Field service has evolved enormously in recent years, and both technology and customer expectations will continue to advance just as quickly.

Twenty years ago, field service might have been centered on break-fix repair, but the direction now is toward long-term, contractual arrangements that are more satisfying for the customer — and more lucrative for the service provider.

Transformational technologies can enable entirely new revenue models that make field-service organizations stickier and more intimate with the customer, even while generating more value. Here are four predictions about how leading service organizations are adapting:

Outcomes-based service takes hold. We will see more companies selling annual maintenance contracts. These contracts are attractive because they give a field-service organization predictable revenue and demand, and can deliver high margins.

For decades, product-centric businesses have been transitioning towards servitizing what they sell. First it was the addition of a warranty, and the availability of after-market parts. Then it was reactive field service or depot repair. As early as 2018, IFS data suggests that 62% of manufacturers were already pursuing some form of aftermarket revenue. But manufacturers are now adopting more advanced forms of aftermarket service, with 16 % of respondents offering maintenance contracts with specific service-level agreements (SLAs).

It’s notable that modern customers not only demand a better service experience, but a holistic outcome. They expect to be left feeling positive, as well as having their specific issues remedied. Technology is a mechanism whereby this change can be implemented, but nobody should overlook the people element as well.

I predict that the share of manufacturers involved in service contracting will rise from 16% in 2018, to between 25% and 30% in 2020.

The move towards servitization in most cases will deliver value-added revenue on top of product sales. In some cases, where it’s attractive to the consumer, a product might be completely servitized, and the end user pays for metered usage or other metric captured in real time. In the 2018 IFS data, only 4% of manufacturers were fully servitized, including companies from the medical device, metal fabrication and oil and gas industries.

For customers who want to push enterprise risk off on their vendors, servitization will be an attractive way to buy. But actually realizing a profit on these contracts poses significant management and enterprise software problems. When the service agreement is sold, a company will be committing to deliver against a contract on which it could make or lose money for years. Executives will need to make sure they have adequate what-if scenario planning capabilities to enable them to deliver quotes that are competitive with minimal risk.

Companies can turn their data into a strategic tool that facilitates service sales while improving the customer experience. For this shift to be successful, they’ve had to put some foundational technologies in place. Research that IFS and Future of Field Service recently conducted with Bill Pollock of Strategies for Growth shows that outcomes-based service operations rely on the foundation of the service management platform, enterprise resource planning (ERP), predictive maintenance, and the internet of things (IoT) in particular.

More than half of respondents were running some type of enterprise system of record that handled the core service transactional business. Fifty-four percent are running their service business on a “dedicated service management platform” like field-service management (FSM) or enterprise asset management, just ahead of ERP) at 50.8%. The survey found 47.6% using software for predictive maintenance, and 42.8% leveraging data from IoT.

With these tools in place, I forecast an increasing emphasis on complete servitization, and that in 2020 we will see the percentage of manufacturers selling products by subscription or metered use surpassing 10%, up from 4% in 2018.

Digital transformation gets harder before it gets easier. People use the term “digital transformation” to sell any number of technologies, but we’re dealing here not with a technology that can be bought, but a fundamentally different way of doing business. The truth is that doing it properly involves a complex journey for service organizations, one that departs from siloed operations, legacy tools, and outdated business processes. The change-management obstacles that surround this are many, as individuals can find the old ways comforting even as the competition is overtaking their business.

What we’ve seen so far is a rollout of transformational technologies at the edge. We track our field technicians’ location through IoT. We schedule them using artificial intelligence. We might have an AI chatbot fielding inquiries online. Maybe we have some AI functionality in our inventory-management processes. We’ll continue to see point solutions using AI and IoT. Where we’re going next, though, is the introduction of AI in particular to the front office and administrative processes.

In its “Top Predictions for 2020” report, Gartner said: “Through 2021, digital transformation initiatives will take large traditional enterprises, on average, twice as long and cost twice as much as anticipated. Large organizations will struggle with digital innovation as they recognize the challenges of technology modernization and the costs of simplifying operational interdependence. Smaller, more agile organizations, by contrast, will have an opportunity to be first to market as larger organizations exhibit lacklustre immediate benefits.”

History is littered with companies that couldn’t change as they needed to: Kodak in the face of digital photography, and Blockbuster Video in the face of servitized and downloadable media, to name just two. Today, we’re at a point where disruptive technologies are embedded at the tip of the spear of forward-thinking service organizations. IoT sensors capture condition-based maintenance information, or an AI algorithm adjusts the field-service schedule in real time based on constantly changing conditions.

In 2020, we’ll see more companies adopt these disruptive technologies in customer- and service-facing settings. But I believe we’ll also see enterprise software vendors move further toward AI-driven automation of the front office in areas like service finance, inventory management, what-if scenario planning and customer interaction. And those who adopt AI as part of a commercial-off-the-shelf solution will win the race against those who take a go-it-alone approach.

IoT grows up, and we’re left with all the data. With more and more organizations saying they have some degree of remote connectivity for their assets, their drivers, and their parts, IoT is now mainstream. We’re collecting large amounts of data and can now develop and apply analytics.

In our study conducted with Strategies for Growth, the biggest area of implementation interest across all industries was in predictive and prescriptive maintenance. Connected assets are the start of the story rather than the destination, and customers are starting to realize that the old adage of “garbage in, garbage out” applies if data collection and hoarding becomes an end in itself.

A good example is multinational telecom company Telefonica, a provider of smart technology to collect data from assets (such as vending machines). The data is then fed into existing analytics for decision support. The power of analytics is significant enough that it will also be important to assure customers that you respect their privacy and data rights. Google faced controversy after its 2019 purchase of Fitbit due to concern over how it would use and monetize end-user data.

During 2020, I predict, businesses will focus less attention on methods by which to collect additional data, and more on making valuable use of the data they’re already collecting.

Companies work to balance AI vs. human experience. As advanced AI becomes more widely adopted among service organizations, companies will seek an equilibrium between the efficiencies of AI and contact with humans that customers and other stakeholders crave. On one hand, greater AI use not only reduces costs but also enables organizations to make better use of resources in sectors facing labor shortages. AI will do a better job meeting certain deliverables, and should automate many repetitive tasks. Will this free up staff for more customer-facing work?

In various service settings, that is exactly what AI has already done. AI-driven schedule optimization, for instance, enables a single dispatcher to support a larger number of field-service technicians, manage by exception, and perhaps spend more time with customers when they need a human touchpoint.

In a Field Services Online article, service intelligence vendor Aquant acknowledged Fortune’s warning in 2016 that 48% of jobs would be lost to AI and robots. At the same time, it said the true future lies in a hybrid between people and AI. As noted by Deloitte in its report Smart Field Service: Connecting Customers, Assets and Employees, “In a digital world, it’s emotional connections that make the difference between satisfying experiences and those that delight the customers and build strong long-term customer relationships.”

Our job now is to use AI to engineer seamless, satisfying automated processes into our business without engineering the human contact out.

Sarah Nicastro is field service evangelist with IFS.