Computer Modelling Group Ltd. (CMG) Earnings Call Transcript & Summary

Thank you, gentlemen. We'll now address any questions or comments from shareholders or proxy holders that are directly related to the appointment of the auditor. Have any questions come in?

No, no questions on this item.

Thank you. We now invite shareholders and proxy holders to submit their vote, if they've not already done so. As a reminder, if you've already voted or sent in a proxy, there's no need to do anything unless you wish to change your vote. We will now have a short pause to allow shareholders and proxy holders to complete their voting on this matter. [Voting]

Thank you. The next item of business is to consider a resolution to approve the renewal of CMG's shareholder rights plan and the amendments, as more particularly described in the information circular. I note that the amended version of the shareholder rights plan was available in full text on CMG's website for review by CMG's shareholders. The full text of the ordinary resolution is set forth on Page 22 of the information circular. Unless there are any objections, I'll dispense with the reading of the resolution and ask for a motion from the floor approving the renewal of CMG's shareholder rights plan and the amendments. I now ask for a motion on this matter.

I move that the ordinary resolution set out on Page 22 of the management information circular with respect to the renewal of the shareholder rights plan of the corporation and the amendments described in the information circular be approved.

I second the motion.

Thank you, gentlemen. We'll now address any questions or comments from shareholders or proxy holders that are directly related to the renewal of the shareholder rights plan and the amendments. Have any questions come in?

No, we have not received any questions on this.

Thanks, Kristin. We now invite shareholders and proxy holders to submit their vote if they've not already done so, and we'll now have a short pause to allow shareholders and proxy holders to complete their voting on this matter. Following this, the polls will close. [Voting]

Thank you. As we've now dealt with all the business items on the agenda, I declare voting on all matters closed.

The polls are closed for all motions.

I understand that the scrutineers have a preliminary tabulation of votes cast in respect of each of the items of business before the meeting. The preliminary result with respect to the appointment of auditors is 99.14% voted in favor of the appointment of KPMG LLP. As a result, I declare that KPMG LLP has been duly appointed as the auditor of the corporation. The preliminary result with respect to the number of directors to be elected at this meeting is 99.99% voted in favor of fixing the number at 9. As a result, I declare the number of directors to be elected at this meeting fixed at 9. The preliminary results with respect to the election of directors are a substantial majority of the votes cast at the meeting were voted in favor of the 9 nominees named in the management information circular, with each nominee receiving in excess of 90.09% in favor. As a result, I declare that the 9 director nominees named in the management information circular have been duly elected. On the resolution to approve the renewal of the corporation's shareholder rights plan agreement and the amendments, 99.97% voted in favor. As a result, I declare that the resolution approving the renewal of the corporation shareholder rights plan agreement and the amendments has been passed. The final results of today's vote will be posted tomorrow on SEDAR. That concludes the formal business of this meeting of shareholders, and I declare the meeting terminated. We will now have a presentation by our President and CEO, Ryan Schneider, following which we will answer additional shareholders' questions as time permits. I remind you that questions should be of general interest and not of a personal nature. Over to you, Ryan.

Thank you, John. Good morning, and good afternoon, everyone, as the case may be. Glad to be here. Glad to have you here as well. And so my name is Ryan Schneider. I'm the President and CEO. I have been for the last several years, 3 years is my account. But with the pandemic in last year, everything is a little bit blurry, and the year seem to grow together a little bit, but I'll be celebrating my tenth year at CMG come the fall. So the 10 years have gone by incredibly quickly, but it's been wonderful to work with the team at CMG as well as the Board. And so I just want to make some opening remarks before I get into my update presentation and say some thank-yous and some other things. But glad to be here this year, for those of you who followed the company and were around last year, right about this time during the actual AGM last year. I was at the hospital due to my appendix. And from the feedback, I heard that Sandra did a much better job of the investor presentation and her delivery. And so I'll do my best this year to be just as good, if not better, as Sandra did last year. In addition, levity aside, I did want to express gratitude to a number of groups within CMG. And I wanted to thank our shareholders for their support and your patience during this past year. I think it suffices to say it was an interesting year and a difficult year, and a new way of doing business and approaching our customers. And so we definitely thank our shareholders for their support. I'd like to thank the Board. Both are inclusive for their sacrifice, their support of myself and CMG as well as their leadership as we navigated through this new environment during the last year. I'd like to thank my fellow executive for their leadership as well as their sacrifice and for just helping CMG to achieve the results that it did and come through this past year together relatively whole. And like to thank our staff, not least -- last but not least, but thank our staff for their sacrifice and support of CMG, and especially for all their hard work so that we could achieve what we did in this past year. And also in transitioning to a virtual world and work from home, not without its hiccups and bumps, but relatively seamlessly, nonetheless. And really, all pulling on the road this past year, even though the environment and the situation changed. And so really, thank you to everyone at CMG for helping with that. I did want to make one special honorable mention with many of our investors dialed in and our Board and executives together. It will be one of the few opportunities that I have over the next several months or quarters. But I just wanted to say thank you and acknowledge Kelly Tomyn. If all goes according to plan, Kelly is serving roughly a 1-year term, finishing in September-October this year. And really grateful for her support and commend her for the minimal amount of time that it seems and apparent ease that it took her to grab onto CMG's business and start contributing. And so I just wanted to recognize Kelly as one of the public opportunities that I might have to say thank you to her for her contributions to CMG over the year that she's been with us. So with that, I'll get into my update presentation. If you do have questions, please type them in. They're only seen by our moderators, and they'll be -- they'll repeat it back to myself and John, and then we can answer them. And happy to take questions from all of you. So with that, on to the presentation. On the first slide, for our AGM. I'll stop for a second on Slide 3, which is the forward-looking statements. We may make statements that are, of course, based on our current expectations or aspirations that have risks and uncertainties and are beyond CMG's ultimate control. Actual results may differ from what's represented in the presentation. And past history is no indicator of future performance. And so please treat the following statements as forward-looking. So on the fourth slide, around what we do, and that hasn't changed, I'll say, although I'd like to add to it, when I speak to the slide, as we engage in software development, primarily. And then license that software to our customers. For the most part, they're using our software on-premise inside their own environment, sometimes in the cloud in their own environment, but generally on-premise, and we do have a cloud option, which I'll talk about as well. And then we engage in specialized consulting services to help our customers use our software, especially in advanced or bleeding edge scenarios where there may not be a lot of experience in terms of new recovery processes or new physics and new science. We do also support a consultant ecosystem that supports our customers for more straightforward and sort of engineering advisory or engineering call-off where they're just working on projects for a more straightforward engineering problems as well. We engage in training and both on site, virtual now and private training, public training, and had a lot of support and messaging around how to better use our tools and get maximum value from the technology that CMG brings to the table. And we engage in collaborative research where our customers are asking us for solutions to problems and want to model them in the simulators and figure out what it means for their assets and the value of their assets and those. So they'll come to us and say, we need you to support this science so that we can make a decision about how applicable this is to our areas. And historically, that evolved into things like FlexWell or enhancements to FlexWell. We recently did something around nanomaterials and injecting those nanomaterials into the reservoir. And that's turned into -- for the customer, ultimately, a pilot or field scale test of that technology. And so prior to 2 years ago, we didn't support nanomaterials in the simulator directly, for example, right? So just groundbreaking and adding functionality that makes sense that our customers can make better decisions and earn higher margins and reduce the risks and so forth. The one thing that I wanted to add or the one twist is we're now more into production engineering as well, especially with CoFlow. And so now we're focused on more than just the reservoir and so up the well string and the downhole equipment. And on to the surface and compressors and pipelines and so on right to the plant gate. We don't necessarily do prosthesis modeling, but getting the various fluids and gases and so on to the plant gate is now part of our expertise and our technology as well. So Slide 5, the value proposition of what we do for our customers. Really, with our software and our technology and also the people that support our software as we help energy companies to generate greater returns and maximize the value of their assets, maximize the value of their production. And I'm more of an opportunities guy. So I see the upside and the opportunities, but we also help them to manage their downside risk and reduce their risks. And then really just optimize their investments, where they should invest, how they should invest, what type of engineering decisions they should make, improve their recoveries on their assets. And it's -- a relatively small cost in software can have great impact -- financial impact, the economic impact, environmental impact as well in a positive sense to what they're doing with their assets. And on the right-hand side is just an example that we've shown. Customer had a rule of thumb for the well spacing and the number of fractures to use in their reservoirs. And running a study with our tools and CMOST, which is our optimizer and then certainly, management on top of everything. We're able to pick a better sweet spot that gave them almost a doubling in net present value per acre. That's an extreme example, even with the amount of money that's being invested, the amount of barrels that flow from assets, even a 10% or 20% improvement is huge value compared to what customers are spending on our software. So the next slide, our specialization. I think that's Slide 6 if you're working by the numbers. We're really well-known for difficult oil, as we call it. The harder recovery processes, the more advanced physics that takes us into areas like unconventional or hydraulic fracture in tight reservoirs, shale oil and gas. Compositional where you're injecting gas or CO2 or other things and compositional has a tie-in to my next slide around carbon capture and other things. We've always been known for thermal, especially in Canada, but also in other places in the world, Middle East, China, Russia, Latin America, around using anything from steam, hot water, and other thermal methods to increase or enhance the oil production. Chemical UR, injecting chemicals to enhance the production as things around our geochemistry and now finally into integrated reservoir and production, the models where you're coupling the reservoir to the surface network or even coupling multiple reservoirs and then when one reservoir is producing too much gas and you choke it back, then another reservoir can produce more. And so you can now start to look at things that the asset view as opposed to, at the reservoir level and make decisions about your entire asset. And one of the things that was an interesting learning for me is that reservoirs can produce a lot and have all kinds of potential. But they're generally limited by sometimes the production network and the downhole equipment in the wellbore and so on. And so it's an important piece to couple to your reservoir to make better decisions. And there's a lot of investment that goes into the equipment and so forth. And so to be able to optimize that also helps our customers to do a better job. And you have all kinds of case studies showing that those sort of optimizations can add incredible value to our customers. So the next slide, regarding the energy transition, we've had differencing capabilities, in some cases, for as much as 30 years, and we continue to evolve GEM and STARS to support these areas. And I think what's interesting even through the pandemic. And with our customers this past year, this has been a bright spot for us in terms of an area of customers using our software as well as additional trainings and additional interest from a broad array of both the traditional energy producers, but also new -- what I'll call new energy companies that are looking to solely produce hydrogen, for example, or solely produce what would be considered green gas and so forth. And so we've been giving a number of talks. We've demonstrated a number of models and capabilities over -- honestly, since I joined the company, there's been a lot of activity in this area, and it was sort of up and down and in Southeast Asia and then in Europe. But interestingly, even this past year, we've done trainings in Houston, and there's a lot of interest around the world. And it's generating real revenue for us. If you squint at your thumb, we estimate and one of the reasons that it's not fully resolved is because our customers use our licenses for many things and don't necessarily need to tell us exactly what they're doing. But we estimate that we're generating somewhere between $2 million and $2.5 million a year in software license revenue tied to the energy transition right now. And it appears to be growing, and there's a lot of interest, and there's money flowing in this area. So it's -- we consider a breakout area, and we're going to try to track it as quantitatively as we can going forward. So in terms of the product suite, that really hasn't changed much in terms of what we've had in terms of the names. But of course, the -- what they do underneath the hood has improved materially since the last time we spoke with you a year ago. And so we continue to make improvements that help our customers to solve the problems of value to them, to help them go faster and make a difference to their workflows. And I would say that we're extremely competitive with our technology compared to our competitors, both in terms of performance and functionality. And again, we're really known for the difficult oil in terms of what we can do and we're the go-to company when it comes to the hard stuff, right? But -- so in terms of the simulators, IMEX, GEM and STARS from left to right, sort of easiest to hardest oil, if you want to call it that or amount of the science and physics that each one models. With those 3, we can model any type of reservoir, anywhere in the world, any stage of life. And so we really have coverage of all the different energy companies and their production processes and mechanisms. I'll call the bottom row there, supporting cast and so building results help us to build models and then consume the results. And CMOST is the sort of workflow manager, if you will, that sits on top of all of those and can help our users to run hundreds of models and hundreds of different scenarios and then drive them either towards quantifying the uncertainty, so that I can give a band of -- I think it's -- the outcome of this is between $45 million and $55 million and with some reasonable certainty as well as to optimize and generate optimum answers that the humans ultimately cannot. And so we have a saying that a human being can remember 3 things in their mind or keep track of 3 things at a time. And obviously, the CMOST deals with up to hundreds, even more variables at a time and generates an optimum solution and does it a lot faster than the engineer can on their own, right? So it really augments what the engineer can do and the quality of the answers that they can deliver. Out there on the right-hand side, of course, is CoFlow, and it really includes a lot of the pieces that it can do black oil and compositional simulation, in some cases, K value, which is STARS as well. And it also handles building the models, consuming the results and how steep most embedded in addition. So really, it rolls a lot of our different technologies into an environment that then also supports the production engineering pieces that I mentioned prior. So in terms of competition, and I think this past year, one of the questions we get from our analysts and others, investors is, has there been much change in the competition and in anything to pay attention to and so forth? And I would say that it was a relatively quiet year on the competitive front, mostly because with the virtual world and working from home and so forth, there wasn't a lot of in-person trade shows and a lot going on. And so we didn't see a lot. The one thing we did see is some of the competitors pulling back on their marketing, investing -- investments and spending and not a lot of super aggressive pricing or freebies or other things, not a lot of change from, say, 2019 to 2020 in terms of behavior tactics. I think that the couple of callouts is that we continue to see Slumber J and Rock Flow Dynamics fight it out at the bottom of the market around black oil. And RFD offers a clone of Eclipse and the differentiator being pricing. And so to the customer, it looks like the same thing, just the difference in price. And so I think they continue to battle it out there. However, the other thing that I'll call out is that we have seen a number of RFPs and customers looking and even one of the wins that we had during the year in Bolivia was a customer that was using Nexus, the Halliburton product and then switch to CNG. And so we are seeing a dwindling of Halliburton's market share. They never had necessarily a large market share to begin with, but I think they're seeing further attrition of customers just as there's questions about investment and the longevity of the product. And also, in terms of in house products, I think what's interesting is there's few in or fewer. Shell is using us right now and looking at a lot of CoFlow and less of their in-house tools. ConocoPhillips uses a lot of CMG right now, and they're contemplating whether they're -- they'll continue to invest in their in-house tool or not. And so there's very, very few companies that are doing their own thing in resource simulation going forward. And I think that's just speaks to that the third-party vendors with multiple customers, significant investment like ours and the technology can do a better job than any one company at furthering the technology and delivering value to the user base. So on to the history slide, you can see the history there. First IPO and first and only raise of capital in 1997, positive cash flows, and then dividends paid and so 70 quarterly dividends paid and $350 million paid back to shareholders since 2004. And not much more I want to say about the history, except we've been doing this a long time. And we believe, in fact, that there's a decent amount of runway for the oil and gas industry, and it's going to be harder and harder to produce those incremental barrels. And our software will be more and more useful and more applicable to the reservoirs that our customers have. In terms of the deep bench of intellectual capital, just our makeup of the company and education background and so forth, we're at about 200 employees. We've been at about 200 for the past several years. Our biggest expense is people, about 80% of our expense is just under. And so as we've held the line on expenses, that means we've held the line on hiring. Just over half of the company is designated towards research and development. These are software developers, mathematicians, physicists, engineers, developing solutions in the software for our customers. We have another 25% roughly of the company, 48 people in consulting, support and training. These are reservoir engineers mainly that support our customers in terms of using the tools, training them and helping them to get more from our software. And then lastly, in the last part of the pie there, we have our sales and marketing teams, our men in finance and IT. And so we're really well educated group, very, very smart people. 75% of the company really focused on customer facing, delivering the technology to our customers. There's the executive team. You heard their names listed off during John's introduction. I won't say much more about them than that. But I've got a great team to work with, and I think they all do a great job in terms of their respective areas. And I am trying to really help serve our customers and also grow the business and grow the future revenue, which I'll talk about in a moment. So global reach slide, in terms of our footprint around the world, we have offices in most of the places in the world where there are reservoir engineers. And so all the development is done in Calgary, but then we have customer-facing offices in the next 2 largest offices are Houston and Dubai. Following that, we have one in Bogotá, one in Kuala Lumpur, one in near London. And then smaller offices in Oman and Rio. And we also have agents in Russia, China, India, which makes sense in order to do business with there. We deal with most of the super majors, if not all of them. We have the top 25 -- 90% of the top 25 companies that have ranked by production. And we have 72% of the top 25 unconventional producers. And we've been incredibly successful in the hydraulic fracture space. And that's one of our -- even though it's a bit pounded in the U.S., it's one of our potential growth areas for the customers there in terms of supplying more of the workflow and the functionality to them. Right now, our business mix is 40% Eastern Hemisphere. Canada is just under a quarter of that. And then United States as well, just over a quarter and a smaller portion in South America at the moment. So just briefly touching on the customer slide, the customers near the top of the slide and with the bigger logos tend to be our larger customers. We have most of the household names in Canada. Slumber J Manager said that CMG is 98% market share. In Canada, I think we're in the incumbent, but in most of the other places in the world, while we've been incredibly successful in certain areas like in Southeast Asia or in India, or Oman. We're not necessarily incumbent and so that's an opportunity because those companies can use more CNG and get the value from the way that we do business and the way that we support them and so forth. But so household names to customers in Canada, customers in the U.S., and a lot of the hydraulic fracture players and some of the new entrants in the market in that space over the past more than 5 to 6 years now, but it's contributed to a lot of our growth in the U.S. Our customers in Latin America, the -- most of the national oil companies in that region, and we've had quite a bit of success since we started well ahead of days, in fact, and Venezuela has been a customer and a member partner since the late '70s, but continuing that into Colombia, into Brazil and so forth. So again, a lot of the customers in Latin America that you'd recognize as the national oil companies from that region. In Europe, a lot of the European international oil companies as well as some of the more green and new entrants into the space, whether that be for hydrogen or CO2 sequestration, again, a lot of the names that you would expect tied to the energy industry. In Asia Pacific, same story. A lot of the national oil companies and significant success, I would say, in Korea, in Japan, in Malaysia and Indonesia and the surrounding areas as well. And lastly, in the Middle East, we opened our Dubai office in 2009. This past year, I think, we're now into 2 years of the contract, but had some decent success in Abu Dhabi after a number of years of working in that area. And that really helped out the Eastern Hemisphere region this past year. So we're grateful for them as a customer and continue to work hard to earn their business and widen CMG's footprint. But we have great success in Oman with both PDO as well as a number of the local companies there, too, as well as in Kuwait, Qatar and other places as well. And so Eastern Hemisphere has been quite healthy and quite resilient even through the pandemic. Just want to talk a little bit about our university strategy, which is the next slide. There's 2 slides called university customers, Slides 20 and 21. And we supply our software to all universities that have undergraduate programs or graduate research programs around engineering. And it's really a bit of a grassroot strategy. People get comfortable with CMG and learn how to use it. And then ultimately, prefer to use it inside the commercial energy companies. And so we -- the universities get it essentially for free or for a small administrative fee. And we get exposure to all of our future users, and that has worked incredibly well. And I think we do this better than our competitors do if they even do it at all. So now Slide 22, our annual software license revenue. There -- and so on the right-hand side there, same as the trailing 12 months, but how we did for 2021. We did get hit by a few things: one, consolidation in Canada; and also a customer that was back on maintenance and then went off maintenance due to the pandemic, but that customer has had a history of doing that during the tough times of backing away from maintenance and then coming back on when things were better. And so I think that, honestly, given what happened in the world around us and what happened with oil price and market values or certain market caps and everything else, I think, that last year was a reasonable year, all things considered. And I think sentiment is improving as we go forward. Looking at the quarterly license revenue that now with the prior periods in there, we sometimes have customers that pay us either on a COD basis. So just due to geopolitical risk or payment risk, we don't book their invoices until they pay us. And in some cases, that's our South American customer that has difficulty paying or difficulty getting access to U.S. dollar currency. And so in the past, some of those big yellow bars would have been that customer. We haven't had payments like that in the past year, and so the yellow bars are smaller, but there are things that they're either in an area of geopolitical risk or the contracts or the paperwork was delayed such that we didn't recognize the revenue until we got paid. In terms of the Slide 24, it's now the quarterly software license revenue versus the WTI price. There's really 3 comments I wanted to make about this. And historically, we've been asked do we correlate to oil price? And when I first joined the company, our answer was no to that, not really. But I think as we got bigger, there was more of a correlation. And so really what we're tied to is customers, budgets and really operational spend, a little bit to their capital spend, but we really prefer lease licenses as opposed to perpetual licenses, which would be CapEx. And so it's really tied to customers' budgets and if customers' budgets are under pressure, then maybe our revenue also comes under pressure. But I would say that from that plot and what's happened over the past year, we're relatively resilient, I would say, is the first point. The second point is that the impact of commodity price or drastic changes don't necessarily impact a sort of way, there's a bit of a lag, if you want to call it that. But that's because of annual contracts or multiyear contracts that are already in place even when the commodity price changes. Now that sort of cuts both ways. So there may be a lag. In the future, I do believe sentiment is positive, and oil prices getting better and so forth. And so where we expect to see the benefits of that is as customer budgets get healthier or improve or increase and we review about 50% of our business at the end of the calendar year. And so we would expect to see the benefits of that as we go into the end of the calendar year. There will be some customers whose budgets change midstream or midyear, but for the most part, they're working with the budget that they had and that they set last year for 2021, so to speak, and so we'll -- I think we'll see better prospects and improvement in that story as of December of this year. I think one thing that's interesting is that we are seeing customers trying to lock in multiyear pricing at the moment and what we see that, as we say it around the sales table, is that the customers are trying to lock in pandemic pricing, right? And so I think our readthrough is that they see better times ahead and they want to try to secure the pricing that they had from say a year ago, right? So -- and then annuity and maintenance revenue, one last slide there, just showing it. And there is a bit of seasonality in, especially, our deferred revenue numbers, which I'll talk about in a minute. But you can see it also broken out by a quarterly basis and how it looks in the comparative quarters down below. So deferred revenue, those are -- that's Slide 26. Those are prebooked licenses, usually a year in length. We don't generally book anything longer than a year in length, outside of a Shell contract that showed up as long-term deferred in prior years, but that was more of an outlier than standard practice. And so as we book all those contracts, as I mentioned, around December 31, January 1, then our Q4 deferred revenue balance spikes up, and we drawdown on it for the subsequent 3 quarters. And so that's why you see the Sawtooth sort of impact that you're seeing. Analysts look to this number quite a bit as it's sort of a good indicator of upcoming revenue, and it's revenue that's already booked or even, in most cases, even paid for. But we have to recognize it as the software licenses are used and time passes. So that's -- it's a good indicator of the revenue, will do going forward. In terms of perpetual licenses, we are quite pleased with the number we did in 2021, towards the middle of the year, and especially in the pandemic and customers pulling their fingers in. We thought that we might do about $2 million a year -- that year. Obviously, we did quite a bit more than that in 2021. And so we are pleased with that, all things considered. It contributes to future maintenance revenue, which is good. And I think the other part is, is that the one part to call out is it's very unpredictable. Even last year, we had a quarter that didn't have any perpetual licenses in it at all, which is pretty extreme. But it's difficult to tell when it will come in and how much it is. We still pick them. This is not our preferred licensing model, but there are customers that still prefer perpetual licenses. And so we're happy to facilitate that if that's a big real fielding. I mean, it does turn into that long tail maintenance stream, which is what we really are focused on. In terms of historical margins and return on equity, Slide 28, you can see our operating margins and so our target really there, historically, outside of the downturn and outside of a pandemic is 50%. EBITDA margins look very similar. There's not a lot of necessarily deltas between the two. And we don't use a lot of sort of -- this isn't a return on capital this here, but I did want to mention that we're a fairly capital-light business. And so our return on capital numbers, for example, look great because we just -- generally speaking, we buy computers, and that's about it. We're not a capital-intense business. And so for fiscal '21, $34.8 million in EBITDA. In terms of free cash flow from operations, $24.4 million. And we do pay the dividend, as I mentioned prior. And so to date, we paid $350 million to shareholders. Coming up on my future is one last slide to close out in terms of just institutional ownership. Same -- relatively same list of names as it's been there in the past, and we tend to attract long-term value type investors. And so edge point in Burgundy being some of our largest shareholders with [ T- roll ] price having helped CMG for a long time. And then the rest of the folks, top 20 reporting institutions making up 72% of the ownership. Directors, officers and other employees hold approximately 6% of the shares, and we have significant shareholders, both on the Board as well as in the executive management team, which I think is great and important. I'm not trying to buy more shares as I can. So I think it's good for alignment with shareholders, obviously. So the CoFlow slide, Slide 31, moving into Slide 32. We've talked about this before, but CoFlow is a little bit different in the sense that it's a multiuser environment. So a team of people can work on different components of an asset, different people, so whether they be production engineers, drill mechanics, specialists or reservoir engineers and then combine their input to come up with a model that incorporates all the learnings and all the things that the different disciplines know. You can also engage in multiple fidelity where you model something at super high-resolution in 3D or you model it with the decline curve and a relatively simplistic view, depending on how much impact you think it has to the asset. And then, on top of that, you can do optimization and uncertainty across multiple reservoirs, the network and the functional assets. So this -- it's harder to do, but it makes a lot more sense to me because of my comments earlier. The production network, gates or throttles, what a given reservoir to produce. It identifies opportunities where if you increase your pipe size or your compressor size, you can get better results, and that's an investment trade-off. And I think one of the really interesting things about CoFlow is that moving maybe to the benefits of CoFlow slide, Slide 33, is that the initial conversations that we've had with our paying customers around CoFlow, it's really been about -- I have a new plant, I have some capacity that I need to fulfill, help me to solve this problem. So it's really business value. Some of the conversations we get in. There's nothing wrong with it. One of the conversations we get in is, do you support chemical 365 that I want to inject into my alkaline surfactant polymer flood here in this reservoir? And so it's very, very detailed down in the deep technical sort of requirements versus the one -- the first customer that we had outside of Shell and Petrobras that came to us, I think, that's now 2 years ago. They had a gas plant, and they were saying, we need to put in a compressor. I need to know how big, I need to know when to put it in. And that -- and if we produce enough gas in the plant, it's going to be a big party for all of our shareholders in the company, right? And so really straightforward sort of business conversation about how we can help them and help them to make better decisions, right? And so when it comes to offshore, which is what Shell and Petrobras focused on, some of our other customers, big, big investments. And if we can help them to do better job, they completely pay for their software investment and reap much, much higher returns in that. And anything that shortens the time to production or shortens the -- or removes any deferred production, so they're producing in an optimal sense. Means a lot of value to our customers. And we're seeing that with the employed customers that we have with CoFlow now. In terms of assets with active copay deployment, there's a lot of dots on the screen. There's a lot of customers that we're working on, and we have 4 publicly-announced paying customers outside of Shell and Petrobras. And so we just added another 1 in Q4, and we're pretty excited about that. And look forward to continue to engage in a lot of proof of concepts around CoFlow, and look forward to turning that into commercial traction and building the story and being able to point back to those customers and saying, look at the value they're getting and bringing that to additional customers. And so we're definitely focused on that. And it's one of -- when I talk about the 4-legged stool of how we're going to grow, CoFlow is definitely part of that mix. It also takes us into an additional addressable market, which is production engineering which we believe is of equal size to the reservoir simulation market as well. I won't spend too much time on CoFlow, except to say that in addition to CoFlow, we can now couple our existing simulators and models for -- in IMEX and GEM and STARS into a network model or a surface network that's running CoFlow and run them in a coupled fashion. And we're even -- this past year, we've added IMEX and GEM to run under the hood, as we call it, inside CoFlow, and so we're blurring the lines between our technology and really just taking the best technology that makes sense to making sure that it can run in CoFlow, with CoFlow and solve these larger problems that customers are interested in. Another area that's been a healthy area for us and a contributor to our growth to date has been unconventional oil and gas. And so we have a long history in that now. I'm going to Slide 37, the oil and gas history. Starting with papers in 2009, so over 12 years ago and then improvements to horizontal and hydraulic fracture modeling. All the way to now thousands of wells and one of the big features in the past, I'll say, 1.5 years was the ability to model overlapping fractures better. So now the fractures touch each other, they go over top of each other. And being able to model that well in our simulators. And so when we started, customers would do, say, 30 or 40 fractures per well and maybe one well. And now we're getting hundreds of wells and hundreds of fractures per well and saw model with 6,000 fractures. And so you can imagine, that's not where we started with the design of the software and the workflow, and we've caught it up to where it can really do a lot more and do what customers are really doing. I think what's important there, why isn't an area of growth and opportunity for us, is that customers want to model not only the production from the fractures, they want to model the fracturing process itself and the native rock. That's more gene mechanics, that's more run time, the pump job itself. And they want to do the [ head all ] in the same software environment, and we're working on that. And it will mean more licenses for our customers in order to model it from end-to-end like that. And so we do believe it's an opportunity, and it's something that we've invested in. I'll stop briefly on cloud and just talk about -- there's now Slide 39, the CMG cloud solutions. We -- cloud is -- one of the things to clarify is that the cloud is viewed as a way to save money on the IT side, and that's not why we got into cloud. We got into cloud to enable our customers because there's a lot of obstacles, whether it be the best hardware, how much you have of it, when you have it. And so what we found is that our customers' needs are quite lumpy, in fact. They'll get busy in, say, the spring for operational planning or in the fall for reserves reporting. And so they have really spiky load. It's not -- doesn't necessarily fit in a nice rectangular box that I use exactly 5 licenses or 10 licenses, Monday to Friday, every week of the year, right? And so it's really an enabler, and that's the way we see it for simulation that customers consume a lot more software. And also do it when they need it against their more lumpy loads. We've see that with customers that are using the solution. They're getting great results. They can tee up jobs and tee up CMOST projects. They leave at 4:30. It runs on a lot of computers overnight, come back in the morning, and they can start to look at their results. And that's really, really powerful compared to lining up behind a constrained set, a smaller set of resources, both computers and licenses from us and waiting to get the answer. And so it's really turned what our customers can do on its head, and it's really powerful and it's valuable, too, I think, right? So all of those things. I'm going -- secure and performance and everything else, we can support data centers around the world. Most of our usage is in North America, primarily U.S. with a few Canadian customers in there. And so just looking at Slide 41 and 42, we've had a number of trials that continue to increase over the years. And then we have a number of commercial customers on the right-hand side there. They are regularly used in cloud for their simulation. I think what's interesting, I made a comment to an investor, just the other day. Where last year, we did just under $1 million in revenue this past fiscal year -- so sorry, 2 fiscal years ago, we did about just under $1 million. This past fiscal year, we did just over $1 million. And it was neither way up or way down. But when you think about it compared to what happened during the pandemic being flat or slightly up means that there was implicit, sort of healthiness in growth in there compared to everything that happened in most of our other markets, right? And so cloud's growing and what our focus is now is getting some of the mid-tier and larger players. They have more security concerns, more IT concerns. By getting them to use cloud and we'll see a significant jump in terms of consumption, which I think is healthy or will be good for both our customers and CMG's business as well. Just to call out a few more things. I'm going to skip towards the end. I want to be conscious of time and give folks a chance to ask questions. But we continue to invest in CMOST. There's a number of things that our customers have asked for in terms of how to run more and more jobs with CMOST, which is a good thing. And do it in a more intelligent fashion and with AI and machine learning overlaid on that. And so it continues to be a flagship product for us. It's also a differentiator in terms of the competitive space as well. And so we continue to make CMOST better and relevant to our customers. In terms of performance enhancement technology, we -- so there's now Slide 46 and 47, we continue to invest in performance. We, in the past several years, started to bring MPI, which doesn't mean a lot as an acronym, but message passing interface, the ability to run models across multiple computers connected together with the network. It allows you to use cheaper hardware to solve a bigger problem and ultimately build a bigger hammer, if you want to look at that way. And for some customers with large models or a lot of models with long run time, we can apply more computer hardware. It means more licenses for us, which is good, and we can apply more hardware to solve the problem faster, and there's business value to doing so. So that's very important. One of the things that's made a lot of waves with our customers and really it's been wonderful. That's I think the only way to phrase it is Autotune and so that's using artificial intelligence inside our simulators to improve the way that we run, so that we get faster results and without user intervention. So it's adaptive and it looks at what's going on inside the simulation, and it makes it run faster. And so average results you've seen anywhere from sort of 2 to 3x. There's some extremes on there of 9 and 17x. I think, 12 hours went down to 40 minutes. But customers that are paying maintenance and leasing our software get access to the new versions of Autotune. And honestly, it kicks b***, right? And so it's value that nobody else has in the market and really helps our customers to run faster, which is awesome. And some -- one of the questions we get is, well, if the Autotune run faster, does that mean they use less license? It's no. They just throw a bigger, more complicated model at it. They're already sacrificing in terms of the resolution and the accuracy that they're getting. So that it fits within a box like overnight or over a weekend. And so we can just allow them to do a better model of higher quality and get them the run time on it. So there's no risk of -- because we make it so fast that, that they stop using it or use less of it. It's the opposite as they use more because they're already compromising. One call out, our IMEX team from a version a year ago to a version this year in terms of IMEX and in combination with Autotune, it's 3x faster. So the release that we shipped in earlier this year was 3x faster than the release prior, right? And so we're doing awesome things around delivering that value and really being competitive in the space with respect to our competitors and delivering value to our customers. So it's just -- it's awesome. It's an awesome feature that we brought to market over the past 3, 4 years and continue to enhance it. Third, from the last slide, I think, just wanted to talk about online training for a minute. We had a huge impact. We pivoted to delivering virtually and instead of used to have classrooms and mostly in-person and structure-led, and we moved to online structure-led in both May-April last year. And so I ended up training almost 10,000 people, which if you look at sort of the cumulative amounts, the 4 years prior, we probably came close to training 10,000 people. So a fundamental shift, I think, in the number of people that we're able to reach. We got more people's attention in the pandemic as they're working from home. And we're able to help them to learn our software and get value. And so I think that's -- I'll call that building for the future. I think that, that's going to have positive impact on our reach and ability to succeed in coming years. So closing up what we usually say to investors and to the folks when we speak to them are new prospects that we talk to that are considering investing in CMG, is that we're a dominant player in a very complex market. There's a high technical moat to getting involved in these types of simulations. A very loyal and recurring customer base. We measure ourselves based on the recurring revenue. And our goal is to increase that recurring revenue over time. We have great margins even through this past year, and we have great staff. A lot of them have been with us for a long time and want to continue to work at CMG, and we've paid dividends for many years to date, right? And so I think there's a lot of good reasons to be an investor at CMG. In closing, this is a new slide for us, but I think it's important is that we recognize in our minds that we're valued based on 2 things, really: our top line growth and our margins, our ability to generate cash. And so those are the 2 things that we're focused on. We get questions too around what's going to happen in terms of expenses and what's going to happen this coming year. We are -- we will continue to hold the line on expenses. And we are managing to margins. So I think it's important that people hear that. And so in terms of the other side of the equation, so that's maybe some statements around bottom line. In terms of top line, it's that 4-legged stool that we want to achieve. We want to get back to double-digit growth. That's an aspirational target, but what we recognized is that, that's what investors would like to see as well. And in order to do that, we'd be firing on all cylinders in terms of 4 areas, as we call it the 4-legged stool, which would be generating revenue in CoFlow. And we've seen a little bit of that, and we're going to continue to work hard to do that going forward. Additional market share gains, especially in areas outside of Canada where we're not the incumbent. Obviously, in Eastern Hemisphere, we're focused. We're going to invest in the Eastern Hemisphere. That's where the rays of Sunshine or the green shoots, if you want to call it that, have been. Most of the prognosticators have said that most production will come out of the Middle East and Eastern Hemisphere. It's also where the lowest cost of production is in the world, is generally Eastern Hemisphere. So we'll continue to focus in that area. And as I mentioned, the unconventional hydraulic fracture workflow and value expansion, doing more for those customers and, therefore, then consuming more licenses from us as well. And so we're focused on growing the topline and holding expenses in check to get back to those historical margins. So with that I am done. Hopefully, I still remain connected. I just talked to my phone for the past 40 minutes or so. And so with that, our moderators, were there any questions while I was speaking?

Ryan, we do have a couple of questions. [Operator Instructions] So the first question that came in a little while ago, is it seems like your business should be sticky, but obviously, the decline in revenue indicates otherwise. Can you explain the kinds of customer situation that result in lost sales? Example, clients moving to competitors or clients shutting down, et cetera?

Yes. So it is sticky and it's hard to transition between simulators and stuff. There's a lot of investment in models and knowledge of the keywords and how the simulators work. And so it's difficult to switch. I give examples like I'm dating myself, but people that new load us 1, 2, 3. And then somebody came in and said, "Hey, when did you try this Excel and convert everything over," and it's not necessarily a straightforward switch, right? And so yes, it is sticky. So what happens then? We've seen a few things. We do see customers that a few customers that have gone bankrupt or gone out of business that's happened a little bit in the unconventional, in the hydraulic fracture space. I mean, national oil companies don't go to business, right? But we have seen a little bit of that. I wouldn't say it's a big part of what's been happening, but it does happen. The other thing we've seen is people go off of maintenance. And so that I made reference to that in Latin America. So they can continue to use their 2019 version of the software, but they don't get support and they don't get updates. And those types of customers have come back. So when the commodity price first rolled over, they maybe went off of maintenance and then right before the pandemic hit actually, say, 18 months ago, a lot of those customers were back on to maintenance. And so that's our goal is to get those customers back, pay maintenance, get them on current versions and get them supported, fully supported and so forth. And so it becomes a budgetary pressure that -- we had this, especially in Canada. I've got a choice between spending on your software versus letting people go. So I need to -- I need a reduction in my contract. I think the one good thing that we've done is that we haven't given price concessions. For the most part, it's okay, you need to take software out of the bucket. And you'll get a lower contract value. And then the goal is, when activity resumes or prospects or investment for the company resumes that they would be getting more licenses. So in the short term, they've been doing okay with less but I believe that that's a short-term fix because they're not necessarily modeling all of their assets or they're just running on sort of their 90-day 3-month forecast, but they can only do that for so long before they need to go back and do the proper engineering work. Lastly, consolidation, we have seen, and I believe that short term. We take a short-term hit when CMG customers consolidate. There's the other side where, in fact, a customer that's not a CMG customer gets consolidated into one of our customers. And then that means more usage and more user base. That's generally a good thing. But what happens in the intervening, say, 2, 3 years is while we've got to make the acquisition look good. So all the budgets are reduced, and we're trying to find those synergies that we published in our press release. So assets moving around, I don't think is a bad thing long-term because they still need to be under simulation. They still need to have the work done, but it may be a short-term hit because of some of the behavior that happened, either they're just busy integrating or they've got reduced budgets to make the acquisition look good. But long term, I think that assets moving around doesn't necessarily hurt. And in fact, you could foresee a situation where those acquisitions go to a customer that uses more CNG or uses more technology to get better value. And that's a good thing ultimately for us. So those are where it comes from is budget pressure, sometimes consolidations. A few customers disappeared. So I think that's why you're seeing that. Last thing, sorry. And the exit from Canada, which is now 3 or 4 years back, but as all the international oil companies exited Canada and sort of they got what was the word? Domestification, as it went to the domestic players. That had an impact on our business because we had large strong customers in the IOCs that were operating in Canada. When they all left, that was like a $10 million to $15 million hit in bookings. It was difficult to recover that in other places. So those are the factors, I think, that led to what you would translate is maybe less sticky, but that's where it came from.

Only one question left so far. How much of your R&D spend do you feel is maintenance as they're necessary just to keep the business as it is versus to grow, get new clients and higher pricing?

Yes. So there's a combination of both. And so in our R&D line, if you look at it, it's not quite 50-50, but some 45-55, I'll say, just as a rough approximation of that is CoFlow. And so that's building for a new revenue stream and for supporting new customers and one of the 4 legs of the stool. So then, I think your question is more tied to, okay, that other 55%, the other that you're spending R&D money on. And so I would say there, we -- it's not necessarily dollars-wise that we track it directly. But 1/3 is on -- we do 3 things. 1/3 is on bug fixes and keeping the existing software going, 1/3 is on enhancements to attract new customers and solve new problems, and 1/3 is on innovation, things that customers haven't even asked us for that we know that they would find valuable. And so things like that are CMOST, which we started on our own initiative, really helps to sell additional simulator licenses. And we knew that we needed to build it and build it without necessarily customers asking for it, right? And so we have those 3 categories. There definitely is a part of our R&D investment that does go towards staying current on the latest hardware and going as fast as possible, and that is a bit of an arms race. Some of our models can run for -- in the black oil space anywhere for minutes to hours. In STARS [indiscernible] gave us the model that ran in 47 days, right? And so they would really like to see that run a lot faster. And so customers will make decisions if one simulator is materially faster than another, right? And we've won every benchmark that we've been invited to. So I think we win that speed game, but it takes a lot of investment from us in order to stay at the front of that and stay relevant to our customers in terms of making sure that we're delivering value there. So there definitely is a component both of fixing the software for things that might not work the way we intended or the customer thought they should. There's an aspect of high-performance computing in terms of making it go faster. And then there are areas where we're building out new functionality. We're investing in energy transition, for example, right now, because we know there's things that we can improve. There's stuff that customers are asking for in terms of pure CO2 injection and other things, which has implications to the technology. And so we're working on those to make our already very, very capable tool chain more applicable to some of those problems. And so I think we'll learn as we get more customer engagement and customers want to spend money with us and then they give us feedback. And we've actually seen this with CoFlow as well. The customer that I talked about in South America with the gas plant. They come back to us and say, "Hey, I need to support this kind of pump and this will help me maybe better and use the software more." And so those are all good things because it makes our software more applicable to more companies and solve more problems and, ultimately, drive more value both for us and for our customers, right? So sorry that I can't give you a definitive answer, but definitely a healthy share of our R&D outside of CoFlow is spent on maintaining existing customers and giving them what they need to be successful.

If there are no further questions, Ryan, let me take this opportunity to first of all, thank you very much for your excellent presentation. And as importantly, thank you for your leadership of CMG through what's been a very trying time of COVID and also the commodity price circumstances of the oil and gas industry over the last few years. And I'm sure I speak for all of us, all shareholders, stakeholders, when I say we look forward to your continuing leadership in the brighter days ahead of better commodity prices and a healthier world situation and all that it holds for the great future of CMG. So with that, let me conclude the meeting. And on behalf of the corporation, thank you all for participating and for your ongoing commitment to Computer Modelling Group. I wish all the best to you and your families and hope very much that you stay safe and healthy. Thank you very much. The meeting is concluded.

Thank you, John. Operator, you can close the call. I wish everybody a safe and wonderful summer.

This concludes the meeting. You may now disconnect.