Search...

Type above and press Enter to search. Press Esc to cancel.

October 22, 2026 | 8 Mins Read

Crossing the T’s and Dotting the I’s: An IFS Unleashed 2026 Recap

October 22, 2026 | 8 Mins Read

Crossing the T’s and Dotting the I’s: An IFS Unleashed 2026 Recap

Share

Assets UNSCRIPTED | Episode 9 | Berend Booms - IFS Unleashed Recap

This episode is a little different. No guest. Just Berend Booms, a solo mic, and the notes he nearly lost when his phone got stuck in a boot loop midway through the week.

What follows is his recap of IFS Unleashed 2026 - structured, as promised, around five T’s and five I’s.

The Five T’s

Trust

Mark Moffat opened his keynote with a line Berend wrote down immediately: everything has changed. It’s the kind of statement that usually washes over you at these events. But the follow-up stuck: the industrial climate is volatile. Geopolitical pressure, supply chain disruption, rising operational complexity, it’s all eating into margins and raising the bar for what good operations look like.

Against that backdrop, Moffat made a point about AI that’s deceptively simple. If you can’t trust it, you won’t use it.

For asset management, that cuts deep. These are expensive, mission-critical assets. If any layer of technology fails to earn the trust of the person running the machine, the technician maintaining it, or the boardroom planning investment around it, adoption stalls. Every time. Berend observed one speaker describe a technician standing arms crossed, telling them AI wasn’t needed, then spending 40 minutes on a task that could have taken five. For that person, seeing is believing. Tangible outcomes are what moves the needle.

There was also a question on stage that Berend hasn’t quite let go of: should you be able to buy a seat for your robot? Joe Marbella of Southwest Airlines raised it in the context of robotics at trade events. It sounds like a joke, but it opens something bigger, why do we keep designing robots to look, move and speak like humans? Berend’s read: trust. Familiarity lowers resistance.

Time

Two completely different stories about time landed in the same episode.

The first was the Siemens partnership at IFS, AI as a time-saving tool, automating repetitive admin and giving teams back hours they can spend on higher-value work. Every hour not spent on paperwork is an hour that goes back into the operation, especially valuable when human expertise is becoming scarcer.

The second was Badi Narasimhan from William Grant & Sons, and it was Berend’s favourite moment of the week. At a whiskey distillery, some expressions sit in casks for decades before they’re considered ready. Time isn’t a pressure. It’s not a KPI to deliver against. It’s part of the strategy. You account for it. You take enough of it to get things right.

In an industry where mean time to repair and time to value dominate every planning conversation, that reframe landed hard.

Tempo

IFS sponsors the Cadillac Formula One team. On the face of it, it’s a great story about a challenger making it onto the biggest stage in motorsport. But Berend sees it alongside William Grant & Sons as two versions of the same discipline: planning far ahead, just at wildly different speeds.

Moffat’s personal motto - speed is life - sits in interesting tension with warnings from Anthropic CEO Dario Amodei that AI is moving faster than we can govern or audit. IFS CTO Vibes Kumar addressed this head-on in a panel: how do you keep the brakes off while keeping the safeguards up? Berend doesn’t have a clean answer. He does think it’s the right question.

Talent

The skills gap has been a fixture of asset management conferences for years. The numbers aren’t getting easier: US manufacturing alone could need 3.8 million workers by 2033, with up to 1.9 million of those jobs potentially going unfilled.

Pedro Buhigas, CIO of Cote Gas, spoke about deploying IFS Loops agents for materials replenishment and reported an 85% efficiency gain. That’s significant. But Berend’s view is that agents don’t solve the skills crisis. They help bridge part of the gap.

The harder problem is tribal knowledge. A 30-year veteran on the plant floor can hear a bearing fail, smell something wrong, sense a vibration that’s slightly off. They’re already three steps ahead. How do you encode that? How do you get someone two months into the job to the same baseline?

Joe Marbella also spoke about attracting young people into skilled trades. Berend’s take: that next generation expects mobile-first, easy-to-use, readily available tools. If the technology feels like 2005, they won’t stick around.

Teamwork

Tom Brady sat down with Jake Humphrey on the Unleashed stage, and it turned out to be one of the better interviews of the week. Brady talked about his approach to showing up in training, not focused on himself, but on finding common ground with every teammate. What do they care about? What are they working towards? Where are they in their career?

His point was that a quarterback, however good, can’t do it alone. Football is a team sport. So is asset management. Every function, every role, every individual has to do their part.

Two quotes Berend wrote down:

“I haven’t failed. I have just found 10,000 ways that don’t work.” (Thomas Edison, via Brady.)

And Brady’s own: “If all you ever do is all you’ve ever done, all you’re going to get is all you’ve ever got.”

Both speak to the mindset you need if you want to turn AI pilots into something that actually sticks.

The Five I’s

Intelligence

IFS Intelligence was the headline product launch of Unleashed. At its core, it acts as an intelligence layer across your operational tech stack, surfacing insights from your data through conversational interfaces, then feeding those insights back into your ERP, EAM, or digital twin as targeted actions.

The part Berend found most interesting wasn’t the insight generation. It was the action. You can have all the data in the world, but if those insights have nowhere to land, nothing changes. IFS Intelligence routes the output back into the systems that actually do things, purchase orders, scheduling, crew deployment. That’s where it starts to matter.

Chamath Palihapitiya made an observation that added nuance here: AI is currently very good at encoding procedural knowledge, the “how we do the thing” layer. It’s far less effective at encoding tribal knowledge, the judgment, the taste, the human experience that makes the difference between a good decision and a great one. That’s the part you shouldn’t try to automate. That’s the part you post-train on, fine-tune around, and keep firmly in human hands.

Investment

Palihapitiya was also Berend’s favourite speaker of the event, partly because he wasn’t avoiding the uncomfortable parts of the AI story.

His view: most organizations haven’t actually seen a return on their AI investment yet. There’s a line he drew between breathless speculative investment and companies doing the real work responsibly. As cloud infrastructure absorbs more and more of the world’s electricity, he put the US energy deficit at six California's’ worth, the cost of running AI is going to increase. His advice to large enterprises: get into the energy game. Generating your own power may stop being a nice-to-have and become a strategic necessity.

Integrity

A line that Berend noted mid-week: machines may take the action, but you own the outcome.

Satya Nadella was quoted on stage making the same point: we cannot outsource responsibility for what a model does on our behalf. Whatever guardrails are in place, whatever speed the operation runs at, there is always a human who is accountable for the result. That’s not going to change.

Berend also raised a question he’s been sitting with: how do you treat your AI? Are you directive, dismissive, polite? He’s not raising it to be philosophical, he’s raising it because how we interact with these systems tells us something about how we see them, and that shapes how we integrate them into the work.

Insincerity

A gentle frustration, shared honestly. At almost every industry event, someone says it’s all about the people or people are our most important asset. Berend doesn’t disagree with the sentiment. He just finds it more powerful when it’s demonstrated than announced.

Brady put it simply: your actions reflect what your priorities are. If people genuinely matter, it shows up everywhere, in investment decisions, in how you design tools, in how you develop and retain staff. It’s easier to say from the stage than to do in the field.

There’s also a claim that’s been circulating, that the way we work has changed forever. Berend’s pushback: asset management is still rooted in the physical world. When a cooler breaks at 2am, you still need someone to show up and turn a wrench. The opportunity isn’t to replace that. It’s to cut the drudgery, the admin, the planning friction and give people who are excellent in the physical world the time to do what they’re actually there for.

Innovation

The final I came from a field service breakout session hosted by Sarah Nicastro. Mark Brewer introduced her with a slight pause, she’s from the Future of Field Service, that made it briefly sound like she’d arrived from the future itself. Berend ran with the idea.

But Sarah’s point landed regardless of the introduction. Digital sophistication must not come at the cost of humanity. Service is a people business. The technology changes what work looks like. It shouldn’t change whether people are at the centre of it.

Especially now, with the workforce challenges Berend outlined under Talent, investing in people isn’t just the right thing to do. It might be the smartest play available. That’s how you attract the next generation, develop them, and hold on to the skills you’re going to need when industrial AI moves from pilot to practice.

These were Berend’s five T’s: Trust, Time, Tempo, Talent, Teamwork and five I’s: Intelligence, Investment, Integrity, Insincerity, Innovation.

A week at IFS Unleashed never leaves you short of things to think about. Some of the threads from this episode will come back in future episodes and articles. For now, this is where the notes end, rescued, just about, from a phone that had other ideas.

Assets UNSCRIPTED is the Future of Assets podcast series hosted by Berend Booms. Each episode brings together leaders from across asset-intensive industries for authentic, unscripted conversations on the ideas, challenges, and innovations shaping the future of asset management. New episodes are published on YouTube, Spotify, and Apple Podcasts.

Stay Connected

Subscribe to The INSIDER, our exclusive monthly newsletter, and get a first look at what’s new, what’s next, and what’s only shared with our inner circle.

October 21, 2026 | 5 Mins Read

My Top 10 Highlights of IFS Unleashed 2026 (Part 2)

October 21, 2026 | 5 Mins Read

My Top 10 Highlights of IFS Unleashed 2026 (Part 2)

Share

UNSCRIPTED Podcast | Episode 380 | IFS Unleashed

IFS Unleashed 2026 brought together more than 3,000 attendees from over 100 countries in Orlando for a week of keynotes, customer stories, and conversations about the future of industrial operations. In part one of this series, Sarah Nicastro covered her first five personal highlights from the event, including industrial AI becoming tangible, the workforce gap, and standout customer examples from Mo Dawood at Eight Group and Pedro Buhigas at Kodiak Gas. You can read that here: https://www.futureoffieldservice.com/my-top-10-highlights-ifs-unleashed-2026-part-1/

This piece covers highlights six through ten and these are the moments that kept coming up in the conversations after the event wrapped.

Highlight #6: Industrial AI is making the servitization journey more attainable

One of the more substantive conversations of the week happened between Sarah and Ali Bigdeli, professor of industrial service innovation at Aston Business School and co-director of the Advanced Services Group. The topic: how advancements in industrial AI are changing the path toward outcomes-based services.

The Advanced Services Group works with a research framework that maps the forces pushing or pulling organizations toward advanced services. Several of those forces are spiking right now. Customer expectations have risen, shaped not just by competitors in the same industry but by every consumer experience people have day to day. The benchmark is no longer the other company in the sector, it's the ease and real-time visibility people expect from any interaction.

What changes with industrial AI is the risk calculation. Outcomes-based models are, at their core, shared risk models and that has historically been a significant barrier. Better predictive capability, stronger operational visibility, and greater confidence in delivery shifts how organizations perceive that risk. The barrier hasn't disappeared, but it's lower than it's ever been. For organizations that have been hesitant to move toward outcomes-based services, that shift is worth paying attention to.

Highlight #7: Trust becomes a more important form of currency

Chamath Palihapitiya, founder of Social Capital and former senior executive at Facebook, brought a clear-eyed perspective to the week's conversation around AI. Two points in particular stood out.

The first was practical: organizations that are strategic about AI adoption can use it to encode tribal knowledge in ways that reduce dependency on individuals and make the business more resilient. That's an opportunity that many organizations haven't yet thought through carefully.

The second was more fundamental. Palihapitiya made the case that as AI becomes more embedded in decisions and customer interactions, trust becomes an even more critical currency. Being an honest broker, helping people make good decisions even when it doesn't benefit you immediately, these things matter more, not less, as automation increases. For service leaders, the question is how bringing AI into more of your operations and customer relationships changes the ways trust gets built or eroded over time.

Highlight #8: Tom Brady on resilience and what leadership requires

Tom Brady was interviewed on stage by Jay Comfrey of the High Performance Podcast, and the conversation covered more than most expected from a sporting keynote.

Brady talked through his journey from youth baseball to a long NFL career, and made a direct point about resilience: shielding yourself from failure only slows you down. The instinct to position yourself only where you'll look good is understandable, but it's a short-term play with long-term costs.

The more substantive thread, though, was about leadership and relationships. As Brady progressed in his career and took his leadership role more seriously, he said he started to see it as his job to extend himself, to go make the connection rather than wait for it. To learn people's names, understand what matters to them, and approach every interaction with the question of what he could learn from them, not just what he could offer. For service leaders managing teams through significant change, that reframe - from dispensing direction to genuine curiosity about the people you're responsible for - is worth taking seriously.

Highlight #9: Community as a competitive advantage in the AI era

Alexis Ohanian, co-founder of Reddit and founder of 776, made a point that has stayed with a lot of people who were in the room. He called community the Trojan horse of the coming years. As AI makes content faster and cheaper to produce, the ability to build genuine community becomes rarer and more valuable. His phrase for the alternative: "AI slop." Those who figure out how to build real community, he said, become more special, not less, as that content floods the market.

For service organizations, the question is what there is to rally around. Internally, that might be how the workforce understands and connects to the mission. Externally, it might be how customers experience the relationship beyond the transaction. Neither happens by accident, and both become more important as AI handles more of the surface-level interaction.

Ohanian was also asked about legacy. His answer: big ideas that become institutionalized have the most reach. But on a personal level, he said he's more focused on what he called his mini legacy, his children. "I care most about what those two humans think of me. I'm still trying to go out and do work in the world that makes a difference, but make it home in time for dinner." A useful counterweight to a week of big-picture thinking.

Highlight #10: Aki Mandar on doing the right thing, every day

Aki Mandar is the CEO and president of Chelsea Football Club Women, and the first dedicated CEO appointed in the Women's Super League. Her session was one of the most talked-about of the week.

Mandar spoke about the work of transforming women's football into something commercially powerful and globally resonant and she was asked, as a prominent leader in that space, about being a role model.

Her answer was direct: "I do not wake up each day thinking about being a role model. If you do, that might be a red flag." What she wakes up thinking about is figuring things out, fixing problems, proving people wrong when necessary, and holding power to account.

"I get up thinking about doing the right thing, working hard, and helping other people."

It's a reminder that's relevant well beyond football. Every leader, in their own context, has the potential to be a role model to someone. The ones who do it well tend not to be performing the role, they're just showing up, doing the work, and leading with integrity. That's what people remember.

Follow Sarah on LinkedIn for daily insights, behind-the-scenes reflections, and the conversations shaping the future of service.

Connect with Sarah Nicastro on LinkedIn

Follow Future of Field Service on LinkedIn for new articles, podcast episodes, and event updates as they go live.

Follow Future of Field Service

Subscribe to The INSIDER - our monthly newsletter with exclusive content you won’t find anywhere else.

Subscribe to The INSIDER

Stay Connected

Subscribe to The INSIDER, our exclusive monthly newsletter, and get a first look at what’s new, what’s next, and what’s only shared with our inner circle.

Most Recent

October 7, 2026 | 5 Mins Read

Honeywell's AI-Powered Service Evolution

October 7, 2026 | 5 Mins Read

Honeywell's AI-Powered Service Evolution

Share

STANDOUT UNSCRIPTED - EPISODE 379 • AI IN SERVICE • DIGITAL TRANSFORMATION

Yasir Sheikh, Global Service Transformation Leader at Honeywell Technologies and 2026 Stand Out 50 leader, joins Sarah Nicastro to explore what AI is actually changing in service delivery. Technology, it turns out, is only half the story.

Yasir Sheikh has spent most of his career in service and operations, working across building technologies, automation, and customer support at Honeywell Technologies. He manages remote service teams supporting customers across a wide range of industries. His focus is helping customers get the most value from their systems while driving innovation within the service organization itself.

The conversation covers a lot of ground. AI and predictive maintenance. The distinction between managed services and outcomes-based service. The evolving role of the service engineer. And what gets missed most often in digital transformation, which, in Yasir's view, has very little to do with technology.

Customers Are Buying Outcomes, Not Hours

The biggest shift Yasir points to is customers moving from ownership to outcomes. Historically, many customers wanted to manage and maintain their own systems and call in support when something broke. Systems are now so interconnected and complex that many customers want someone else to take on more responsibility for them.

The examples are practical. A hospital doesn't want to know how many maintenance visits happened this month. It wants to know that critical environments stayed operational and patients weren't impacted. A data center isn't focused on individual assets. It cares about uptime and resilience. Managed services exist to continuously monitor systems, analyze data, and prevent problems before they become disruptions.

"Customers are looking for partners who can predict issues, optimize performance, and reduce risk. That's where managed services becomes so valuable.", Yasir Sheikh

What AI Is Actually Changing at Honeywell

As a technology company, Honeywell has been an early adopter. The biggest practical change Yasir describes is the shift from reactive to predictive service. AI analyzes data from connected equipment, identifies patterns that suggest something isn't operating correctly, and flags what needs attention before failure occurs.

Engineers have always had data available. What AI does is help them identify what's important and what needs to be prioritized. Scheduling, troubleshooting recommendations, emerging asset issues, AI is making decisions faster and better informed. The framing Yasir is clear on: this isn't about replacing engineers. It's about making them more effective by giving them better insights and cutting time spent on repetitive tasks.

The next evolution he sees is prescriptive service, AI that doesn't just identify a potential problem but recommends the best course of action, the resources required, and when to intervene. Customers may not even know there was a problem because it was identified and resolved before anything was impacted.

"Service conversations will shift from talking about maintenance activities to talking about business outcomes.", Yasir Sheikh

Managed Services and Outcomes-Based Service Are Not the Same Thing

The two terms get used interchangeably. Yasir draws a clear line between them.

Managed services means taking responsibility for operating and maintaining a customer's environment: actively monitoring, managing, and optimizing systems on their behalf. Outcomes-based service goes a step further. It's measured against specific business results, uptime, energy savings, operational efficiency, agreed performance metrics.

Yasir's framing is useful: managed services is the foundation that enables outcomes-based service. You need visibility, data, and operational control before you can confidently commit to outcomes. The leap from traditional service straight to outcomes-based commitments is a much harder one to make. Build the foundation first.

The Trifecta: Technical, Digital, Human

The engineer of the future, in Yasir's view, will need three things. Strong technical skills, that hasn't changed. But also comfort with data, digital tools, and AI-generated insights. And something that's becoming more important as AI takes on more of the diagnostic work: human skills.

Sarah connects this to a point from a recent Service Council event, the growing need to upskill technicians on prompt engineering, the ability to get good outputs from generative AI tools. Yasir agrees and goes further. Organizations need to be investing in analytics capabilities, consultative skills, and digital acumen alongside technical training. The engineer who succeeds will be the one who can combine human expertise with digital intelligence.

"The engineer who succeeds will be one who combines human expertise and digital intelligence. That's what the future engineer looks like.", Yasir Sheikh

Why Change Management Gets Missed

Organizations tend to focus on the technology first. It's easy to get excited about AI platforms, analytics, and digital transformation tools. What's harder is helping people adopt a new way of working. That's where most transformations struggle.

People naturally wonder what change means for them. Whether they'll need new skills, whether their roles will shift. If leaders don't spend enough time addressing those concerns, even good technology can fail. The most successful transformations Yasir has seen happen when employees can clearly see how change benefits them as well as the business.

The most impactful thing in the process, in his view, is trust and understanding. Not just communication, understanding. Making sure people genuinely grasp where the organization is going, why, and how they'll be supported. When people understand that AI can eliminate repetitive tasks and help them deliver more value, they tend to get excited rather than concerned.

"Transformation is a people journey as much as a technology journey.", Yasir Sheikh

Transformation Is a Slow-Cooked Sunday Roast

Sarah brought in a reference from her conversation with Peter George of Alfa Laval, who described pace as the hardest part of transformation. The excitement to move fast is real. But moving too fast means you can't build the trust you need, and you can't take people with you.

Yasir's analogy landed well: transformation is a slow-cooked Sunday roast, not a two-minute noodle. It takes longer. But the result is better. Organizations that reconcile their enthusiasm for the opportunity with the pace at which they can genuinely bring their people along are the ones that go furthest.

His closing thought: we're at a pivotal moment in service. Technology is a huge enabler. But people remain at the center of success. The organizations that combine technology investment with genuine investment in people development are the ones that will realize the full opportunity ahead.

Follow Sarah on LinkedIn for daily insights, behind-the-scenes reflections, and the conversations shaping the future of service.

Connect with Sarah Nicastro on LinkedIn

Follow Future of Field Service on LinkedIn for new articles, podcast episodes, and event updates as they go live.

Follow Future of Field Service

Subscribe to The INSIDER - our monthly newsletter with exclusive content you won’t find anywhere else.

Subscribe to The INSIDER

Stay Connected

Subscribe to The INSIDER, our exclusive monthly newsletter, and get a first look at what’s new, what’s next, and what’s only shared with our inner circle.

Most Recent

October 6, 2026 | 4 Mins Read

Why Does Your Win Need to Come from AI? A Conversation with Ashley Gross, The Alliance

October 6, 2026 | 4 Mins Read

Why Does Your Win Need to Come from AI? A Conversation with Ashley Gross, The Alliance

Share

Episode 8 | Assets UNSCRIPTED with Ashley Gross, The Alliance

he gap between what organizations say they want from AI and what they're actually ready for is rarely a technology problem. It's an operational one. Processes that don't work before AI don't start working after it. Departments that don't trust each other's data don't suddenly align because a new tool is in the room. Ashley Gross has spent the last several years helping organizations figure this out the hard way, so they don't have to.

In this episode of Assets UNSCRIPTED, Berend Booms sits down with Ashley Gross, founder and CEO of AI Workforce Alliance and author of The AI Workweek, published by Wiley. Ashley's entry into AI productivity came not from a technology background but from necessity: becoming a new mum in 2020 with no maternity leave and an 80-hour work week that had to change. She rebuilt it down to 15 hours before ChatGPT became a household name, and has since built a career helping Fortune 500 companies and mid-market leadership teams figure out what AI actually delivers in practice.

The conversation covers how AI readiness depends far more on communication and alignment than on tools, why AI is putting more pressure on senior leaders than on entry-level staff, and what Ashley does when a CEO walks in and asks for an AI win. It also touches on AI fatigue among frontline workers, the tension between boardroom mandates and shop-floor reality, and whether Dario Amodei's recent call to slow AI development is a genuine concern or well-timed positioning. Watch the full episode on YouTube, or listen on Spotify and Apple Podcasts.

From 80 Hours to 15: What Ashley Actually Did

Ashley's story starts in 2020, before the current wave of AI tools existed. Facing a new baby, no maternity leave, and an 80-hour work week, she started by doing something basic: timing her tasks. Every activity she logged, documented, and recorded herself talking through. That process surfaced what was actually eating her time, mostly administrative work she was doing for visibility and gratitude rather than real impact. The step after awareness was finding the right tools and testing them, one by one, with no playbook to follow. The discipline of that process, not the tools themselves, is what she's been teaching organization's since.

AI Is Coming for the Executives

One of the sharpest observations in the episode is Ashley's view on who AI is actually disrupting. Entry-level workers get the headlines. But Ashley's experience points somewhere else. It's coming for the executives, the senior managers, the C-suite. Not because AI is better than them at their jobs, but because the habits they built to get where they are no longer fit the pace of change. The leaders still running daily check-ins with every team member, still expecting two Slack updates a day, still operating the same way they did five years ago: those are the ones getting replaced. Not by AI, but by people who've adapted to it.

Why Does Your Win Need to Come from AI?

When a CEO walks into a meeting and says they need an AI win, Ashley's first question is the one that gives the episode its title: why does the win need to come from AI? It's a genuinely disorienting question to ask a CEO, which is exactly the point. AI is a vehicle that gets you from A to B faster. But you still need a driver who knows where they're going. Her case, made consistently throughout the conversation, is that organisations reach for AI before they've figured out what they're trying to do, and then blame the technology when the results don't materialise.

The Business Therapist

Ashley describes her role as sitting in rooms between IT and business teams doing what she calls business therapy. The problems she sees are rarely technical. They're communication failures dressed up as technology problems. Her example is sharp: a 500-person company where two analysts spend hours creating a report that Salesforce can generate automatically, because marketing doesn't trust a report pulled by sales and sales doesn't trust one pulled by marketing. The fix isn't a new tool. It's a 30-minute monthly check-in and a shared definition of what they're trying to measure.

AI Fatigue and the Frontline

Berend raises a tension that will resonate with anyone working in asset-intensive industries: frontline workers who couldn't care less about AI. Ashley's approach is to stop talking about the technology entirely. She asks people what their five-year goal looks like. Almost everyone says they want more time. Then she reverse architects from that number back to what AI could actually free up in their current week. It's a conversation about their goals, not about the technology. That reframe is usually what gets people to listen.

Assets UNSCRIPTED is the Future of Assets podcast series hosted by Berend Booms. Each episode brings together leaders from across asset-intensive industries for authentic, unscripted conversations on the ideas, challenges, and innovations shaping the future of asset management. New episodes are published on YouTube, Spotify, and Apple Podcasts.

Stay Connected

Subscribe to The INSIDER, our exclusive monthly newsletter, and get a first look at what’s new, what’s next, and what’s only shared with our inner circle.

Most Recent

September 30, 2026 | 6 Mins Read

Podcast Takeover: Future of Field Service Live London Highlights

September 30, 2026 | 6 Mins Read

Podcast Takeover: Future of Field Service Live London Highlights

Share

UNSCRIPTED - EPISODE 378 • EVENT RECAP • PODCAST TAKEOVER

Berend Booms, Associate Editor of Future of Assets, takes over the mic to share his key takeaways from Future of Field Service Live London. The event through fresh eyes, from someone who came in from outside the field service world and left genuinely surprised, moved, and energized.

Sarah handed the mic to Berend Booms, Associate Editor of Future of Assets, to recap Future of Field Service Live London. Berend came to the event as an outsider to the field service world. That perspective makes this episode something different from a standard recap. He noticed things that those inside the industry might have taken for granted, and he wasn't afraid to say so.

He also took Rob Stevenson's keynote idea and ran with it. For every part of the day's agenda, Berend picks a song. The playlist alone is worth the listen.

The Venue

*Alexa play We Are Going to Be Friends, The White Stripes*

The Orchard, eighth floor, central London, close to Hyde Park and Marble Arch. The views were good. The coffee was good. The balcony access was a bonus. What stood out most to Berend was that it didn't feel like a conference centre. No stuffy carpets, no rows of identical chairs. It felt like somewhere people actually wanted to be.

Sarah's Opening Keynote: Amplification, Not Optimization

Sarah opened by talking about priorities. She's a mom first and a leader second. Berend found that grounding. It's an honest acknowledgement that everyone in the room is juggling multiple roles, and that the balance between them is real and often difficult.

The substance of the keynote centred on what Berend describes as three seismic shifts redefining service. The one that stuck with him most: operational excellence has moved from being the goal to being the starting line. The bar has risen. The ambition now has to be amplification, doing the best work possible with the least expenditure of energy and resource, and doing it without sacrificing the humanity that makes service what it is.

Leadership, Sarah argued, is the bridge between what service has been and where it is headed. And the shift she's advocating for is from change management, which tends to be project-focused, to change leadership, which is ongoing and continuous. Berend found that a genuinely useful reframe.

"Leadership is action. It is not a position." -- Sarah Nicastro

Rob Stevenson: Energy Is the Currency of Sustainable Performance

*Alexa play Lovely Day, Bill Withers*

Rob Stephenson, author of Music is the Answer, was the speaker Berend was most grateful for. Not because the session was the most technically relevant to field service, but because it didn't try to be. It was about people. About energy. About wellbeing as something that should be built into how an operation runs, not bolted on as a perk.

The exercise Rob ran, getting everyone in the room to share the one song that never fails to give them energy and seeing it appear on the big screen, is what inspired Berend's song picks throughout the rest of this episode. His own pick for Rob's session reflects exactly what the session gave him.

"Your wellbeing is not a perk. It is not a policy. It should be baked into how the business actually operates." -- Rob Stephenson

Natasha Nelson, Schneider Electric: The Talent Crisis Is Compounding

*Alexa play High Voltage, AC/DC*

Natasha Nelson, CTO of Services at Schneider Electric, spoke about the talent shortage. The detail that landed for Berend: a plumber or HVAC engineer can now get a mortgage more easily than a lawyer or a dentist, because the demand for trades is that high. And Sarah added the number that frames the scale of the problem. For every five technicians leaving the workforce, only two are joining.

Amplification starts to make sense in that context. If you can't solve the workforce gap through recruitment alone, the answer has to involve getting more out of the people already there. AI is the obvious tool for that conversation. But Natasha's point was clear: the challenges in AI deployment are almost always human ones. Technology isn't the hard part. People are.

Patrycja Sobera, Unisys: Drink Your Own Champagne

*Alexa play X, Xzibit*

Patrycja Sobera, SVP and General Manager of Digital Workspace Solutions at Unisys, introduced Berend to a concept he hadn't encountered before: XLAs, experience level agreements. The difference from SLAs is that they're intentionally short-lived. They force evaluation. They allow adjustment. They let you respond to change faster than a standard contract ever could.

She also gave him a phrase he's replacing an old habit with. He's spent years saying "eat your own dog food." He's now going with Patrycja's version: drink your own champagne.

Simon-Kucher Workshop: What If We Just Train the Salespeople to Do Maintenance?

*Alexa play Holding Out for a Hero, Bonnie Tyler*

Simone Schneider and Torben Petersen from Simon-Kucher ran a workshop on how to unlock commercial value from frontline workers. The basic premise: technicians are already with customers. They have a level of customer intimacy that sales functions often don't. How do you help them surface opportunity without turning them into something they're not?

The pushback from the audience was immediate: isn't that just making them salespeople? One person's response cut through: why don't we just train the salespeople to do the maintenance work instead? It was tongue-in-cheek. But it reframed the whole debate and made the room think differently. Berend's takeaway: there is no one-size-fits-all solution, and the moments where someone flips the script entirely are often the most useful.

Think Tank: Innovation That Doesn't Start or End with AI

*Alexa play Go with the Flow, Queens of the Stone Age*

The think tank brought together Clinton von der Merbech of Tomra Cycling, Yasir Sheikh of Honeywell, and Nick Vanderveer of Nexus Black to talk about the human side of digital transformation.

Clinton's observation about flying back from China and feeling like he was returning to the Stone Ages said a great deal about the pace of innovation elsewhere and the gap that's widening. Nick's point about frontline feedback was equally sharp: if you only build what the field asks for, you end up with a slightly nicer version of what already exists. Real innovation requires going beyond what people can currently imagine needing.

Berend's broader takeaway: fear of AI is real, but conflating job loss from economic conditions with job loss from AI isn't fair. The more useful question is what do we do about it.

The Stand Out 50 Announcement

*Alexa play With a Little Help from My Friends, The Beatles*

Berend's favourite part of the day. Not just because of the celebration, but because of what it represents. Roy Dockery, one of the judges, articulated it well: we recognise sales, we recognise revenue. Where do we recognise leadership? These awards exist to close that gap. To make people feel seen. Understood. Appreciated.

"Joining a live discussion, experiencing a sense of community, asking someone about their favourite song. It is something that is much more rewarding than anything any prompt will ever be able to generate." -- Berend Booms

Follow Sarah on LinkedIn for daily insights, behind-the-scenes reflections, and the conversations shaping the future of service.

Connect with Sarah Nicastro on LinkedIn

Follow Future of Field Service on LinkedIn for new articles, podcast episodes, and event updates as they go live.

Follow Future of Field Service

Subscribe to The INSIDER - our monthly newsletter with exclusive content you won’t find anywhere else.

Subscribe to The INSIDER

Stay Connected

Subscribe to The INSIDER, our exclusive monthly newsletter, and get a first look at what’s new, what’s next, and what’s only shared with our inner circle.

Most Recent

September 23, 2026 | 5 Mins Read

5 Takeaways from the Service Council Symposium 2026

September 23, 2026 | 5 Mins Read

5 Takeaways from the Service Council Symposium 2026

Share

UNSCRIPTED - Episode 377 | Solo Episode | Event recap

The Service Council Symposium ran Monday through Wednesday in Chicago. Three days, a new venue, and a lot of familiar faces alongside some new ones. Sarah didn't attend every session, this isn't a full recap. These are the five things that stood out to her personally.

1. Employee Experience Is Getting More Attention: and the Conversation Is Finally Going Deeper

The link between employee experience and customer experience isn't new. Most people in service will tell you they already know it's true. What was different at this year's Symposium was that the conversation didn't stop at the acknowledgement.

On a panel Monday with leaders from Hitachi, Otis, Staris, and Microsoft, Laura Mather of Staris made the point directly: if you take care of your people, they will take care of your customers. That's a principle that sounds straightforward, but in practice requires a very different set of priorities than most organizations operate with. The discussion at the event reflected that more leaders are starting to make those decisions differently.

The same panel raised something worth sitting with: leaders now need to manage both humans and agents. That sounds like a technology question, but it's really a leadership question. Strong people leaders need new skills to work effectively with AI systems. Technically skilled leaders need to make sure they're not losing the human connection with their teams. Neither of those transitions is easy, and the panel suggested we're not talking about them enough yet.

One practical note that stood out: nobody on the panel felt that AI has yet improved or shortened onboarding for their organizations. Eight panelists. Zero. That gap between what's possible and where organizations actually are is significant, and closing it will require work that's less glamorous than the vision.

"Focus on the digital rocks, not the digital popcorn.", Marina Guimaraes, Microsoft

2. The Benefits of Digital Transformation Need to Land on Both Sides

Sarah moderated a session on Monday with Dave Mencarini of Ecolab on the company's AI journey. One of the threads that came through most clearly: when we measure the gains of digital transformation, we tend to measure them in terms of efficiency and cost. Seven hours saved per technician per week. Headcount held flat. Those numbers matter. But they tell a one-sided story.

If employee experience and customer experience are genuinely linked, that has to show up in how we think about the benefits of AI. The tools being deployed to increase efficiency can also meaningfully improve the day-to-day experience of the people using them, if that's part of the design intention.

Corey Aulet from IFS Nexus Black shared an example from a recent meeting with a technician team. Several members raised concerns about a new AI-powered mobile app because a number of people on the team have dyslexia. His response was to see it as an opportunity to incorporate voice and photo-based elements. That's the same technology, evaluated through a completely different lens. The outcome is better for everyone.

3. Outcomes-Based Service: Dating vs. Getting Married

Sarah moderated a panel on Tuesday with Sasha Ilyukhin of Tetra Pak and Emrah Ercan of Hitachi Energy on outcomes-based service. Emrah opened with an analogy that reframed the whole conversation: traditional service is like dating; outcomes-based service is like getting married. The commitment level is categorically different.

The panel covered pricing, commercial strategy, risk appetite, and the mindset shift the model requires. At the end, during audience Q&A, someone asked who actually benefits more in an outcomes-based agreement: the provider or the customer. Emrah's response brought the analogy back around: in a marriage, who benefits more, the husband or the wife? Long-term success requires mutually beneficial gain, or it doesn't hold.

4. What Frontline Workers Actually Want

The Service Council runs a Voice of the Frontline panel each year. Four frontline representatives were asked one question: what's one thing that would make your life better?

The answers were simple. Less silos and more collaboration across departments. Better route planning to cut wasted time. Clearer communication. And the one that stayed with Sarah: removing friction so we can do more of what we love.

These aren't complicated asks. They're not requests for new technology or higher pay. They're requests for organizations to remove the things that get in the way of people doing good work. The gap between what frontline workers need and what gets prioritized in service leadership conversations is real, and it keeps showing up.

5. Culture Allows People to Do What They Already Know Is Right

Wednesday morning's panel on co-elevating humanity and technology was Sarah's favourite session of the week. Rob Orager of EchoStar made the point that KPIs only tell part of the story. If they're used in isolation, they can incentivize the wrong behavior. Efficiency matters. Lead with efficiency above everything else, the unspoken message to your team is that efficiency is everything. That's rarely what you actually want.

The point that landed hardest came from Melissa Perkle Bailey, also of EchoStar, who leads training. She was asked how to increase empathy in a team. Her answer: it's far less important to teach someone how to display empathy than to create a culture that allows them the opportunity to display it.

"It's far less important to try to teach someone how to display empathy than it is to create a culture that allows them the opportunity to display it.", Melissa Perkle Bailey, EchoStar

The principle applies beyond empathy. Most people already know what the right thing to do is in a given situation. What gets in the way is whether they feel allowed to do it. That's a culture question, not a training question. Sarah says she'll come back to this one.

Follow Sarah on LinkedIn for daily insights, behind-the-scenes reflections, and the conversations shaping the future of service.

Connect with Sarah Nicastro on LinkedIn

Follow Future of Field Service on LinkedIn for new articles, podcast episodes, and event updates as they go live.

Follow Future of Field Service

Subscribe to The INSIDER - our monthly newsletter with exclusive content you won’t find anywhere else.

Subscribe to The INSIDER

Stay Connected

Subscribe to The INSIDER, our exclusive monthly newsletter, and get a first look at what’s new, what’s next, and what’s only shared with our inner circle.

Most Recent

September 16, 2026 | 6 Mins Read

The Hidden Risks of AI in Service: Why a Strong AI Defense Matters

September 16, 2026 | 6 Mins Read

The Hidden Risks of AI in Service: Why a Strong AI Defense Matters

Share

UNSCRIPTED | Episode 376

Dr. Aymen Gatri, Professor of Cybersecurity and AI Defense and Director of PMO Products at Konecranes, joins Sarah Nicastro to add a layer of pragmatic necessity to the AI buzz and make the case for why AI defense deserves as much attention as AI offense.

This episode is not about dampening excitement around AI’s potential. Aymen Gatri makes that clear from the start. The opportunity is real. The urgency is real. If organizations don’t engage with AI now, they will fall behind. The point of this conversation is something different: making sure that the genuine excitement around what AI can do doesn’t obscure a clear view of what organizations need to prepare for.

Aymen brings an unusual combination of perspectives to this conversation. He teaches cybersecurity and AI defense at the Digital Business University of Applied Science in Berlin. He also works in aftermarket service at Konecranes. The intersection of those two worlds is where the most important, and least discussed, risks to service organizations live.

The Gap Between Hype and Opportunity

Aymen’s framing of the risk is precise. There is hype around AI. There is genuine opportunity. The gap between those two things is where risk lives. Organizations that rush into AI adoption because everyone else is, because the PR is overwhelming, or because the excitement has outrun the analysis, those organizations are taking on risk they haven’t priced in.

The core case for AI in service is clear. The field service workforce shortage is real, growing, and not going away. AI’s ability to capture knowledge, distribute expertise, and support technicians in the field addresses something that was previously very hard to address at scale. First-level support, second-level support, knowledge transfer, technician guidance, these are the use cases that make immediate sense, and where the benefits are already visible.

“There is so much PR around AI. Is this hype, or is this a chance? Or is there a gap between both? That gap is where the risk exists.” — Dr. Aymen Gatri

The Three Moves and Where the Real Risks Begin

Aymen’s framework for understanding AI risk in service is structured around three moves. The first move is using AI as a tool: knowledge management, first-level support, decision assistance. This is where most organizations currently are, and where the risk is manageable.

The second move is where the risk profile changes fundamentally. When AI transitions from assisting decisions to making them, when AI agents start to act autonomously on behalf of the organization, questions of governance and accountability that haven’t been answered yet start to matter enormously. Who is accountable for what the agent does? Who owns the data it touches? What happens when it hallucinates in a mission-critical context?

The third move, where AI agents are making complex service decisions with significant operational or financial consequences, is where the risk is greatest and where the least preparation has been done.

“Starting from the next move, it’s not just an AI system doing service instead of several people. It is an AI agent that is asking, questioning, thinking, doing, reflecting, and maybe deciding. And this is where we start to encounter risks we haven’t considered.” - Dr. Aymen Gatri

Cybersecurity, Data Ownership, and the Usability Risk

The cybersecurity risk in AI adoption is not theoretical. In Europe, there are already significant regulations around product cybersecurity, organizational cybersecurity, and critical infrastructure and the AI layer adds complexity to all of them. Organizations implementing AI systems are creating new vectors for data exposure, and in many cases they haven’t mapped the data lifecycle clearly enough to understand what is being accessed, processed, or shared.

The usability risk in aftermarket service is equally significant and less widely discussed. An AI agent that isn’t governed well can make its own service decisions, discounts, escalation paths, recommendations, in ways the organization didn’t intend and can’t control. The reputational and financial exposure from that is real, and it can move fast.

There is also a trust dimension that mirrors what the service industry experienced with remote access a decade ago. Customers who were resistant to allowing remote access to their machinery are now being asked to trust AI agents with even greater access and autonomy. That trust has to be earned incrementally, built on a foundation of data privacy and cybersecurity assurance. Organizations that rush past that foundation are taking a risk with the customer relationship that is hard to recover from.

“Consider that you have an AI agent which maybe you don’t know how to control. How will customer acceptance be? This journey is not an easy journey. It’s not a quick journey.” - Dr. Aymen Gatri

AI Offense and AI Defense

The framing that sharpens everything in this conversation is the distinction between AI offense and AI defense. AI offense is what most organizations are currently focused on: implementing AI capabilities as quickly as possible, capturing the opportunity, keeping pace. AI defense is what most organizations are not focused on: building the governance, accountability, and risk management structures that make sustainable AI adoption possible.

Sarah’s basketball analogy lands well here. Offense gets the glory. But defense wins games. An organization that builds impressive AI capabilities without an equivalent AI defense capability is vulnerable, to data breaches, to hallucinations with real-world consequences, to regulatory exposure, to loss of customer trust, and to governance failures that become visible only when something goes wrong.

“AI defense is about building a very important strategy for implementing AI. You want to defend your business, your information, your privacy, your intellectual property. This is the strategy that allows you to take the opportunity while keeping the risk under control.” - Dr. Aymen Gatri

The Role of Governance and the CISO Parallel

Aymen’s recommendation is that organizations need a dedicated governance role for AI, someone whose job is not to build AI systems, but to hold the strategy for how AI is implemented, what risks it introduces, and how those risks are managed across the business. He draws the parallel to the Chief Information Security Officer role, which emerged as cybersecurity expanded from an IT function into a cross-organizational strategic concern.

Sarah pushes back thoughtfully on whether a standalone Chief AI Officer is the right structure, raising the risk of silos and the importance of AI literacy as much as AI execution capability. Aymen’s response is that the need isn’t driven by hype: it’s driven by risk. The role exists to provide a home for accountability, governance, and strategic vision, not to build things, but to ensure the organization knows what it is building toward, and what it is building against.

The message both agree on: organizations don’t have to build every AI capability themselves. The solutions exist. What they need to build is the internal capacity to evaluate them intelligently, implement them responsibly, and govern them continuously.

“Risk is not constant. Risk is variable. You never know how big the risk could be. Nevertheless, there is always a huge opportunity beyond the risk. So adapt — with a strategy behind that adaptation.” — Dr. Aymen Gatri

Follow Sarah on LinkedIn for daily insights, behind-the-scenes reflections, and the conversations shaping the future of service.

Connect with Sarah Nicastro on LinkedIn

Follow Future of Field Service on LinkedIn for new articles, podcast episodes, and event updates as they go live.

Follow Future of Field Service

Subscribe to The INSIDER - our monthly newsletter with exclusive content you won’t find anywhere else.

Subscribe to The INSIDER

Join us at Future of Field Service Live London on 24 September. A day of conversation, insight, and community with service leaders across industries. Register here.

Stay Connected

Subscribe to The INSIDER, our exclusive monthly newsletter, and get a first look at what’s new, what’s next, and what’s only shared with our inner circle.

Most Recent

September 9, 2026 | 6 Mins Read

Building High-Performing Field Service Teams: From Technical Skills to Transformational Behaviors

September 9, 2026 | 6 Mins Read

Building High-Performing Field Service Teams: From Technical Skills to Transformational Behaviors

Share

UNSCRIPTED - Unscripted 375

David DeAngelis, Senior Director of Technology Training and Support Services at Brightstar Lottery, joins Sarah Nicastro for a detailed, practical conversation about how to identify the behaviors that distinguish top-performing field service talent — and build a program that scales.

David DeAngelis has been with Brightstar Lottery, then called GTEC, since 1988. He built the training organization from the ground up, and today oversees training and support for more than 600 field service technicians across 26 states. His team trains every person involved in making a lottery work: the technicians who install and service terminals, the retailers who sell tickets, the data center staff who run the systems.

When Brightstar’s field service leadership started telling technicians they were the face of the company, that customer experience was as important as technical execution, the technicians asked a reasonable question: where’s the training for that? This episode is the story of the three and a half years that followed.

The Light Bulb Moment

Brightstar had a rigorous technical competency model. Every task a field service technician needed to perform was logged, cataloged, and tracked through a career development system. There was no ambiguity about what technicians needed to learn technically, or what they needed to demonstrate to move up.

There was nothing equivalent for professional skills. No framework. No language. No training. And when supervisors were asked to describe what made their top performers stand out, many of them struggled to articulate it beyond “I just know.”

“When the field service leadership was out talking to our technicians and saying, ‘You’re much more than technicians to our company - you’re the face of our company,’ the questions started coming back: where is the training for that? That was the genesis of this program.” - David DeAngelis

18 Months of Research: Finding What Actually Differentiates Top Performers

Before building anything, David’s team went looking for the answer. They asked supervisors across 26 states and multiple territories to identify top performers and average performers at each of the five FST levels - people who were roughly equivalent technically, but who differed in ways that supervisors noticed and customers responded to.

Then they ran behavioral interviews. Thirty to ninety minutes each. Recording every conversation. Asking open questions about what behaviors people believed had made them successful in difficult situations, in learning the job, in building relationships with peers and customers. The analysis took a year and a half.

What they found was striking in its consistency. Across geography, across levels, the differentiating behaviors were clear and they changed as technicians moved up. At entry level, credibility and interpersonal skills were the biggest differentiators: new FSTs who could build trust quickly with the strangers they needed to rely on were the ones who stood out. At senior levels, mentoring behaviors and early leadership qualities became the distinguishing factors.

“The competencies that made an FST four a top performer were different from what made an FST one stand out. The expectations changed. The role changed. And so what we were looking for had to change too.” - David DeAngelis

The Program: One Week to Learn, Four Weeks to Practice

The professional competency program at Brightstar now covers 15 core competencies, including self-confidence, customer orientation, adaptability, interpersonal effectiveness, fostering teamwork, and forward thinking. Within each competency there are seven to ten specific behaviors. Technicians work through the program in six-month blocks, focusing on one competency at a time.

The training structure is deliberately different from technical training. One week of learning: understanding what the competency is, why it matters, and what it looks like at the individual contributor level, using examples that field service technicians can actually relate to, not examples written for executives. Then four weeks of practice: each week, technicians are given an exercise to put a specific behavior into action in their everyday job, and they check in with their supervisor to discuss what happened.

The supervisors who were initially unable to articulate why certain technicians stood out found that the program’s language gave them a framework they had always needed. For many of them, the debrief conversations became the most valuable part.

“You can teach behaviors. You just can’t teach them once, observe it, and walk away. You have to reinforce it. You have to build exercises so that people are doing it for six months before moving on. It’s a slow cultural change, not a training event.” - David DeAngelis

The Blind Spot: Adaptability and Commercial Awareness

Asked which competency field service technicians most commonly struggle with, David’s answer is immediate: adaptability — specifically, awareness of their commercial impact. Technicians are often focused on what they understand their job to be: install the terminal, fix the machine, move to the next site. The idea that every interaction is also a brand moment — that the impression they leave with a retailer or a customer’s customer shapes how that customer feels about the business — can genuinely come as a surprise.

Brightstar’s approach has been to share direct evidence. At town halls, they read technicians feedback from RFP evaluations — moments where Brightstar’s investment in this program was cited as a differentiating factor that helped win business. Connecting the behavior to the commercial outcome, and then connecting the commercial outcome back to the technician’s own career and security, is what makes it land.

The Feedback That Made It Worth It

Three and a half years in, the anecdotal evidence is strong. Supervisors report more engagement in team meetings. Technicians are thinking about problems from the customer’s perspective before their own. People are bringing potential solutions, not just problems. The first six months were hard. The next six were better. Each cycle since has built on the last.

But the piece of feedback David returns to most is simpler than any survey. A technician wrote in to the team newsletter to say that an audiobook on self-confidence had changed them forever. That they had always been quiet and reserved. That the program had made them think: maybe I need to voice myself more. Maybe I need to show that I’m here and I want to move up. Maybe I need to start speaking out and showing what I know.

“To get that kind of feedback - to know that you influenced not just your customers and your RFP results, but your people, and that someone said something like ‘it changed my life’ - that made the whole year and a half of putting this together worthwhile.” - David DeAngelis

The Business Case: Build a Blueprint First

David’s advice to any service leader considering this work is direct. Do the due diligence first. Don’t assume you know what behaviors differentiate your top performers, do the behavioral interviews. Talk to supervisors. Talk to technicians at every level. Look for the patterns. Map what you find to a professional competency model.

Without that blueprint, you’re guessing. And the cost of getting it wrong, training that doesn’t resonate, programs that don’t stick, technicians who can’t see the point, is harder to recover from than the cost of doing the research upfront.

The other message: be prepared for the timeline. It took two years to research and pilot. The third year was when the turnaround became visible. Cultural change moves like a ship. Steering it requires patience, communication, and the willingness to show people evidence that it’s working before they’re ready to believe it.

Follow Sarah on LinkedIn for daily insights, behind-the-scenes reflections, and the conversations shaping the future of service.

Connect with Sarah Nicastro on LinkedIn

Follow Future of Field Service on LinkedIn for new articles, podcast episodes, and event updates as they go live.

Follow Future of Field Service

Subscribe to The INSIDER - our monthly newsletter with exclusive content you won’t find anywhere else.

Subscribe to The INSIDER

Join us at Future of Field Service Live London on 24 September. A day of conversation, insight, and community with service leaders across industries. Register here.

Stay Connected

Subscribe to The INSIDER, our exclusive monthly newsletter, and get a first look at what’s new, what’s next, and what’s only shared with our inner circle.

Most Recent

September 2, 2026 | 5 Mins Read

From Crisis to Innovation: How Authenticity and AI Are Reshaping B2B Publishing and Field Service

September 2, 2026 | 5 Mins Read

From Crisis to Innovation: How Authenticity and AI Are Reshaping B2B Publishing and Field Service

Share

UNSCRIPTED | EPISODE 374

Kris Oldland, Managing Director of 1927 Media and former founder of Field Service News, joins Sarah Nicastro for a candid conversation about crisis, rebuilding, what B2B publishing gets wrong, and the platform he’s built to fix it.

Kris Oldland spent thirteen years building Field Service News into one of the most recognised voices in the field service industry. In late 2023 he sold it to Copperberg — not entirely by choice. What followed was a period of processing, rebuilding, and eventually designing something new from first principles. The result is 1927 Media: a B2B publishing platform that challenges almost every assumption the industry has made about how content, readers, and solution providers should relate to each other.

The Sale, the Aftermath, and Letting Go

Kris doesn’t shy away from what happened. Financial pressures outside his control, an accountant who let him down, invoices landing in the wrong columns at exactly the wrong time. The details aren’t the point. The point is what he did with them.

The hardest part, he says, wasn’t the loss itself. It was being 90% of the way back and watching the people around him lose faith — while he knew, with complete clarity, that he was on the right path. That gap between self-knowledge and external perception is something most leaders encounter. Very few talk about it this directly.

“You’ve always tried to do things the right way, and everything goes wrong for a short period. The hardest bit is when you’re 90% of the way there and everyone else around you is losing faith. You’ve gotta dig a little deeper. You’ve gotta tell yourself: you know this. You’re on the right path.” — Kris Oldland

What helped was a podcast — deliberately doing the opposite of what his friend John Lamberton’s show usually does. Instead of a success story, Kris told the real one. It was, he says, one of the most healing things he did.

What He Saw When He Stepped Back

One of the most interesting observations in this episode comes from Kris’s time away from writing and speaking publicly. When you stop talking about the future and just watch the industry, the gap becomes visible.

The players change. The characters — and the conversations — often don’t. Companies are still struggling with preventive maintenance. Still debating whether service should be a cost centre. The distance between the boardroom conversation and the plant floor reality remains, he says, genuinely surprising.

“Sometimes our whole job is talking ahead. And then you get into the real world, and you’re like — people are still struggling with preventative, not predictive. People are still struggling with: how do we move from a cost centre? And I’m just falling off my seat.” — Kris Oldland

On the other side of that same observation: AI. Kris makes the point that the technology itself isn’t new to field service. Scheduling algorithms, Darwinian machine learning models, intelligent triage — the industry has been using AI for years. What’s new is accessibility. And what that accessibility has done, above all, is change visibility.

The B2B Publishing Model Is Broken — And Kris Can Prove It

The core problem Kris spent two months mapping — in a white paper he wrote for himself — is one he saw from the inside: B2B publishing serves two masters, and what you need to do to serve one is often at odds with what you need to do to serve the other.

Solution providers pay to reach readers. Readers exchange their data for content. Everyone in the system knows, at some level, that most of those leads won’t convert — the 3% rule suggests only 3% of your target market are actively buying at any given time. That means 97% of lead generation activity is, as he puts it, fishing. The content may be excellent. The association is real. But the transaction underneath it is a form of smoke and mirrors for everyone involved.

“The idea that the audience is a commodity — that’s how every media outlet is geared. And if you stop treating them like a commodity, and you just say: let’s treat both sides equally — it seems so stupidly simple, but nobody’s doing it.” — Kris Oldland

His solution starts from a different premise: what if both the reader and the solution provider were treated as equal participants? Not seller and target, but two parties who have a genuine interest in finding each other at the right moment.

The Platform: 1927 Media and the 5x5 Model

The 1927 Media platform launches in September with 10 focused titles — five vertical (manufacturing, aviation, built environment, heavy industry, energy and utilities) and five functional (including field service, spare parts, pricing, and logistics). Articles that cover a topic relevant to two titles get a double rewrite — appearing in both, with the primary and secondary focus inverted. The intent is to help readers learn from adjacent industries, naturally and without friction.

178+ articles go live simultaneously at launch. Providers can see the editorial calendar, submit commentary in their own words, and be surfaced to readers whose content behaviour signals they’re at the right stage of their journey — without cold outreach, lead forms, or data extraction.

Underpinning all of it is a content gap analysis tool and AI-powered search that maps where each reader is on their knowledge journey — and surfaces what they need next. The AI does everything except the writing. That, Kris is clear, will always be human.

Authenticity as a Competitive Advantage

The thread running through this entire conversation is authenticity — as a value, as a publishing principle, and increasingly as a strategic advantage. In a world where the majority of professional content is AI-generated and largely indistinguishable, the people and platforms that sound like real human beings with a real point of view stand out.

Kris has always had one speed. He is candid about his failures as readily as his intentions, and the audience he’s built over two decades in this industry trusts him for it. That trust is, ultimately, what 1927 Media is built on.

“In the world of AI slop, authenticity is rising to the floor. I’m quite lucky that authenticity is now rising to it.” — Kris Oldland

Follow Sarah on LinkedIn for daily insights, behind-the-scenes reflections, and the conversations shaping the future of service.

Connect with Sarah Nicastro on LinkedIn

Follow Future of Field Service on LinkedIn for new articles, podcast episodes, and event updates as they go live.

Follow Future of Field Service

Subscribe to The INSIDER - our monthly newsletter with exclusive content you won’t find anywhere else.

Subscribe to The INSIDER

Join us at Future of Field Service Live London on 24 September. A day of conversation, insight, and community with service leaders across industries. Register here.

Stay Connected

Subscribe to The INSIDER, our exclusive monthly newsletter, and get a first look at what’s new, what’s next, and what’s only shared with our inner circle.

Most Recent

August 26, 2026 | 5 Mins Read

The As-a-Service Evolution: Lessons from Kaer on Scaling, Customer Trust and AI

August 26, 2026 | 5 Mins Read

The As-a-Service Evolution: Lessons from Kaer on Scaling, Customer Trust and AI

Share

Standout UNSCRIPTED - Episode 373

Dave Mackerness, Director at Kaer and 2024 Stand Out 50 leader, returns to UNSCRIPTED five years after his first appearance to share what scaling cooling-as-a-service has actually looked like, from regional expansion to managing the customer behavior shift to an AI platform that makes their oldest plant their best.

Kaer has been in the cooling business for 70 years. The company began as a distributor, evolved into a system integrator, moved to performance-based contracts, and committed in 2018 to going all-in on cooling-as-a-service, a model it had been building since 2012. When Dave Mackerness last appeared on UNSCRIPTED in 2021, 80% of the business had transitioned. Today, Kaer operates across India, Malaysia, and Indonesia, with strategic investment from Patrizia and Mitsui backing the next phase of growth.

This conversation picks up where that one left off.

The Sweet Spot and Why Both Small and Large Companies Struggle

One of the most practically useful frameworks Dave shares is his thinking on which companies are well-positioned to offer as-a-service and which face structural barriers on both ends of the size spectrum.

Smaller companies often lack the capital. Cooling-as-a-service is capital-intensive, you build and own the infrastructure rather than sell it. And beyond capital, there is a trust issue: mission-critical infrastructure demands the confidence of large customers, which smaller companies may struggle to earn.

Larger MNCs face the opposite problem. They have the capital and the trust. What they lack is the ability to move. The bigger the company, the more legacy it has in the old model, factories producing hardware, revenue targets tied to box sales, reporting metrics on listed exchanges that shift uncomfortably if the model changes. Many have tried to test as-a-service in a separate business unit, found it hard, and watched investment get redirected back to the core product.

“That’s where I find the sweet spot. A medium-sized business in the middle, one that has the capital, has the trust, but isn’t so big that the legacy holds it back.” - Dave Mackerness

The Complexity-Cost Matrix: Is As-a-Service Right for Your Product?

The conversation that sparked Dave’s thinking on this came from a conference where he was seated next to someone in the crane and lifting industry who wanted to explore servitization. By the end of the conversation, Dave had talked himself out of recommending it.

That experience led him to a framework: plot complexity against cost. Products that are cheap and simple may benefit from as-a-service for customer experience reasons, music streaming being the classic example. Products that are complex and expensive sit in the sweet spot where the provider’s expertise and scale create genuine value the customer cannot replicate themselves. Cooling fits squarely there. So does aircraft maintenance, Rolls-Royce’s Power by the Hour model being the poster child. Cranes do not: the customer manages and operates the equipment themselves, leaving little room for the provider to add value through operations.

The Glass Box: Transparency as a Competitive Advantage

When Kaer started, they operated as a black box. This is my system, you receive the outcome, don’t ask how. Customers didn’t trust it. The model was right but the relationship wasn’t working.

The shift was to build a glass box instead, not giving customers control over how the outcome is achieved, but giving them full visibility. The result was the Kaer Connect app: supply temperatures tracked minute by minute, service reports signed by engineers, compliance documentation available at any time. In the old model, you report to the customer. In the as-a-service model, you make everything available and let them audit at will.

“80% of cooling-as-a-service is managing customers, making sure they’re happy, confident, and that what you’re delivering is auditable. We didn’t realize that at the beginning.” - Dave Mackerness

AI Has Inverted the Performance Curve

Kaer began building its AI optimization platform in 2017. Cooling systems are highly complex to operate, and small adjustments to set points across multiple pieces of equipment can have a significant impact on energy efficiency and carbon emissions.

The unexpected outcome: their oldest installation is now their best-performing one. Not because the hardware has been upgraded, it hasn’t. Because it has accumulated the most data. The models trained on that data now optimize the system better than any engineer could manually. The performance curve, traditionally a decline from day one, has been inverted.

“Your best plant is not the newest one. Your best plant is the one with the most data.” - Dave Mackerness

For mission-critical applications, the challenge is the variability AI can produce. Kaer’s solution was a second layer, a thermodynamic validation layer that checks any proposed action from the machine learning model against physical constraints before allowing it to execute. If the action would affect chilled water supply temperature, it does not happen. The result is a system engineers can trust.

The Energy Paradox and Why It Echoes Cooling’s Own History

Dave draws a direct parallel between how air conditioning was perceived a decade ago and how AI is perceived today. Cooling is vital but energy-intensive. The criticism: it’s good, but the carbon cost is too high. Kaer’s response was to decouple the two, delivering low-carbon cooling as an outcome.

AI faces the same challenge. The output is extraordinary, but energy consumption is increasingly scrutinized. Kaer’s position is unusual: they are using AI to reduce the energy consumption of the cooling systems they operate. The energy their AI platform consumes is, in their model, more than offset by the efficiency gains it drives.

What Comes Next

Dave’s clearest signal of where the model is heading: the conversation has shifted from buildings to portfolios. REITs, developers, and portfolio owners are approaching Kaer about outsourcing cooling across their entire estate, not building by building, but as a strategic decision made at CFO or CEO level. That conversation is in a different category entirely.

The other area of active development: agentic AI. Kaer is building compliance agents that monitor service and maintenance activity across the portfolio, with data going back to 1993. As the data library grows, so does the value of the models built on top of it.

“Once you buy Spotify, you’re not buying CDs anymore.” - Dave Mackerness

Follow Sarah on LinkedIn for daily insights, behind-the-scenes reflections, and the conversations shaping the future of service.

Connect with Sarah Nicastro on LinkedIn

Follow Future of Field Service on LinkedIn for new articles, podcast episodes, and event updates as they go live.

Follow Future of Field Service

Subscribe to The INSIDER - our monthly newsletter with exclusive content you won’t find anywhere else.

Subscribe to The INSIDER

Join us at Future of Field Service Live London on 24 September. A day of conversation, insight, and community with service leaders across industries. Register here.

Stay Connected

Subscribe to The INSIDER, our exclusive monthly newsletter, and get a first look at what’s new, what’s next, and what’s only shared with our inner circle.

Most Recent