI that will help running your system
  • Build a prioritized portfolio of initiatives to implement that get reviewed and reprioritized regularly. This helps in developing a think big – act small mindset to show results and create value frequently.

But most importantly: Communicate with and involve your employees. The introduction of smart software agents will create fear and uncertainty. Use their creativity instead.

Thomas Wieberneit

Thomas helps organisations of different industries and sizes to unlock their potential through digital transformation initiatives using a Think Big - Act Small approach. He is a long standing CRM practitioner, covering sales, marketing, service, collaboration, customer engagement and -experience. Coming from the technology side Thomas has the ability to translate business needs into technology solutions that add value. In his successful leadership positions and consulting engagements he has initiated, designed and implemented transformational change and delivered mission critical systems.


Please use comments to add value to the discussion. Maximum one link to an educational blog post or article. We will NOT PUBLISH brief comments like "good post," comments that mainly promote links, or comments with links to companies, products, or services.

Please enter your comment!
Please enter your name here