IFS customers shed light on the core role of data in servitization and XaaS
- Manufacturing industry started building servitization business models even before the software industry discovered SaaS and XaaS. Connected data is central to both.
The Software-as-a-Service (SaaS) industry has pioneered the as-a-service model of doing business, characterized by subscription-based recurring revenue, ongoing engagement with customers, and continuous monitoring of usage to help tune the product for successful customer outcomes. Over the years, I've written extensively about this model, calling it XaaS, or Everything-as-a-Service, to reflect its broadening adoption across every industry. But this model is also known as servitization and it wasn't invented by the software industry. After listening to manufacturing companies speaking at the recent IFS Unleashed conference, last month I wrote:
These IFS customers haven't followed that SaaS-derived XaaS template. They've arrived at digitally enabled servitization by their own, independent route. Rolls-Royce first introduced its XaaS model of selling what it calls 'power as a service' in 2002, at a time when the SaaS industry was so nascent that the term 'SaaS' hadn't even been invented yet. Jotun decided to start selling efficient ship hulls — as an outcome of applying its paint — in 2012.
So how did these other industries get there first? The unifying thread is their use of data gathered about how customers use their products. Whereas SaaS vendors essentially stumbled upon the benefits of being able to monitor usage as a by-product of delivering their products digitally from the cloud, these manufacturers actively set out to collect usage data with the express intent of transitioning their physical products to a servitization business model.
Power by the hour at Rolls-Royce
Engine maker Rolls-Royce adopted servitization two decades ago as a competitive play to build market share for its jet engines from a lowly third place at the time. About six percent of the typical operating cost of an airline is the cost of maintaining and overhauling the jet engines on its aircraft. A big part of that cost is the regular overhaul, when the engines are removed and transferred back to the manufacturer after every five million miles or so for a thorough check-up and refit. But each overhaul of these massively complex and sophisticated engines has an unpredictable cost. The insight that Rolls-Royce had was that the manufacturer could smooth out this cost for its customers. Nick Ward, VP Digital Systems at Rolls-Royce explains:
We're taking the jagged cost curve away from our customers, and we're taking that onto ourselves. So the customers get to pay a flat rate — 'power by the hour' is the other phrase — a certain number of dollars per flight. We take on the uncertainty of the cost, and we then manage that. That's our service concept.
That's worked fantastically well. I think in the twenty years since we introduced that, we've certainly grown enormously. A high 90% of all our engines are provided on these contracts, and we are a strong number two on our market now. In our particular market for wide-bodied aircraft, around 50% of those.
Data is key to the success of this model. The company does 1,500 overhauls every year, and each one must be meticulously planned, with detailed information about how the engine has performed and which components may need to be refreshed or replaced. Over time, the volume and detail of data collected from each engine in flight has expanded massively. As Stuart Hughes, CIO at Rolls-Royce, told diginomica last year:
In the past, we would have 30 sensors, and they would capture data five times in the flight at key points. Whereas, the latest generation of engines going into service now capture hundreds of data points every second. So we've moved from [a] floppy disk to half a gig of data per engine flight.
The company now brands that connected data stream as the Blue Data Thread, joining up data from customers, their engines, their maintenance systems and the manufacturer's own systems, where it can model a digital twin of each individual engine. It even maps the environment each engine has flown through to help plan the best maintenance window and help avoid unplanned maintenance incidents, when a plane might be unexpectedly grounded and cause costly disruption to flight schedules. A Databricks lakehouse running on Microsoft Azure powers the analysis and machine learning. Ward comments:
As I explained, we're taking that uncertainty from the customers and into Rolls-Royce. The only way that you can do that and survive as an organization is if you're going to understand the risk, if you can manage the uncertainty. You can only manage uncertainty if you have the information and the ability to understand that information and plot your course to make your operations as efficient and as lean and as predictable as you can.
Data technology is absolutely core to doing that. We could not do it without that. And we've gone from first generations of digital capability, and as we go on, we're obviously talking cloud, AI and all of those things. So we move to those as we need to.
A mother lode of servitization
Paint manufacturer Jotun's move into servitization in 2012 was prompted by a desire to secure its position as a market leader in anti-fouling paints for ship hulls, and was also enabled by data collection. The company realized that its customers were buying its paints to achieve the outcome of more efficient hulls, and so it decided that was what it would sell. It created a proposition that provided the underwater paint, technical service at the shipyard, and equipment to measure hull efficiency, which can report data as frequently as every 15 seconds. The selling point for customers was a guarantee of lower fuel consumption, and now also encompasses lower emissions. In recent years, the service has added a robotic hull scraper that constantly cleans the ship's hull underwater. Morten Fon, CEO of Jotun, comments:
From launch in 2012, we started to collect data from vessels ... to see how the efficiency of the hull develops. So we can help the client — the customer that owns the ship, or the customer that operates the ship — to save fuel, so they can optimize their operation and help them with huge amounts of data.
The IFS customer base is becoming a mother lode of such stories. Iconic packaging brand Tetra Pak is delivering outcomes such as line performance, waste reduction, cost reduction, and even emissions reduction under as-a-service contracts to customers who use its production line equipment. The next step is full servitization, as Sasha Ilyukin, SVP Customer Service Operations at Tetra Pak, explains:
Full servitization is where we sell all of our services, everything that Tetra Pak has to offer — equipment, materials, services — that are priced on the basis of a liter of package produced or packages output.
Other examples include tire production automation provider Cimcorp, which plans to use data collected from IoT sensors to help customers improve their equipment and operations. In a session during last year's Mindfuel series of virtual discussions, Klaus Glatz, Chief Digital Officer at industrial equipment manufacturer Andritz, described how it has sold a process optimization software solution on a shared revenue model to its pulp mill customers. This puts Andritz's existing domain knowledge together with live data from sensors in the mill to help operators make the right process decisions to ensure quality output — in part as a response to the loss of experienced operators through retirement. Andritz has sold the solution to more than 70 customers, using a shared revenue model to encourage take-up. He elaborates:
What we understood is that, if we as a manufacturing company are offering now a software platform, it's hard to convince the customer that we're able to do that. And we said, 'Okay, then give it a try. I guess you will get the solution for six months for free. And after six months, if we can guarantee you or we can prove that the output has improved, let's do a shared revenue model because it's beneficial for you. But then it's also beneficial for us. And if after six months, we cannot prove that it's now better than it was before, then I guess it was a nice play.' This lowered the entry barrier significantly because the customers understood they don't need to invest anything, and if it's working, then it's beneficial for both sides.
At an earlier event last year, Cubic Transportation Systems gave an example from what the transportation industry calls the Mobility-as-a-Service (MaaS) sector. The company spoke about delivering its maintenance services for Transport for London's contactless ticket gates and scanners on a results-based contract. Mike Gosling, IT Service Platforms Manager at Cubic, says:
We no longer supply ten gates here and four machines there. We supply hours of retail and hours of validation across the whole board — and those hours of retail and validation, we have to maintain to a service level.
This totally changed the way that we had to do our service, it totally changed the way that we had to manage our business. We couldn't just stick people at every place, because that would have blown the budget. There are devices all over the place. So to try to maintain these really high levels of service in place, we had to get smarter in the way we do it.
Gosling's final comment illustrates another theme that's common to both servitization and XaaS. This is much more than a move to a different business model or pricing mechanism. It requires a complete change in mindset and culture across the organization to ensure that processes and performance align with the new outcome-based goals. Having the right data, and interpreting it appropriately, is fundamental to getting this right.
These asset-centric businesses in the IFS customer base have come to servitization by instrumenting customer usage. They have adopted various types of proposition, not all of which have meant turning the core product itself into a service. It's interesting, however, that many of those who started out with an offering that bundles enhancing services around the core product appear to be on a long-term path towards what Tetra Pak's Ilyukin calls "full servitization" of both the product and surrounding services to create a pay-by-results proposition.
My conclusion is that it is the shared thread of connected data that is at the heart of this model, whether you choose to call it servitization or XaaS, and the trend is far larger than what has emerged from the SaaS industry. Indeed, the example of Rolls-Royce shows that the phenomenon of data-driven servitization may even predate the advent of SaaS, although the first examples of computing delivered as a service date back to the very first application service providers in 1998. Nevertheless, it's clear that manufacturing industry has been on this path at least as long as the software industry, which got a 'leg up' from cloud computing but in many ways has been playing catch-up in learning to focus on customer outcomes.
That focus has long been clear to manufacturers, as Professor Tim Baines, Executive Director of The Advanced Services Group at Aston Business School, told me in a discussion of servitization back in 2016:
That shift really does move away from, 'It's simply about product and the ownership of a product.' It moves much more to, 'What it is that we're trying to do with things?' And that moves it much more to a conversation about services. With the services, you're buying an outcome. You're buying a process, rather than simply buying a thing which you then want to figure out what to do with. We're moving to this world.
While I think there may be some things that manufacturers could learn from the software industry's experience of XaaS, in particular around the automation of subscription management and instrumentation of customer success, I suspect there's much more that the software industry can learn from manufacturing's journey to servitization, especially around the shift to outcome-based pricing. This is a topic that I plan to explore in future articles.