From Engineering License Data to Business Value

[00:00] Branislav Potoček: Before we go forward, I would just like to see how many people are actually dealing with engineering licenses in the audience since you came to the presentation. One, two, three. I’m like an oil and gas guy, I’m going to talk oil and gas. I’m automotive. Okay, so I guess we can actually see ourselves there. So from engineering license data to business value, this is sort of the plan here and Dave is the practitioner. The platform is to kind of talk about the differences and similarities between the regular SAM, between what is being done in engineering, and how in the end it ties into the most scary thing that we see today in the world. But I won’t bore you too much with that. That’s just one slide that I have towards there at the end.

[00:51] Branislav Potoček: So the agenda is the practitioner, the platform. Where do we see this go next and how it actually aligns and why it shouldn’t be the problem. It shouldn’t be too scary. So the way I see it, I did it all: I did SAM, I did engineering and I did both of them at the same time. And now I’m trying to do the AI tokens as well. And with standard SAM, it’s always, to sum it up at the high level, deployment versus entitlement, right? What you own versus what you deploy. I mean, the deployment should never overtake it. Entitlement is very complex. Engineering is all about metering the usage data. And I’ve been sitting with people from Deloitte, people from KPMG trying to sell us a unified holistic SAM solution and completely missing that all of engineering is completely different.

[01:42] Branislav Potoček: Completely, completely different. It’s something else. There are so many license models, so many different metrics, and they will elaborate on that. And at the end, like I said, one slide about the token economics. So Dave, I guess you need to introduce yourself first.

[01:54] David Boyle: Sure. David Boyle. I’ve been a practitioner and a user of ServiceNow in a couple of large oil and gas companies. So that’s ConocoPhillips 20 years, both in Europe and in the US. And lastly, I was with BP for five years and was charged with portfolio management for a large, large suite of applications. So using ServiceNow for the small part of asset management that was applicable for engineering applications. But I was using ServiceNow to a large extent, of course, for ticketing. Because I also did the specific applications for those specialist parts of the business as well. So first of all, what is it that’s different about engineering applications versus classic software asset management or mainstream software asset management?

[02:51] David Boyle: And it depends what you mean by engineering. So software for engineers, or in my case, I’ve worked. So I’m actually a geologist. So I’ve worked with subsurface interpretation systems used for oil and gas explorations. And they’re very expensive, very specific. So really, we’re talking about software that is very specific to your industry. We have practitioners who are driving value of your company with a specific IP. They’re specific scientists, rather than generalists. So those two portfolios are dealt with very differently. So, so first of all, I put this in two parts. There’s technological reasons why engineering or scientific applications dealt with differently.

[03:41] David Boyle: But there’s also a lot of human organizational differences that have a certain inertia between the two. So when it comes to the technological differences, there is no compliance issue with engineering applications. Typically, they will be licensed with a licensed server, and you’ll have concurrent licensing. The piece of software that I’ve dealt with the most is from Schlumberger, SLB: it’s Petrel. It’s a big petro-technical platform. If you’d gone back 10, 15 years, it probably would have been Halliburton. You know, these guys, big, big, very large suites that are delivered to oil and gas companies. So you can install them as many times as you like. They can be on workstations, they can be on HPC environments, but they’re managed with concurrency. So there is no how many licenses are installed, it’s how many licenses you own and how many can deliver at any given time.

[04:39] David Boyle: So that makes the, and also the versioning, the versioning isn’t really so relevant by that. That’s very specific for what you want to be doing at any given time. Having old versions that sit around is often perfectly valid because it’s more scientifically more valid. So that’s fine too. But the kind of asset management systems that you use for enterprise applications don’t really apply. And most asset management systems, and I’ll, to a large extent, ServiceNow, is guilty of this as well, is that you tend to look at applications as being on a package or a deployment level. Whereas in the engineering and scientific world, it’s about the breakdown, the modular suite. So you have a basic license, and then you have a huge amount of bespoke application components that go to make up that.

[05:31] David Boyle: And your asset base is handled very differently from that than it is for enterprise service. That’s one of the key things. So those are kind of, you know, the, and I remember when I, we did the ServiceNow implementation. They were moving away from BMC Remedy. And it was absolutely great for ticketing knowledge management. It was going to be really, really good. But it didn’t really handle the number of weird and wonderful applications that we had. We had to deal with those separately. We could support them through ServiceNow, but we couldn’t actually get an idea of what our asset base was. So they’re all kind of technological reasons. There’s also a lot of organizational differences between applications I work with.

[06:21] David Boyle: And when I started the oil and gas industry, it was just moving into the digital age in terms of seismic interpretation and reservoir simulation. We were moving away from sheets of paper with seismic on the draw, and it was going into mainframes. So it was a world dominated by not Linux, but Unix, but also Silicon Graphics. So I worked Silicon Graphics system, but at the same time in parallel, you had the PC world, which was the normal world. So the two worlds never used to meet. And so you’d have specialist teams running HPC specialist teams with the data, specialist teams supporting the application. Also practitioners in the subject. So the two never really met. And take something like HPC organizations, Shell, BP, ExxonMobil, all these big companies, they have their own organizations that manage them and look after them that are separate from their kind of mainstream IT.

[07:21] David Boyle: And it’s only in recently that with this, the Schlumberger application Petrel runs really well on PCs now, primarily driven by games technology, which brought that price, price of that down. So it’s taken a long time to bring those two worlds back together again. And so managing the usage of these applications is really critical. So we look at usage management because we want to know who’s checking in, from where are they checking out? But most importantly, what is the modular breakdown that they’re using? Because when we think what our kind of procurement lifecycle is, or our renewal lifecycle is, and that’s where the two worlds of mainstream SAM, to use that term, engineering SAM, tend to, only to pay our bills, we all need to have our monthly, our yearly cycle.

[08:14] David Boyle: But I think how it’s done is quite different. So when I was in my portfolio management seat in BP, we would, the process that we would go through would be to, you need to be, before the bill comes in, you need to be looking at your usage profile, who is using what application when and for what. And often the decision makers that you want to approach aren’t the ones that you think are in a senior management position. They can actually be quite far down the organizational ladder. They are the experts that you need to go to, to look at your usage patterns that you then go forward to make a decision about what you’re going to renew or not. And I’m looking at my slide here, the title usage intelligence drives real portfolio decisions.

[09:07] David Boyle: And what that means is it’s not simply about optimization, saving licenses. Clearly, that’s very much part of it, but it’s, are you using the right tool for the right job? For example, if you, if you’re in your own business, if you’re working Gulf of Mexico, or you’re working North Sea, or the Far East, deep sea, or hugely complex environments, complex drilling organizations, it’s very different from onshore U.S. where you’ve got unconventional plays in Oklahoma or places like that where you’re just punching lots of wells. And I’m not saying that’s easy, but from a geological perspective it’s different. So you actually don’t want the same tool for the same roles necessarily. You want the right tool for those roles.

[10:01] David Boyle: And the difference could be, for geophysicists working Gulf of Mexico, they probably run $200,000 to $300,000 per user; for an onshore U.S. person, it’s probably 10 to 20,000 pounds. So the costs involved are much, much higher for those people. So you’re going through that process. And if you have a change in the business as well, you’re also looking at those usage patterns. And change in the business could be anything, divestment, an acquisition, a joint venture, a joint venture with a different fiscal regime. It’s very relevant now when you’re going between oil and gas exploration to renewables, the same company is doing it, taxation regimes.

[10:51] David Boyle: So you want to know exactly what the usage is and what you’re doing. So once you go through that process, you can then take your renewal of what you want to do to the vendor. And it’s quite common. You might think the vendors will be resistant and that they don’t like too much usage analysis. But it’s better for them to be right sized, have the correct modular configuration for what they want to achieve than to have the wrong and having too many denials. And not enough of one toolset, too many of another. So that’s extremely important, as well as usage in terms of check in and check out, the other really important thing we look at is actual idle time. So that’s where people are taking licenses of an application, but they’re not necessarily actually using it. So it’s quite important to know that.

[11:48] David Boyle: And it might seem, it might seem that it’s bad behavior for users to take an application they might use it, but in some situations, real-time drilling operations, for example, is one where it’s kind of legit to have to, if it takes an hour, an hour or so to load an application to get all this stuff, you know, then you have to go to sleep and get up between the morning to check it. It’s quite legitimate to leave that open. And having the technology to harvest the licenses and put them back into a pool is really valuable when managing this kind of high value portfolio. And when it comes to that, you might think that users don’t take my licenses away. That’s a terrible thing, but in reality, people have a lot of freedom to do what they want to do. And in fact, when you’re dealing with these users, I’ve dealt with them over the years and they’re often very intellectual experts in the field of one and only people who know what they’re doing. If you can go to them with good information about what they’re using or what their peers are using, they’re much less defensive because it’s certainly been the case in the traditional world of science, engineering. I’ve got my HPC. I’ve got my Linux cluster.

[13:04] David Boyle: They’re like, don’t go near it. It’s mine. You can’t analyze it. It’s a real thing. So you have to get the buy-in with the good data and understanding what it is. It’s all about the right sizing. It’s all about weird and wonderful licensing techniques as well. I mean, we all love to talk about tokens, tokenization systems are all apparently very new, but in engineering space, tokens have been around for years and years, using either ordinary token-based systems or consumption token-based systems are very normal, strange licensing system involving dongles and only being able to use an application within 500 meters of a location are very common. And because we deal in the strange world of behemoth companies like Halliburton, but also one-man operations that are, they are where they are because they are experts in their fields and big companies rely on them for particular intellectual properties.

[14:13] David Boyle: And so there it’s more about understanding the criticality where that tool fits in rather than its volume. So you might have usage patterns with applications like that that have very low usage. Apparently people, why would I need that? But that criticality is absolutely huge. Reserves, you know, all the big oil and gas companies report reserves to the US government the week before you report your reserves, it’s like that’s unbelievably stressful and getting that job right and no amount of compliance or auditing compares to the stress of that kind of exposure. So criticality versus volume is a very important thing.

[15:04] David Boyle: Very important. So there’s some numbers there about the kind of scale of money that you can save that compared to what you need to spend if you manage these portfolios properly. And with that, I’ll probably hand over to Branislav for more of that kind of because I’m the kind of why, and he’s the kind of how: the platform.

[15:25] Branislav Potoček: Yeah, I think that’s the best way to put it. So you were the what is it and then why it’s being done. And I’m going to talk from the platform perspective, even though we are actually interchangeable. He was oil and gas, I was automotive, both customers, and now both of us are with the vendor. And it’s funny, it’s funny how we like to think we’re so special and our industry is so unique. But the similarity of challenge between automotive and oil and gas is actually it’s practically the same from a little bit of a different angle, but it’s almost the same. So I’m Branislav, I’m not going to go back to the slide that talks about me. I did SAM and SAM for engineering for I think at this point, 12 years, implemented a SAM solution from the ground up, did ServiceNow SAM Pro back then, and then did engineering. And now I’m very heavily invested in AI. So let’s talk a little bit on the how, how we actually do what they’ve talked about. And it all starts at actually being able to get the data from the source, which is actually the first part on my slide, connected to license servers connected to the machines. This is where the data resides overall, maybe an HPC cluster thrown here and there, but that’s what is that’s where we get the data.

[16:36] Branislav Potoček: Now we want to also capture active use. This is one of the special features of the engineering licensing world that okay, so an engineer may have been checking the license, was he actually working or was he sitting outside having a coffee or smoking, or maybe he just went home for a small vacation, the license is still being checked out. When we talk about M365, which is $25 a month, not that big of a problem, even on the bigger scale, it could be when we talk about licensing that can be used by someone else and it costs $100,000. I think the peak that I’ve seen was $500,000 per license per year. Those are different numbers. Somebody taking the time for 48 hours using the license and preventing someone else. That’s already something that you can see on the budget. That’s something that actually is a problem. Now, we have the integration to SAM Pro, which allows us to actually normalize the data, bring some attributes into it, and ultimately provide a holistic view to the customers. I always talk about it’s like an X and Y scale. On the left side, I have technical people. On the right side, I have the business people. On the Y axis, I have the people who are the engineers and on the top, I have the CEO. The more I move to the business side and towards the CEO in the hierarchy, the more I’m going to be interested in what SAM Pro with the Open iT integration actually has to offer. I’ve seen more high level of the data. The engineering license server manager will probably actually still use the other interface, which contains a lot more details, but I bring from practical implementation. The managers, it’s like, again, working in the engineering world, a higher manager probably is not actually doing the drawing.

[18:13] Branislav Potoček: He’s just reviewing it. He may not even need the license. He may be good with just an AutoCAD reader. So how does the data flow? Well, first, and we already established that we need to collect it from where it resides. Now, in the engineering world, when you look at the normalization, you’ll get a license file that has 200 features. Nobody’s interested in reading and checking the usage of 200 features, especially not somebody from procurement or from finance who wants to see the numbers. They are because especially, again, engineering world is special. They’ll get a quote from Ansys and there’s five products, but the license file actually contains 200 sort of products in quotes, how to map it. So that’s where the normalization comes in. Now we would like to attribute some factors to it, right? So assign it, align it with teams, projects, cost centers, group based, whatever the group is in your organization. It differs.

[19:07] Branislav Potoček: I’ve seen people cross charge. I’ve seen organizations cross charge from IT to engineering, just the whole cost as one charge. And you deal with it the way you want. We have customers where they do this actually very in depth and in detail. And they look at not just group of individual user, how much time they spend using these licenses, how much time maybe they just actively spend using these licenses. And based on that, do the cross charge from the person who’s paying for all of it to the people who are actually using. Now, the second-to-last step is publishing. So we publish the data, right? Whether it’s in the Open iT UI or you see it in the SAM Pro dashboard, that’s where you see what is happening. I have the last one in red and it’s not because these colors are Open iT colors, but I wanted to really stress it out. The hardest part comes at the end. And this is where most people or most organizations actually struggle. So we see what is happening. What do we do? Who can tell us what kind of decisions can we now take? And it actually perfectly mirrors what happens in the regular SAM, right? You get an implementer like SoftwareOne, for example, and their most premium package will actually give you the fact that they not only bring a tool and deploy and set it up and give you the reports, but they also tell you, now this is what you can do. And we stand by our words. This is where you need more. This is where you can probably have less. So the whole ecosystem is very straightforward. So we have the ServiceNow SAM Pro platform and probably everybody here is familiar with that since we are at Nexus for Asset Management, Open iT brings the specialized usage, the specialized license metric into this pool and together allows you sort of a holistic view to how to reconcile all this consumption, all this data. And again, it’s in red. Even together, you still need an expert that will be able to interpret what’s there and allow you to take some decisions. Now, this one is my favorite one because it shows why ServiceNow is a big thing.

[21:11] Branislav Potoček: And that was there when it really started as a ticketing system. And it’s amazing what came out of a ticketing system. Now it’s an ecosystem that covers basically every angle in information technology and IT in practically every medium to big organization in the world. And if they’re not there, I’m sure that they will be there very soon. And that’s because so we have the models, we bring the data. But what then if nothing happens with it, it was just probably wasted time. So ServiceNow ecosystem and especially SAM Pro allows people to go and see it in their dashboards. You can actually take the dashboard from SAM Pro and actually see it in a completely different ServiceNow application. You can use the workflows and automation, create your own flow that fits your organization. This is what I really love about ServiceNow world. You can actually mold it for it to become your own, right? They don’t really strictly tell you this is how you have to do it.

[22:07] Branislav Potoček: So if you need a ticket because a license is reaching a denial rate of XYZ, you can have a workflow that actually automates it and tells the people responsible. You have a renewal coming and maybe the renewal is tied to the license expiration. Maybe it’s not. We’ve seen so many times that the expiration of the license is completely outside of the expiration of the contract. Suddenly there’s 1000 engineers who come in the morning to the office that they cannot work. Again, it can be very easily automated with the ServiceNow ecosystem. And I mean, back end of it all is the CMDB and the cost context, which is something that we have customers that do it. And we have customers that would really, really like to do it because just the usage data on its own only tells you a part of the story. We need to know how much you’re actually paying and how to tie that cost to usage.

[22:59] Branislav Potoček: This is one of my last ones, and I wanted to talk a little bit about it. And David already kind of hinted that everything is about the token stuff, right? So we can talk to a CEO who doesn’t even know that there’s the Dassault Systèmes or Siemens inside the organization. He doesn’t care. He knows there’s Microsoft, there’s SAP, the big players. That’s where the money really is. The money’s in the engineering. But now the tokens are being shown in our faces from all sides. And I understand it may be a little bit scary. But in principle, the token is just a currency, like David already actually mentioned that. And it’s a currency, whether you pay for AutoCAD for 24 hours to be able to use it with your token for X enterprise, or you pay to Anthropic to get X million of tokens for a license model that helps you automate your work, it’s the same principle. It’s just a matter of currency. I’m listing some of the most favorite of mine in the engineering world.

[23:52] Branislav Potoček: Autodesk, always a pioneer in kind of figuring out how to make customers pay. Autodesk is the pioneer in licensing and engineering. They always figure out a way how to make the customers pay more. And they’ve been trying and you hear it from so many angles. Move it from the concurrent usage. Just get the license for every part, right? It’s going to make more sense. You won’t have to worry about it. And then actually it was the Flex token wasn’t even there in the beginning. The tokens were only part of the Autodesk ecosystem for the enterprise customers, but based on the push from customers like all of us, they had to introduce some form of currency. Again, if you think about it, so an AutoCAD user that only works for two hours a month, he’s actually much better off with the tokens because you just pay as you go.

[24:43] Branislav Potoček: And he doesn’t need a subscription for the whole month, not to mention the subscription for a whole year. At the same time, this is actually the same case with all the AI that is being now used on the side of, let’s say, regular users who use that application. You go to the Gemini web page or you use Claude Desktop. If this is what you do and you don’t code, you don’t automate, you don’t connect to the APIs, you don’t really use that inference thing. It’s the same principle. Somebody who only does 10 prompts a month, very straightforward, you can just use the pay as use and use the token. Somebody will be better with the subscription, especially if they work every day and maybe they are real heavy hitters. And we were in San Diego at the FinOps X in June, and there was a story of a company where one mistake over weekend cost $250,000, the token consumption.

[25:34] Branislav Potoček: So I kind of wanted to bring it to the audience a little bit. The metric, the currency is not new. It’s been for engineering old timers. It’s exactly the same thing that’s been used for many, many, many years now. And even though there may not be solutions yet on the market, holistic solutions like what we present for engineering today, they’re coming. Nobody has to worry too much. Are we good on time? I think we’re good. I think we’re good on time. So this is just, I don’t want to talk too deep about this slide, but it kind of compares how the engineering and the AI spend, kind of mirrors each other. And again, it’s something that we’ve been used to for so long. So it’s not too much of a worry. So whether you look at the license units in the pool or token budgets for a team that you can already actually do with Anthropic or the ChatGPT application or through the Entra administrator for Microsoft Copilot, it is the same thing.

[26:30] Branislav Potoček: In engineering, we’ve always looked at, let’s harvest those idle licenses. So if somebody’s really actually sitting in the cafeteria for three hours and I have a seat that costs me 100,000 a year, maybe David can work with it while I have a smoke for 30 minutes outside. In the AI, again, it’s the same principle. So we’ve all seen probably the horrific stories of an agent going loose and repeatedly being stuck in a loop and just burning, burning, burning those tokens over the weekend. I think that’s actually the case from San Diego 250,000 over weekend. Nice bill from Anthropic there. Overall, as you can see, it’s the same thing with a different coat of paint. And I don’t think we should be worried too much. But at the same time, I like to say we have only a finite amount of electricity on this planet and unless we actually discover the fusion and can make it work, let’s make sure we don’t burn all of it.

[27:28] David Boyle: Another token system. AspenTech SLM, which again is a completely different model and it further upsets traditional SAM. So rather than having tokens that you would consume, like you would burn them and go away, you buy, say, a lot of a thousand tokens. So that’s what you own. However, you can use them on any tool in the suite and that can be anywhere in that case, anything from geophysics to reservoir engineering. So what you get is, when somebody loads up one particular tool, takes a hundred tokens, somebody else loads another one, another one, two hundred and fifty. So you’ll have a mix of users until it goes nine, nine, nine, then thousand, bang, that’s it, denied, the last person denied. Now the software company love it because they can feed the users whatever software they want, oh great, brilliant, this is a playground of software.

[28:23] David Boyle: From a traditional SAM perspective, it’s a nightmare because if you do any kind of discovery, what are these tools? Who’s paying for these tools? Because the vendor’s pushing them, the users are using them and installing them and it’s all across the board, but it’s controlled by this weird finite token model. So that’s a whole other, so it’s back to, we’ve seen all these weird how to plan for them, how to look at them, it’s something really familiar.

[28:55] Branislav Potoček: Maybe one of the organizations will become big enough that you will just buy their tokens and then you’ll license all the software on the planet. One or two years, who knows, maybe we’ll get there. So I’m not really too much into these numbers, but again, like a comparison of the principles. So for the individual unit, I mean in the SAM, that’s it, that installation. And in the engineering, we had the concurrent license for so long and the tokens are now coming for the AI and reused from concepts like Autodesk Flex, or if you go further back, IBM is using tokens for some of their license applications, probably like two decades ago. Although I will say something that in the engineering, it’s slowly but steadily, they’re trying to move away from the concurrency, right? Because for the vendor, it’s always better to just buy, for you to buy a named-user license and be done with it. And I mean, everybody probably who’s sitting here knows that the era of perpetual and maintenance has slowly come to an end and we are now in subscription, everything is subscription.

[29:52] Branislav Potoček: This is what we struggle with sometimes actually guiding our customers. Sure, the data may fully support you to remove half the licenses from the pool. But then when you think about it, you bought them in 2006 for a reasonable price from Dassault or Siemens and now you get rid of them and the next year you explode and suddenly you need them back. Well, then they’re going to sell them back for the price from 2006, they’re gonna be 10 times more expensive. That’s probably just another variable that can come in too, like the whole decision-making process. And I wanna make sure that we don’t run out of time. So this has been the polls that we’ve been running that we’ve been posting on the event app. And I’m not sure how many people in the audience are, the people who actually asked for them, but it’s interesting to see, this actually does support my theory that in engineering, still a lot of organizations that actually maintain the spreadsheet. And don’t get me wrong, I’ve been doing engineering renewal for a 55,000-employee organization, 12,000 engineers and all the renewals work for maintaining the spreadsheet for many, many, many years because we couldn’t get anybody to approve a tool.

[30:57] David Boyle: So when I joined BP, which you’ve all heard of, multinational company, they honestly had their engineering portfolio in a spreadsheet. Yeah. You know, and that really wasn’t that long ago. Happily today, it’s in ServiceNow. Happily today. You know, it’s not so long.

[31:15] Branislav Potoček: I have a good story when I was maintaining a spreadsheet and I tried to get the tool. And the person who was holding all the engineering budget, which we are talking about around $150 million told me, “Why would I want to reduce my budget? It’s better for me if I just keep the money flowing year by year. It’s not going out of my pocket.” And we had to wait for him to retire. So we could actually go and get the tool to help us reduce the cost in the organization. So I guess you will, you know where we leave it from here. I really want to make sure that we give some time for the Q&A, if there are any questions from the audience. So please, thank you very much for listening.

[32:07] Audience member: [Question partly inaudible] …engineering software and SAM separately, or do you just…

[32:14] David Boyle: So when I was at ConocoPhillips, we had to touch on that. Like we had to keep it completely separately. Actually, you know, in an Access database. Later, we actually created fields specifically in ServiceNow. So we actually had fields created at the application package level to contain, okay, I want to be able to essentially have your asset base, which is, you know, how many modules you own, because there’s always a pyramid of functionality. We have a lot of base licenses that move up to really expensive things. And we knew it was really important to keep it in the, you know, the company’s repository, which was ServiceNow, but we had to have the bespoke fields created just for the job.

[33:04] Audience member: On the dashboard, you can actually see, like these are the engineering software needs to be handled separately.

[33:10] David Boyle: Right, exactly. So you’d still get all the same dashboards because you’d have the… You’d still have your infrastructure, topology, and you’d have all your ticketing, so Petrel from SLB, which had in BP had about 70 or 80 modules from Schlumberger, and maybe had about 30 or 40 from other companies, a whole load that were in-house developed, over a hundred modules, all with costs and things, but in the deployment, in ServiceNow, it just said Petrel, just had one line for their cloud infrastructure and maybe into every environment. That was it. But actually, we had a hundred different tools and products. Managing that difference was the most difficult part.