The Role of Data in As-a-Service Models

Behdad:
Hello, and welcome to the data for subscriptions podcast, where we discuss and explore how to run better subscription businesses. I’m your host, Behdad Banian. And today I have the pleasure of welcoming Holger Pietzsch to, the show. Holger is the director of business development at Moog Construction, and also leading the digital task force task force at the committee for European construction equipment. Welcome Holger. 

Holger: 
Thank you. Thank you for having me on. 

Behdad:
So today Holger, we want to discuss the role of data when running as a service or subscription businesses. And I’ll just right off the bat, take the opportunity and, mention that we want to talk about the usage data specifically because data can mean a lot of different things. And in the case of Moo Construction, when we say usage data is data that is looking at how machines operate and how machines are utilized. So with that, Holger, let’s talk about your background because you’ve been in the business of running as a service for a long time. Talk us through your experience of running such in different companies, different industries, and what led you then to move construction. 

Holger:
Sure. Yeah. Yeah. Thanks for opening the door here. So, yeah, construction runs a little bit in the family. My dad used to run a few, quarries, in Germany, and then from there eventually through Caterpillar dealers, I worked my way to, to Caterpillar. Caterpillar has been very early bird in in leveraging data. You know, some of it trickled down from the mining industry, which is obviously, in need of highly standardized operations 247, and then it started to trickle down into construction, in terms of telematics. You know, telematics, I would say, around the 2000, became an increasingly standardized feature in a lot of machines. And at this stage, Caterpillar sort of, you know, made the decision to not necessarily leverage data as a new business, but as the CEO then coined it was, hey. Data is an enabler, and it became a key part of optimizing the operating model. That means how can I, design machines better, leveraging data, and now knowing how they’re being used? How can I, support these machines better in terms of knowing where and what kind of parts will fail, in terms of optimizing sales and marketing efforts, towards, the machines in the areas that have higher utilization than others? So I would coin this as my experience on how do you leverage data to optimize a an operating model. That means delivering customer value, better, cheaper, faster, but not necessarily different type of value. And then from there, I moved to a company called, Hexagon, Leica Geosystems. Hexagon, Leica Geosystem, is also all about data, but it’s more around topography, surveying, geospatial data. And, so there are lots of the traditional, surveying instrumentations that you that you know, you know, to measure landscape. These days, you can you can measure that landscape as it is. You can then, define the to be landscape in terms of here’s where I need a road, here’s where I need a slope, here is where I need whatever decline or ditch. And then you can use that data to actually, you know, you could say, control machines, almost program machines to then change the landscape. So now, given that construction customers get paid for moving dirt, you know, now you get very close to their revenue generation, their own revenue generation. And data in that case allowed them to get paid faster. It allows them to prove the quality of the outcome they have achieved with leveraging their data. And, overall, it falls, I would say, within that category now where we’re no longer talking about the operating model of the manufacturer, but, really, the business model in terms of how can, customers create customer value or value and capturing part of that. And then fairly recently, I joined the morgue, which is, I would say, at the cost of a completely new technology in this space, which is electrification of, construction equipment where we’re now, substituting, you know, diesel powered construction equipment with a very heavy c 02 footprint to be run by electrified machines, and battery enabled, and in some places, even, replacing the hydraulics. So their data plays an important role, a, just to make the whole system work. So, you know, they didn’t you need to be very smart in terms of how you actually use the battery. So you optimize its life. You optimize its power. You plan the recharging. And then, also, data plays an important role here in terms of, making sure, you know, that your investments work out and that, you keep the object safe. So very different companies, very different stages of their maturity. All of them somewhat leader in their fields, but very different value propositions and very different data that is required to make these value propositions work. 

Behdad:
You already hinted to, to some extent to why customers, when it comes to Moog, prefer the as a service model over the onetime transactional sales model. Why don’t we elaborate a little bit on it? Why is it so that the customers do prefer this approach? 

Holger:
Yes. Good point. So, I think in the construction, you see, there are three very important factors, you know, that allow customers, to make money. You know? Assuming everybody could build the same road, it is road by road. So then, you know, what is what is important for their own value creation is, on one side, is asset utilization. Right? So these are very heavy, pieces of equipment, and you want to make sure that you basically, identify an underutilized assets very quickly as well as overutilized, you know, because those are basically indicators for a bottleneck. So data around, machine utilization and asset utilization is very important. The second one is there is a potentially very high cost of downtime. So one of these machines break down, potentially whole construction site could be down. And last but not least, you know, you then also want to optimize the timing of investing and divesting. So if you now come to the party with a very, with a value proposition that has a very high acquisition price, even though running and maintaining electrified equipment is extremely low compared to, you know, keeping a diesel engine running, let’s say, but the upfront cost is significant. And because customers do know what they pay upfront, but they don’t know to what degree they’re going to use it as much as planned for as long as they planned with the intensity they planned. And there is still today some concern about, well, what is the residual of a of a battery in residual value of a battery in 10 years. Right? Right. So people in terms of getting their toes wet prefer, I would say, a paper use or a power by the hour model that allows them to kind of, well, get their feet wet. You know? And it’s not just you put a an electrified piece of a construction equipment on a site. You then have to adjust your own operations. Right? You need to see how do I get electricity to the machine. How do I train my own people differently? And so all of these costs and all of these, concerns, obviously, can be are easier to be shared for customers if all of that becomes a pay per use service. So that is a lot of feedback that we got. Obviously, at some point for a customer as well, at some point, they need to decide, okay. Well, how much do I need to learn and how far do I go with this before I switch to a new operating system myself? Yep. And I think at this case for high utilization assets, you know, a pay-per use model might or might not be the right solution. We’re in the very early stages of this, so it’s going to be exciting to see how it maps out. 

Behdad:
It’s interesting with Caterpillar who obviously with the example that you gave your own experience, I’ve been doing this for about 10 years, and 10 years feel like a long time. But if we think about the construction industry, we’re often speaking about companies that have been around for a 100 years, if not longer. So this is you’re absolutely right to say that in the bigger picture, we’re just at the beginning of much more to come. But it’s still very interesting to hear you talk about why customers are choosing the, as a service or a subscription model over. I I do see myself as well. There’s a convergence between technology, such as for example, the electrification and the batteries is a big ex a big factor here, in fact. I also think that there is just a general maturity, towards this kind of a business model as opposed to 10 years ago, if not a little bit longer back where it was fairly new. It was something that you wanted to just test and trial where while we’ve seen many, many industries dipping their toes. I want to ask you, what are the key metrics that you track? 

Holger:
Yeah. Obviously. So we have a few internal operational metrics. We have a few customer metrics, obviously, of course. So, I think one key element for us, given that it’s the source of powering all of the electrified machines, are these energy modules. And so, they’re basically, you know, we think of a battery as what you put in your remote control at home. These days, you know, there’s still some chemistry, but everything that runs around it is like highly sophisticated software that regulates, the machine safety, the machine charging times, the machine, how many life cycles, the temperature, the expected usage. All of those are metrics that are extremely important to optimize, the expected life, the run time, the safety for batteries. So those are areas where today we basically depend on data to a large degree to optimize, to help customers optimize, these investments. You know? So, and then there’s another element that is about, like, optimizing the running of the batteries. And then there are similar elements in terms of, maintaining the batteries, charging, charging the batteries. So it’s a very, very data, intensive business both in terms of, you know, running the product as well as, making sure you help customers themselves adapt their operations to, you know, how do I now run an electric asset on my construction site as compared to running around with a fuel truck, which we’ve done for several generations now. 

Behdad:
Holger, the fact that the usage data that we framed is important in running such business models, I think, is unanimously agreed upon. But in Moog Construction’s perspective, why is this data particularly important for you guys? 

Holger:
Well, I want to name just 2 sound bites here. One is about mission critical. Right? So these machines, you know, they operate in a team. You know? A construction site is like a little bit like a factory without a roof. And, you know, one element going out of whack can have an impact on the entire operation. So very important, from that point of view. And the other one is, obviously, the ability to have this in real time. You know? I mean, you need to be able to have that data available quickly, and also as compared to, what is the history of this particular machine so I can compare the current data as compared to, well, what does that mean in terms of that same asset, what it did in the past? You want to be able to compare your asset compared to other assets in terms of, you know, am I working with an outlier here? Is that a bottleneck in any way, or is there anything different about this piece of equipment as compared to others? So you have the history. You have the comparison, and I think the third one, also very important, is a standard. You know? In order to make profit, I need to have an asset utilization of x percent. Well, you know, how do I do against this target percentage? So data extremely important for our customers in terms of smooth, profitable, and safe operations. 

Behdad:
Do you monetize any of this data? I mean, you hinted to it already in terms of how you provide your different services. But perhaps you can elaborate on how you monetize, the data. 

Holger:
So, yeah, I think when you when you look at the construction industry 30 years ago, right, you would you would buy a piece of equipment, hopefully with somebody who has a very good product support utilization. But then in terms of the risk, the utilization risk, as well as the, the timing of that was totally yours. Right? So if you bought the wrong excavator at the wrong time, well, then you were stuck with an excavator for as long as you could get rid of it. And we’ve seen about, you know, 20, 25 years ago already a change that people, especially for smaller machines, that get used in shorter jobs, some smaller duration, smaller budgets, and a large part of that industry already has moved to a rental business, where now, at least, the time based element I have control over as a customer, I can say, look. You know? I would need this thing between April October because for that period, I feel fairly safe that I have something to do. And so but I don’t have that full lifetime acquisition risk anymore. And then the next level then really moving from eliminating this time based thing is now going down to utilization where I don’t just have, time availability, but I also have the financial ability to adapt my cash flow with my in terms of outlays with my, cash coming in. So if you take that a little bit further, it doesn’t take very long to think like, well, why wouldn’t that adopt also that same, logic to, to, like, a battery energy module where now you’d say, look. I just want to pay this power by the hour. You know? So as I consume the energy, I’ll pay you. And if I don’t consume the energy, I won’t pay you. And you already today in a lot of generators, you know, diesel enabled generators, you know, you have these power by the hour concept. It exists in a lot of standby business models. It hasn’t fully worked its way yet into the construction industry and better enabled assets. But I would think if you had to 2 if you had, and we pride ourselves of still having very, very differentiated products. But if you are in an area now where value propositions from a product perspective start to, commoditize, being able to offer customers a a paper use model makes a significant difference. And anybody who has the data to do this, it would be a game changer, right, with immediate customer adoption. We ourselves see a possibility that a market will go this way. I don’t necessarily think it’s going to be the case for high utilization assets, but, you know, the word is still out there. And, obviously, the cost of non utilization, is another element. It’s a cost that needs to be appropriately shared. You know? With the system we have, we think there’s going to be very high asset utilization, which makes it more interesting for customers to own an asset. But if you start ending up being in applications, or sectors where there are cyclical, that that definitely makes a risk for makes a change in the risk for customers and also their willingness to adopt technologies a lot quicker. No doubt. 

Behdad:
But this is quite interesting and a strong point, Holger, that beyond the traditional product differentiation that the business model itself is kind of growing in importance of being a differentiator on its own. I wanted to just to play this back in terms of going back to the monetization question. So my understanding is that the way that you offer your services today is on some form of subscription. You have, different we can call them SLAs or service level agreements or where you could pay more or less depending on the, let’s say, quality of the service or the uptime that you want. Is that correct? The pay per use that you mentioned a couple of times, is that already something that you are doing or something that you’re embarking on with soon? 

Holger:
Look. We’re I don’t want to call us a stoplight anymore, but we’re still in the early stages of, you know, optimizing product design or distribution and all of the challenges you have as you start. As we scale, I think, you know, this this frequently comes up as a critical customer requirement and a preference, for any supplier. Yeah. The paper use. Exactly. Hey, by the way, I think I think the paper use is definitely something they would like. But, look, they would like it, but they also want to have a reliable product. You know, they also want to have reliable support when things go wrong. And, so it’s one of the many milestones we have on the road. I think in terms of scaling at the later stage, this is going to be it’s going to be very, very difficult to avoid seizing that opportunity. You know? That said, for us at this stage, product, superior product is a is a must. And I think as the industry matures, one of the next steps will really be making more out of our leveraging data for us in terms of our business model, but also for data for customers leveraging their data more. Right? Think about, you know, you now have several electric items running on a construction site. And, look, you might get your charging logistics wrong and one machine runs out of power, and then you don’t pay for it because you don’t consume that power. But the benefit of not paying for that as compared to what you’re losing out by a vending machine that doesn’t operate, doesn’t make sense either. So I think, there is some opportunity here, and we’re very happy to have, some customers in some areas that want to learn along with us to optimize the equation from both sides, which I think that’s probably the best way, you know, to make this happen. 

Behdad:
So let’s shift the feet a little bit and talk about data accuracy now that we’ve elaborate a little bit on the role of data and the monetization that you’re doing upon it. So I mean, data accuracy in general, of course, is extremely important that you, double clicked on it yourself in terms of why it is for more construction. Just to mention a few areas that we typically talk about when we talk about why data accuracy is important is because you can imagine customer satisfaction if, you, you’re billed incorrectly, ability to manage disputes, data transparency. You’ve already mentioned that, in in your in your sharing as both transparency of usage that is we’re speaking about, or revenue leakage if we take it from the, supplier side, if you will. And a common, I would say, industry, learning is that on average, there could be 3 to 8% of revenue leakage. And by that, we mean that there is usage of one’s products, for example, that, we that gets lost in translation data that gets lost in translation. So imagine you’re running a subscription business that has a revenue of a 100,000,000 USD. Well, you’re looking at maybe 3 to 8,000,000, of revenue that is left on the table. These are just some of the examples. There are more, of course. What I wanted to play back to you is, since data is so important, how do you manage usage data at more construction from a system and tool perspective? 

Holger: 
Well, I think there we’re really I mean, for most of our standard operations, you know, we run on SAP as in as in ERP. And I think, you know, it’s very difficult today to run a business as a manufacturer without, you know, solid processes running on solid data. You know? We e have a little bit of an opt out here because we’re developing newer business models in terms of, you know, we’re taking the companies into new areas. And I think to a degree, that also means that we do some things still on spreadsheets, you know, and do a lot of data testing and validation around the product, feedback loops. But as I mentioned before, you know, the goal of a manufacturer is volumes. And, if you have variability in your processes, it just kills you. So and I don’t think you can eliminate variability in processes without data. Now the interesting part is the same applies to customers. So if you get to a point where data is at the source of your value proposition or the actual value proposition and they can see that it allows them to reduce variability, to reduce time, to reduce cost, makes a big difference. To give you an example, in the construction industry, you know, you can spend a long time building a road, you know, which is great. Operationally can be perfect, but you want to get paid as early as possible. You don’t want to get paid after, you know, you an 18 months project on building a road. And a lot of it being paid is a function of you being able to show to the to the end user, right, the owner of the infrastructure, like a government or somebody who needs a warehouse or a car park. You have to show them that you’ve actually done what we were supposed to do within the specifications that were agreed upon. And you don’t want to do this with a lot of people driving around, taking measurements, and testing the quality of infrastructure. You want this to be based upon a a data, stream, a dataset that both parties agreed upon, and that anybody can monitor at any given point of time. And then as you get close to how customers then improve their own work to cash cycle, let’s say, you know, building the road and then getting paid for it, if you can help them, accelerate that cycle in terms of you can give them visibility of what the machines did, what material has been used, with which specification have been accomplished, there’s a big shift in the industry and requires a lot of data to flow, as you said, accurately, you know, because these are multi multimillion dollars of, taxpayer money or other investments, you know, that people need to put on the table to build and run infrastructure. 

Behdad:
So you mentioned that you already have in in when it comes to the management of this user’s data that you’re still running, manual processes and the Excel, which is not necessarily rare. We actually encounter that fairly often. But if we talk about, areas that you’re looking to improve, what would you highlight are the next coming areas in the next, say, 12 months, for example, in the short to medium term? What are important areas that you need to improve within? 

Holger:
Well, I think one thing we want to make sure is that that we have complete, transparency of for customers and for us in how to optimize, the usage of energy on a construction site. Like, how many modules do I need? When is the best time to charge them? You know, where do I charge them? How do I charge them? Who charges them? So I think that is a that is an extremely important area. Obviously, product design and especially product safety is always on the top of the list. You know? And that is not just, you know, ideally data, not just about the product itself, but how people operate with these, assets. You know, there can be some risk in that also. And last but not least, you know, I I would love to see a module where you just really, don’t need to become an expert in construction site electrification, but you have a provider like Mo Construction that that make it easy for you to focus back on moving earth and building infrastructure as compared to becoming an expert on electrification. And so I could very well see then our value proposition and our income streams linked, all linked together to that utilization element. So for me personally, you know, I’m on the commercial side of the business. You know, I hear this a lot. You know, if you are very customer centric, that’s definitely on the top of my list. There are also some financial elements associated with that. You know, you have to mitigate the risk the financial risk of now, basically owning a lot of the assets that you hand out as a paper use service, which, again, the only way to mitigate those risks is data about the usage. And last but not least, then, obviously, we need to design the product so they give us the best possible data, the highest accuracy, as well as the platforms and processes to manage it all. You know? Because just have having it on your phone, is not going to cut the mustard. You know? This this needs to be on a solid platform that can integrate with other platforms as well. 

Behdad:
So we’ve, I think to a large degree, discussed, that they were speaking about multiple sources of data. We’ve talked about the data accuracy and demonetization. One question I have left really, Olga, is we’re speaking sizable volumes of data. And you mentioned before as well that oftentimes we’re speaking about real time data. So what are we talking about? So if you take the construction industry and just generally in the industry perhaps, what volumes of data we’re speaking about on a constant basis here? 

Holger:
Well, I think the construction industry today, you can assume that any piece of equipment over 20 tons, say, less than 7 years old, is probably connected, and generating data multiple times a day or across multiple channels. So that makes it millions of assets. And just in my time, you know, we were we’re collecting over a 1000000000 records a year, 1,000,000,000 records of data on machine temperature, data on idle time, data on utilization. So I think there was an there’s an extreme this this industry is becoming more and more data driven. It is becoming more and more data driven because it’s very cyclical. And, you know, when the economy, the construction economy collapses in a given industry, you want to know about it very, very early because inventories start filling up quickly, and, and factories cannot afford to just build based on guesses. So I think Caterpillar was a very good example. And increasingly, also, we see customers driving a lot of their forecasts, their investments, their own operational processes based upon data.
 

Behdad:
Right. And you mentioned data transparency several times, actually, that, while we’ve been speaking to a large degree, the business model, the monetization of value for customers, the transparency of that data has been mentioned several times. And this is perhaps a good bridge to another question that I have, which is we there’s a factor of data ownership and data integrity here as well, and we have, an upcoming EU data act. Perhaps you want to just, run us through briefly what should we expect? 

Holger:
So this is actually I think this can be changing. You know? This this can be changing, not just the construction industry. Right? The EU Data Act is lateral legislation. It’s going to basically impact the usage of any connected object. Yeah. From your razor to your bicycle to your excavator. Right? And so I think what was very interesting is the EU has, to a degree, has made the idea of data ownership obsolete. As in it’s not really the question of who owns the data. If you are the user of a of a connected object that generates data, you have the right to that data for free. And I think that’s going to be a significant change because it’s going to allow a lot more people now to get data that previously was proprietary, that previously was inaccessible, and the idea of the EU in this particular case that this will feed, a new economy of platforms, of value proposition, and, hopefully, allow them to keep the position in the b to b markets where, without doubt in terms of if you think of b to c and anything that happens with data on the Internet, Europe has lost out. So, I think it’s going to be also a big change for manufacturers and OEMs, in terms of how do I provide this data now. How do I provide it for free? You know, if if you don’t have the platforms and systems in place, then, right, how many people can you possibly hire or send data around to users that ask for it? Yeah. Very, very exciting. And, CC so CC is the organization that represents all of the machine, manufacturers towards Brussels, and they’ve been very involved in finding out, you know, what or do some of these, definitions mean. So there is there is a lot of collaboration here to kind of put this in place smoothly and successfully. But, you know, it’s going to provide a lot of people with a lot of data, and, that will then also very quickly, I think, drive the question of a look, let’s say you have 5 different brands of machines on your job site, you know, and they all give you an idea of battery status. Right? So one says 62%, the other one says 80%, the other one says 50%. What does that even mean? What is what is 50% full for battery mean? And there are probably a 100 different ways to calculate it. But unless there is some standardization that will take place, that data, however, very interesting is for you to have for 1 machine, it needs to be usable for fleets and for sites. So very excited to see this, taking place. And I think for folks like yourself, it it’s also going to, like, open a lot of doors and needs that none of us can even imagine at this stage. 

Behdad:
Yeah. We had an interesting, speaker earlier in the year, Pavana Kumar, who, works out of an agency in the US. The FTC in the US has been driving quite a lot of initiatives around data integrity and making sure that there’s transparency, auditability. Of course, here in, Europe, we’ve had GDPR. But I think that the EU data act will be quite a significant, let’s say, pathway to unlocking both value of course, and opportunities for manufacturers and constructors. However, I also think that it’s quite important to understand that it it’s easy to get overwhelmed when it comes to requirements of data transparency, auditability to make sure that everything is according to standard. This, I do believe it deserves its own topic. I think it’s super exciting as you’re saying, but I can also just understand that as businesses, it’s very easy to be overwhelmed when these questions and requirements come in. So it’s a it’s a topic to be taken extremely seriously, I would say. 

Holger:
It’s yeah. Yeah. It’s very very easy to get excited in the candy shop. Yep. Right? But once as an organization, you make, the switch towards, hey. I now, make some of my operational, commercial, or financial processes dependent on data Yep. That I get from a supplier. You want to know what you’re doing before, before you kind of switch over and you want to my advice be don’t try this on your own, match up with people that know what they do. Yeah. 

Behdad:
I think that’s a that’s a good advice. There , of course, are several good ways of doing this, but you, it’s safe to say you probably want to avoid, avoiding manual, efforts and Excel shops and so on because it could easily entangle you in a space where it’s going to be, difficult to sort out. But with that, Holger, thank you very much for also, this piece about the upcoming EU data. As a final, piece in the wrap up, for everybody who has been listening to us now and have been following your experience, what are the 2, 3 advices you want to leave us with if one is looking to accelerate and optimize one’s as a service business? What should we do tomorrow? 

Holger:
Well, I think I go back maybe to some of the known recipes for starting a business and scaling a business successfully. Right? And that’s one is addressing that pillar of feasibility. You know? Can I can I actually build what customers want me to have, and can I build this reliably, consistently, and safely? And so you need to look at the data that it requires you to do that and build your assumptions and then and leverage data even in simulations. You know? Not everything needs to be in the real world, but, you know, feasibility is a is a key element. Right? And, obviously, a lot of Moog’s history is in in building motion control for airplanes, so simulation is absolutely key, right, for some of us. So look at how what data do I need to make sure that the product I design, deliver what they’re supposed to do. The second one is viability, and that means, you know, like, we talk about some of these models that we have before, but make sure you have the usage data and all of these other data’s that allow you to understand, you know, how commercially successful it is. I think adoption rates will be significantly higher, but so will be the cost of anything you carry on your balance sheet. And the last one, probably most importantly, is that desirability. And here, I think, the sky is the limit to a degree what you can do with data to make sure that customers can very easily perceive value out of your products because they fit their operations better, because they allow themselves to get, cash faster, more reliably. And so I think both will be the 3rd area where I would where I would really look at my data in terms of getting usage data about connected objects. We’re going back to feasibility, viability, desirability. Nothing new but, prioritizing which datasets allow you to get there quicker and better. That’s really a session worth having with your management team as you go into some of these data enabled service models. 

Behdad:
Holger, thank you very much, for your time. I’ve appreciated the conversation. Also, thank you for everybody tuning in today. 

Holger:
Thank you.