Good morning, and thank you for joining this presentation today in London. Thank you, François-José. I'm going to review briefly with you the business dynamic in our sector. First, as you may notice, the software revenue is at the high end of the guidance, with double-digit growth for new license -- licenses, as you have seen excluding exchange rate. As you will notice, the exchange rate is having a direct impact on the EPS, and Thibault will go through the detail of that effect. We are having -- we are seeing a very strong dynamic for the 3DEXPERIENCE platform. You have seen some recent announcement especially in July, of course, with the significant milestone with Boeing, which is showing the way things are going forward, but also now with the selection with Scania and other big clients adopting the platform effect in their company which is a very different approach. The dynamic with SOLIDWORKS is strong; similarly on GEOVIA too, we'll cover that. And we continue to show good momentum in Industrial Equipment, High-Tech, Consumer Goods and Package Goods, Energy, Process & Utility. And we have announced this quarter 2 acquisitions, which are really very well aligned with our strategy. I will cover them also, one; Exa Corp., still under way, we need to go through the approval process; as well as No Magic. I will talk about it in a minute for the expansion of our solutions in both the world of multiphysics, multi-simulation as well as system. And we are upgrading the guidance of the 2017 to take into account the currency and the service, while we reconfirm our new license software revenue growth at the range of 8% to 10% excluding exchange rate. So with that, Q3. Q3 is up 7 -- 2%. And if you exclude the exchange rate, it's 6%, which means year-to-date, 7%. The software growth is 8% and 12% excluding exchange rate, which gives you an idea about the operation. The operating margin is at 32%. And as you can see, the effect on EPS evolution is, of course, having a direct effect with almost all currencies having a negative effect to the euro. So Q3 EPS is up 8% excluding exchange rate, and the new license growth is on the top of the expectation. So now if you look at by region, you can see that it's quite well-balanced. I -- we were asked at the beginning of the year about the new policy in America and the effect for us. We mentioned that it was a positive effect. And I think we see it in the number on the -- in dynamic. I was in America last week for the 3DEXPERIENCE forum and the dynamic with clients there is very, very positive toward the adoption of the platform which we think is a way to expand the footprint in all those customer base that we have on [ a results course ], acquiring new customers. Around Europe, strong drive from South Europe and France, where we see a positive dynamic. Asia: China was weak. There was a kind of postpones of decisions, especially for government-owned companies, and there are a lot of them in key sectors like, infrastructure, energy and Transportation & Mobility, or Aerospace & Defense, but was very well compensated by the strong performance of Korea and India. So total revenue, 8% up in Q3, year-to-date 7%. On the domain of coverage, which are well represented with the brands. As you will notice, the Q3 CATIA license is down 1%, 1% up for the year. ENOVIA is up 5% this quarter, SOLIDWORKS 16% and other software 13%, for an aggregate of 8%. What should be the takeaway here? The dynamic of CATIA in the many previous quarters has been strong in China, so the temporary slowdown is easy below the CATIA growth. ENOVIA is up 27% excluding exchange rate, so we have a strong effect of the platform effect. I remind you that the ENOVIA portfolio are really all categories of business application on top of the platform. And SOLIDWORKS is also having a very strong dynamic as well as SIMULIA and GEOVIA. About customer decisions, the takeaway, if we look at year-to-date and if we take a perspective of 2016/2017, is that the question for our customers are not anymore, do we need to go platform or not. I think they got the point. They want to go there. It's the speed at which they are going to adopt and deploy. The Scania decision is a very significant decision. Scania, as you know, is a part of the Volkswagen Group. It's a huge group, and we are -- it's a significant customer for Dassault Systèmes. So all the brands that they have -- 12 brands that they have and the very significant importance of Scania is they do extremely complex configured solutions for the market, all tracks are different. And they need really a platform to run that business. Many of those companies are using legacy software, which are most of them 30 years old, almost all players in that sector. And we believe that this is the signal that the modernization of the digital infrastructure for those companies is coming, and their decision is a very innovative decision because it's about adopting the platform for modular architecture, reuse and adaptation to different market segments that they serve. So very important evaluation that was done over the last 2 years, and we are very pleased to see that as a strong signal for their platform validation for extremely complex program like this. SOLIDWORKS, double-digit growth Q3 and across all geos. We continue to expand the footprint with SOLIDWORKS. And I remind you that we have explicitly said mid of the year that the SOLIDWORKS connection with the 3DEXPERIENCE platform is going to be -- it already exists, first of all, because we have customers with the extend platform of SOLIDWORKS, but new steps will be presented that SOLIDWORKS World, 2018 in February, where we're going to show how the 3DEXPERIENCE platform is powering those kind of customer, which we believe is now very critical for them to go and adopt further products on their product family. We have added 15,000-plus new clients since the beginning of 2017, so it's truly a strong dynamic. If you notice, we believe we are building an industry standard. When we started this 20 years ago, it was the race between Inventor from Autodesk and SOLIDWORKS. I think today, there is a clear winner, which is SOLIDWORKS. The Solid Edge is not anymore a regular player in this industry. So we believe that the footprint for 3D adoption which is needed for many mid- small-sized companies is a significant position to take on this market, because we believe that those customers are going to grow and they need 3D design or 3D printing on many other processes. So the world is going 3D, it takes time but we have to be there. And I think we are there for the right things. You also noticed that the average license price is well sustained over the last 20 years, not the last 20 quarters. 80 -- 800,000 commercial users. We think that is this dynamic is positive. The SOLIDWORKS 2018 was just introduced between August and September and it's extremely well received. It's really focusing on 4 areas designed to manufacturing, distributed data management for collaboration between small- and mid-sized companies and integrating the analyst tools inside, and, of course, integrating electrical and IoT to take advantage of those solutions for what we call mechatronic products. Here is a quick illustration of the SOLIDWORKS 2018 scope and capability. [Presentation]

So the takeaway from that is we believe that when you talk about Industry 4.0, the Fourth Industry Revolution, if you don't digitize design, you cannot digitize manufacturing, because you do not have the right source of data. That's the first point. The second point is that clearly, we see that many of the players in the landscape have completely forgotten that design is important and it makes the difference. So the 2 categories that we have, the power of CATIA and the power of SOLIDWORKS are foundations for every other businesses. There is no way in the future where simulation, manufacturing, additive manufacturing and even sales content for marketing will come out on the digital environment if the offering is not right. So with this dynamic, we have continued to invest and we are investing also in the GEOVIA side from that perspective to do geo modeling and natural resources for mining, for example. We are back to growth with 12% growth excluding exchange rate with GEOVIA. It has been a hard time in the last 3 years on that brand because of the reasons you know. But we are also expanding the portfolio to not only address the question of natural resources, but the underground for cities and underground for many, many other projects for construction on infrastructures where there is not many players in town. So that we have a diversification to serve Architecture, Engineering & Construction, Energy Process & Utilities as well as cities with GEOVIA. This is an example of win with GEOVIA. We're developing accurate 3D geological models for mineral deposits. It's a great showcase. This industry now, with the improved situation in our business, will also digitize progressively and we have to be there. So now about industry highlights, I would like to cover a topic on the High-Tech performance. As you see every quarter, we take one of the strategic sectors. The revenue is up 26% excluding exchange rate. We cover things on consumer electronics to security control, software, computing, communication, contract manufacturer and services, technology suppliers and semiconductor. So it's a full ecosystem. There is no other players doing that today in that industry. Why is it important? Because we think that it will be a further fusion between heavy industries and High-Tech and, on many aspect, autonomous vehicle, transportation, smart infrastructure, smart production system and energy, and that's why we are doing what we are doing. So you can see that we are growing our position in semiconductor in North America, Korea, Europe and Japan. Strong investment in simulation, in consumer electronics. We cannot name the customers unfortunately. And in the automotive electronics, as you know, a lot of things are evolving there. So the trends and challenge is really to address capacity to manage portfolio, capacity to create complex systems and capacity to really create new value chain in the system. Thank you, Thibault. Product modularity, therefore, compliancy on quality and regularization of standards. It's very critical now for personalized electrical systems. The investment we are doing in simulation is not only for engineering, it's also for the impact on people, for example, magnetism. Simulation is not only doing a good product, but a product that will fit well with the environment and especially, that will not hurt people. So there is really there a new game plan on -- I think, as you can see, you can -- in the semiconductor, for example, our platform is becoming the reference for IP management in that sector, and you see here what are the effects of those platforms. 18 out of the 20 -- 17 out of the top 20 semiconductor companies today are our clients. On SIMULIA, we continue to make progress. We want each of the discipline to be #1 in their domain and the assembly of the multiphysic approach being a differentiator. So we are #1 in structural simulation. We are now #1 in electromagnetic simulation. You can understand that if you look at the number of communication point in autonomous vehicle, it's an incredible number of communication points with extremely complex problems to solve for reliability and safety. Investment in fluid simulation, and we already talked a few weeks ago to you all about the Exa Corp. intended acquisition. And, of course, simulation is taking an important role in all industries, and we are using the platform to power all those solutions and integrate them. This is the landscape. We continue to use the same framework. It's not new. It has been published 4 years or 5 years ago, and we continue to fill in the domain of structure, thermal, fluids, electromagnetic control, geo physic, biology and chemistry on multiscale on the other side, and we think that this is where AI system and Big Data are going to play a big role. So for the fluids simulation, it's everywhere, fluid. An interesting case is the position of Exa Corp. in automotive. When you remove the gasoline engine, there is no noise in the car, yes, there is a lot of noise inside and outside, and the problem is to optimize that noise, and the most important thing is to simulate the CFD for the sector. So it's very interesting to see. I personally drive now an hybrid car, I will not tell you which one it is, but of course, it's designed by our software. And the experience is a unique experience, for those of you who will know that and this is going to evolve in a big way, where, when you remove the traditional engines, you have completely different design architecture, you have completely different kind of connection with external system but also noise structure, infrastructure are very different. So you see here the wide differentiation that we want to produce in every aspect, cities, home, energy, health, but also reservoir, mobility, infrastructure. And we believe with Exa and Xflow -- Exa Corp. and Xflow, PowerFLOW and Xflow, we can really change the game here. You remember that when we bought Abaqus, we made the bet for nonlinear and nonlinear is winning against traditional solution like Nastran and others, which were standouts at this point in time. We're going to play the same thing with Lattice Boltzmann. Now I'm not going to go through the math theory, but I'm still a mathematician. And I think that what we are going to do with those teams is uniquely positioned and will become indispensable in many industries which is what we want Dassault Systèmes to be, not just a player, but an indispensable player. So you have multiple examples here about how things are done, basically high-fidelity fluid on unsteady, high dynamic, this is the Lattice Boltzmann theory that applies the best and for traditional steady. Mildly unsteady, this is the Navier-Stokes maths, so back to your books, and you will see that those are beautiful mathematical model that need to be implemented well to make this a workable set of solutions. They have amazing customer base with outstanding degradation and we need to lever our channel to really go faster and provide solutions. Customers' feedbacks are absolutely positive, and it's up to us now to -- when we come together to take advantage of that. So you have the numbers here, Thibault will come back to you. It's a company that is doing $73 million, up 9% excluding exchange rate; 84% is recurring; small service revenue, 16%; and 300 top talented people, highly specialized. So we believe that this will be an interesting race with existing players, where many of those technologies are old technology. And with the 2 acquisitions, PowerFLOW and Xflow which are the name of the product solutions, we believe we can make the difference. So about CATIA and No Magic. This is also a very interesting, very specialized acquisition, and the focus of this acquisition is very special technology to master system complexity. They have the world's best clients managing this kind of system and system of systems in the defense, aerospace and also for autonomous vehicle. We plan to use that technology in 2 direction: one, in the world of CATIA for extremely connected, sophisticated systems, what you call system of systems; and also, we were planning to use MagicDraw, the name of the product, to do enterprise modeling. If you look at the press release, you will be probably astonished with some of the customers there. And in the case, for example, of aerospace customers that I will not name again, they are using for both managing the system complexity as well as enterprise modeling for process and organization, 2 types of usage which are very different. Some illustration of their clients here, NASA, Boeing, Porsche. You have more references in the press release. Please look at them. It's a small company today, $12 million. They have 170 top specialists in the domain. Those are highly specialized talents. We need them because our customers are really looking for breakthrough for autonomous vehicle and for safety in many sectors. So we are very pleased with that. I have to tell you that this privately owned company, they made the decision to join Dassault Systèmes before we agreed on price. And they refused to talk with anyone else because they said our home is Dassault Systèmes. It's an interesting situation, but it's a sincere one from their side. They do want to be with us, because they see how we can play together to create new connected world of experience. Many people are talking about IoT, and many people are Internet of Things. I believe that the basic infrastructure for communication between objects is going to be commodity and that the problem is above that line, because it's going to be standardized like TCP/IP and HTML and it's going to be done by standard definition as well as machine and equipment providers providing these basic layers. But the question is, what do you do above this layer? So we are really clearly positioned at Dassault Systèmes to create Internet of Experience, not Internet of Things, which is how do you orchestrate those objects together to provide a cool experience for the citizen, the consumers or the business. And that's the focus we have. And this is where No Magic is playing a role. So the -- really, the adoption of standard language like UML, SysML and UPDM, those are well formalized languages for system. So we are building the solutions above those well-accepted standards across big firms, across industries and this is what's going on right now. So with that, I will give the floor to Thibault to give you more insights about the exchange -- the currency effect as well as cash flow and other important parameters for you. Thank you very much.

Good morning. So a few figures. As Bernard said, total revenue was up by 6% excluding currency impact in the third quarter; 7% if you look at the first 9 months of the year. And from an organic standpoint, total revenue was up 5% for the first 9 months of the year. New licenses were up by 12% in third quarter, and 9% if you look at the first 9 months of the year, where the recurring revenue was up 6% in third quarter and 7% for the first 9 months of the year. Services was, as you have seen, the weak spot in the third quarter, software revenue was well aligned with the high end of our guidance, but services revenue was not. And it was not for 2 reasons, essentially. The first reason is really our willingness to continue to transfer more services to system integrators. This accounted for roughly half of the weakness, and we also experienced a few delays in contracting. In some areas of our services, customers are very demanding in contracting, and so the negotiations of these contracts is taking time, and this is something that happened in the third quarter with some additional impact on top of system integrators. However, we need to keep our resources in services because when we look at the many deployments that are ongoing for the 3DEXPERIENCE platform and other software as Bernard highlighted, we can also see how system integrators are going to request our resources in these engagements. So the message here is, it was difficult to manage the margin for services in third quarter and even fourth quarter because our top resources are going to continue to be very useful. And so for next year, I assume that based on the many engagements we see, there will be again some growth in our services line. So for that reason, the gross margin of our services went down to 10.4% in third quarter compared to a better performance 1 year ago. But if you look at the margin as a whole, we were able to manage it. And as you can see here, the 32 margin -- the 32% margin we had -- thank you, François -- was equivalent to last year, but this was actually a negative of 30 basis points caused by the strengthening of the euro against all other currencies, essentially, and an improvement from an activity standpoint by 30 basis points. And the acquisitions, interestingly, didn't weight on our margin in third quarter. EPS was up 2% at EUR 0.64. Actually, EPS was very consistent with our guidance. We had a guidance, whose high end of the range was EUR 0.65, but this was with currency rates of the guidance. And as you know, the euro strengthened more than forecasted. So in the actual rate of the quarter, the high end of our guidance was actually between EUR 0.63 and EUR 0.64. And by the way, the consensus as well, there was a consensus at EUR 0.65, but the consensus also was not at all the actual rates in currencies as the ones we had during the quarter. So we were also slightly above consensus from an EPS standpoint if you exclude the currency movements. And you can see that the EPS, if we were to exclude the currency impact, was up by 8% in third quarter. Cash flow was very good. The third quarter of cash flow was roughly equivalent to last year. But looking at the first 9 months of the year, we are up by 28% in cash flow from operations at EUR 672 million which, I think, is a solid figure. Looking into it more in detail, you can see that the bulk of the improvement in cash flow is coming from exactly the elements you want them to contribute. The net income adjusted for noncash, EUR 515 million improvement, and the decrease in trade accounts receivable that we continue to manage, EUR 229 million. No movement on unearned revenue. By the way, the unearned revenue at the balance sheet at the end of third quarter, if you look at the evolution of it, ex-FX was up by 6%, very consistent with the increase in recurring revenue. And the other elements were not truly a factor. There was no true decrease in the income taxes payable in 2017, so the first 2 lines are truly the ones which matter. Now, the update in our financial objectives. Before stepping into financial objectives, a few more elements on our acquisitions. Exa, we signed towards the end of September, the merger agreement. The tender was launched on October 12 at $24.25 per share. That's essentially a fully diluted equity value of $400 million. The transaction is going to be completed when stockholders will tender their shares and when we will obtain the regulatory approvals. We've got actually the U.S. approval from FTC, but we still have to get the approvals from Germany anticartel organization and the Austrian one as well. So the acquisition should be complete towards the end of Q4, potentially end of November or early December. And we will take it into account, of course, when we will release our guidance for 2018, when we release our Q4 results early February. The acquisition of No Magic is not going to close this year because we will have all the same approvals to get, FTC, et cetera. And based on the work they do, we will also have to go through the CFIUS approval, which takes a little longer. So No Magic acquisition should, in principle, close somewhere in January of next year, hopefully, in time for us to include it in the guidance for 2018. So what are we doing with the Q4 objectives? We are reconfirming new licenses and recurring. We are also impacting services for Q4, essentially in sync with what we saw in third quarter of this year. We're also adapting our currency exchange rate. We cannot keep the euro at $1.15, so we are basing it at $1.20. As well for the yen, the yen has a marginal impact because we have hedges in yen, but we also update it to JPY 135. But we also have other impacts which matter, like the won, the renminbi, the Australian dollar, the Indian rupee. Essentially, the euro strengthened against all of them. And since we are in London, I should also mention the U.K., the British pound. So here is the bridge of the update for the full year of our objectives. You can see the impact of FX in third quarter was EUR 10 million. Based on the currency rates we are taking for Q4, it's an impact of EUR 24 million. The impact on services was EUR 9 million in Q3. And because Q4 is much larger in services, the same impact represents EUR 13 million for Q4. And we have some miscellaneous, essentially for rounding purposes, of EUR 4 million. So there is a new objective for reported revenue of EUR 3,185,000,000, EUR 3,205,000,000 in euros, and the EPS is impacted by $0.05, and the $0.05 are just coming from the currency adjustment we are bringing into Q4. There is no other impact on EPS than this one. So Q4 is expected with a dynamic new license growth between 7% and 13%, recurring between 4% and 5%. Maybe this deserves an explanation. We don't have the CST acquisition anymore in Q4. So this is essentially purely organic. We also had a lot of recoveries of maintenance in Q4 of 2016. You see here, you know the nice increase of 9%. So in reality, recurring is growing at, in average, 6%, and this is also due to the fact that rentals are flattish this year as I indicated before, and the recurring license component is growing, of course, slightly faster than 6%. Services revenue, between minus 10% and minus 8% in Q4. EPS up, between 7% and 12% ex-FX. So for the full year, you can see that we are keeping the same growth rate as the ones we were displaying in February of this year, 8% to 10% for new licenses with the recurring of 6%, and this is, therefore, unchanged. The services are now flattish. This is the modification we are bringing and the EPS would be up 5% to 7% ex-FX or 3% to 5% in reported currencies. With that, I think Bernard and I will be happy to answer your questions. François-José Bordonado: We'll take the first question from the room, maybe the front line of the attendants.

It's Adam Wood from Morgan Stanley. Can I just maybe ask first of all on the 3DEXPERIENCE side of things, the more PLM side of the business? You've talked for a little while about growth accelerating as we look through this year and into next year. And I think, Bernard, in the statement, you said yourself that the progress you're making with the platform is not yet reflected in the financial performance. I know the licenses are better this quarter, but that looks as if SOLIDWORKS was better rather than in the core PLM business. Could you help us understand what's going on there? Is that still a sales challenge for you that you need to educate the sales people and retailers better? Is there anything to do with product maturity? The difficulty of migration from V5? What is it that's holding the growth back on that platform, so the financial performance isn't reflecting where you are with the platform and what can you do to improve that as you look into the fourth quarter and into next year?

I think the 3DEXPERIENCE platform represents for clients the new generation of digitalization they are doing for their company. The situation today is they still use in the engineering and production a lot of fragmented solutions. We did a lot for the digital mockup process. We did a lot for the document based -- what with the PDM and PLM side. With the 3DEXPERIENCE, it is really a business integration. I think we have very large customers running through the experience today around the globe, extremely large installation. Also, we see that the scope of reach, the reach of new type of users, a new type of domain of coverage is significant. Why do I say that? To come back to more detail about your question is because it shows us the total accessible market, as we transform these companies from an IT-centric set of tools to a business platform infrastructure. So the [ carbon part ] are significantly bigger than what we ever could reach. The consequences of what I just said, is that it needs to be well organized. The road map needs to be well organized for clients. We are going there. I was in India, in Russia, in America in the last 4 weeks. All customer presentations were about the adoption of a business platform for creation, design, simulation, production on life-cycle management, and also service. I think the tip of the iceberg this year is Boeing. It's a huge decision. What we have communicated is not even close to what we think will be the real business. But it's too early to talk about those big numbers. But the reality is it's becoming the core platform for everything they do, including connecting with supply. So for us, it's a very exciting perspective and we need to stage it in a proper way. So that's basically the phenomenon going on right now. It's not like what we have seen in the past 10 years. Now it's going to take time. Of course, I know all of you want to know when this labor is going to come and it's a fair question, okay. It's a fair question. I think we are learning ourself. We are learning ourself about what does it mean and at which speed each industries which are going in a very different approach. And I think each year, quarter-after-quarter, we'll show you showcases, and we are going to establish the typical metrics of how industries can deploy. I don't -- I think, for example, in the Boeing case, they are very keen to accelerate across the company, because they have a top line and a bottom line objective that they believe they can only reach with them unified. And there are companies like the Renault Group where we have mentioned many times and I think you see the performance of the group becoming #1 in volume today, they could not achieve that without the platform, so we have a large installation continue to grow. I don't think we have the metrics yet about what will be the cycle for each of the industries and segments. In infrastructure, engineering and construction, same thing is happening. The game is changing. So in some way, I apologize not being able to give you precise data -- more precise data. I think, at the end of the day, the best we can communicate is what Thibault and I are communicating to you, okay. Here is the perspective for the next quarter, and I'm sure early next year, we'll tell you the perspective for 2018. Because at the end, it's the synthesis. I don't think the synthesis represents a reality of the field. I think the potential is far bigger, but we have to only speak about it when the numbers that you're looking for are there. That's basically the way I would sum up. We are confident for '14 to '19, commitment that we did on the 2014 to 2019 doubling the EPS. I don't think it would be prudent to say more on at this point in time. I prefer to do a solid track record on market share we are winning on many fronts. I think on almost all fronts. So at some point in time, it should be more visible. We said phone lines, so please, we go.

It's John King from Merrill Lynch. Just 3 questions if I can. Firstly, on China, maybe some comments. I think it was also weak in Q2. Any optimism there that the deals are starting to get done? Or perhaps what's the delay? What do you see in China? And then on the services side, I think -- well, I guess, first of all, this delay that we've seen, is it right to think that relates to deals that have already been closed? Or is that in some way impacting the licenses? And then, I guess, conversely, as you look into '18, you did reference, obviously, you've got Scania, you've got Boeing ramping up, is there actually an issue with signing up? Is that actually potentially precluding you from signing up any bigger deals or do you have the capacity there to continue to sign deals?

I will start with the last aspect. I think what Industry Solutions, the way they are now engineered, are bringing to us is the capacity to do more with less resources. It's very key. We need to be able to enable customers to implement large scale with less resources. And that's what the platform effect plus the Industry Solutions integration play a key role there. In the last 2 years, we have been in the learning process of making sure we can do that in large scale. I think we are there. And to your point, we believe that this lever of doing more with less or more with the same, because it's to continue to invest, so I don't want to imply that we reduce our investment, is about the quality of application. So I think on the latter part of your question, we are in good shape, and I don't think this will be the limiting factor while we still are, we believe, too slow in scaling up with system integrators because they are learning the new business of what means Fourth Industry Revolution. Not many of them knows it and understand what it is really. They speak about it but they don't really know what it is. Sorry for being so clear. On the first question about big deals -- about China. In China, there are sectors where the decisions are done by the people we know. I mean even if it's not state-owned companies, there are sectors where the decision is done by the government. AMD, even transportation and infrastructure, even if it's public sector or private sector for the operators themself, there are policies. We are extremely welcome in China, and I'm confident about long term in China. We are trusted partner, probably one of the best trusted partner in China and they know they need us to have an industry that performs. So I think it's just a question of specific situation related to election. Whatever were the political context, but I don't see more than that. And we announce core decisions in the recent quarters that should pay off. They're also learning on their side. On the service side, Thibault, do you want to...

Yes, the delays in contracting. So no, actually, these contracts are not closed yet. And to be clear, in our services, part of our services are very tied to software and -- but we still have another part of our services which is actually not directly linked to software implementations, what we do, for example, in terms of helping customers do brochures and videos, marketing materials for their product. And this is more here that we faced a few delays in contracting based on difficulties to reach closure on these agreements, but not in the other side where we are key for the implementation of software, right.

Great. So we've obviously seen some interesting customer referencing on 3DEXPERIENCE coming through. A couple of wins seem to suggest that you're making interesting progress at the Volkswagen Group. I mean, German automotive has perhaps have been a bit of a closed door. Can you perhaps help us understand, kind of how you're progressing there? And will the customers be quite different in that VW is a big group with many brands? Will they, kind of, have a procurement model that's kind of brand by brand versus Boeing which made a kind of maybe a more centralized decision? So as you take the platform and engage with customers, what's the kind of -- the buying behavior like? And how different is it as you then look across all your different verticals? And is that kind of why this is a more progressive shift? And secondly, just want to come back on the sales execution. I know in North America this sort of platform selling model has begun to sort of stabilize. It sounds like it's Asia where perhaps you need to sort of upgrade capabilities, and I know in China, you said it was not necessarily -- there was some macro, but -- so where are we in terms of, sort of, upscaling and upgrading, that some of the drag goes away? Are we in the final few quarters now before, or will it take a bit longer than that?

Thank you for the first question -- for your question. I will take the first part of it which is really the evolving decision process within large groups. As you mentioned, more the decision like what Boeing is making is because they have -- they are reaching a new maturity in terms of integration, and they want to enable this to happen across all divisions on all SKUs, all the kind of characteristics of products they do on solutions. They would not be able to go to this level. We have to remember that they were the first in the world to create the first ever digital airplane, and that was in '94 with the Boeing 777 when it was -- it has created a revolution that it has never stopped across all industry. I think they are doing it again. That's for the first remark. The second remark about the group you mentioned in Germany, I think the CEO was very clear about asking each of the big brands they own to really take the lead to increase their performances in each of the business lines that they do and letting less influence of IT unification. That's basically what you can read in what was published from the CEO statement of this big group, which is exactly what we need, because the business needs to decide for digitalization, not necessarily the IT. The IT is the enabler but the decision-maker should be the business. So we see that and the decision at Scania was clearly, clearly facilitated by the fact that the business made a decision to go with full speed platform. You will see more of that in the group with each of the prestigious brands. And we see that as a very positive aspect for us. In Japan, the Japanese industry is really -- is now at the point where they want to move platform. They are going their home now. Other clients in the energy sector, infrastructure, they are there. And they are putting their house in order to make sure that it can happen rather quickly, because the competition in the auto sector, for example, not my competition, the competition between themselves, is extremely, extremely high in terms of technology breakthrough for systems, autonomous vehicle, battery, energy management and many other things. It's a very different world, and we play a key role there. I think I don't want to do any preannouncement, but I'm confident that that's going to play big into our advantage and that's justified in the acquisition we did in the last quarters. So it's a very interesting moment, and I think should -- things should come together properly. Do you want to add something, Thibault? I can see you want to add something.

No, no. I think I cannot expand. But simply on the sales rep skills, it's an endless process in reality. So it is true that we are certainly paying more attention to, first of all, our sales academy, which is the ongoing process we have in order to educate our sales force. And in a few geographies also, replacing salespeople, leaders where we see the need do it in order to get the right skills for what is becoming our business. Things are going very well for SOLIDWORKS. For mid-tier companies, it's going well, and there are spots where this revolution with the platform has to be better understood. So we continue to work on that, but a lot of progress has been made already. François-José Bordonado: We'll take one question from the call. And after, that will be for you, Michael.

[Operator Instructions] We will now take the first question from Stacy Pollard from JP Morgan.

A couple of product questions, if I may. Firstly, SOLIDWORKS was very strong, double-digit growth for more than a year now. And can you just discuss more the drivers? And are you still taking share from a competitor which was maybe forcing clients to the cloud? Also, any comments you have on your cloud business and transition or what percentage of the business is transitioning? And then second question, kind of contrasting SOLIDWORKS to CATIA, and I know you addressed that in some ways across the geographies, but can CATIA, as a brand, can that begin to recover in Q4 and into 2018? And maybe what would your expectations be there?

The solid -- thank you for the question. The SOLIDWORKS dynamic is -- can be justified by multiple levers. First of all, this is far from being a saturated market, and there are still a lot of companies using drawings. So I believe we have still a lot to do for the many years to come. First thing, it's a reality. The second thing is you might notice that in -- the players in the landscape, for whatever reasons, have, in some way, weakened their offer, and we continue to strengthen our offer. The proof points is the average license value. 20 years in a row, we have increased while all The Street price for everything else around was going down dramatically. That's a proof point, 20 years. SOLIDWORKS was at 5K in 1997 July, when we bought the company and it's above today, 20 years later exactly. So we have never cannibalized the CATIA world. And 20 years later, it's the case. In fact, CATIA is acquiring new type of market space in complex system-integrated systems. So the clarity of the strategy and the clarity of the investment is creating that environment. So non-saturated market, a differentiation product, the wide adoption by community, the weakness of the players and the [ personal ] adoption of 3D as a best practice for design, they are all explaining the dynamic across the world. It's not in 1 country or other. I don't -- I think this is a foundation for the next steps to expand. On the CATIA side, let's be clear. The CATIA dynamic is going to be massively levered with the adoption of the experience platform, because the new generation of 3DEXPERIENCE CATIA is revealed with the platform. CATIA and 3DEXPERIENCE is not V5 plus 1. It's a new world of design and system integration. So our first priority is the platform adoption, and that will lever so many new roles, domains of coverage that the phenomena I see -- the phenomenon I see similar to the Digital Mock-up adoption in the '90s but now, on the platform side. And of course, there is another dimension that I think we want to leverage, which is using the platform is a transaction machine. As you know, we have been clear that we are going to open the marketplace next year starting with certain type of services, but we will expand to really have transaction-based business as opposed to software-based selling, and this is going to be -- we have announced it, we have make it publicly visible and it's going to be made available in 2018. So those are the -- this is the dynamic. It's a very differentiating dynamic. And that's why I think we are building the infrastructure for the long-term growth on those parameters. There was another point I wanted to say, but I forget it. So Thibault, maybe you remember what I wanted to say.

I think, actually, you answered both questions. [Audio Gap] [indiscernible]

Thibault and then one for Bernard. Just in terms of V6 within the PLM license sales, what proportion of PLM licenses were V6? And then what proportion of CATIA users are now on V6 or 3DEXPERIENCE? Then, Bernard, my understanding on the Boeing deal is that they will be remaining on V5 CATIA and then consuming the 3DEXPERIENCE extensions. What's the platform strategy here going forwards? I mean do you expect over time that they will move to a V6-based data model for CATIA? Or that you'll support both data models sort of indefinitely?

So [ Michael ], the proportion of 3DEXPERIENCE in new licenses in third quarter was slightly higher than 34%, and this will continue to expand, of course. And the proportion of CATIA 3DEXPERIENCE in the installed base of CATIA is still relatively small, between 6% and 7%, so that's -- the flip side is the opportunity that we have for the future.

On the, what we call, Power BI, using the platform to integrate the current digital situation of our clients, whether it's V5 or other PDM systems, that's a huge value for clients. When the decision is made and that's what we are doing for the clients you mentioned, the lever is a significant layer -- lever, because you create the collaborative platforms that provides an environment where you can reach much more users and you can integrate, connect data that were never connected before to do analytics, to do manufacturing, to do more [ Modelica ] operation and many other things. To your specific technical question related to the V -- well, as you said, V5 data model on 3DEXPERIENCE, V6 data model, depending about the program, if you are -- to make it simple, if you are using the 3DEXPERIENCE platform to power an existing program, which is already done, and you want to really create a new version of truth, to really do the digital reference for life cycle management, we will keep the existing model as they are because there is not much value, if you have -- if you are maintaining an airplane through its life cycle to move to the next level, if you do derivative work or if you do new programs or if you digitalize the assembly line moving to moving lines or robots, then clients will upgrade to the new level of semantic. It's not an upgrade, it's basically the modernization of their digital infrastructure. Those paths today are extremely well defined. They are validated by clients, by large clients. They like it and they think they are unique as compared to anything else. In fact, we are applying that now, to apply it also on our competitor's data model. We have said clearly that we are connecting with [ OE ], with Autodesk solutions, with Siemens solutions to be able to [ swallow ] their data in the experience platform, because this is what customers want to have. So we are applying the same practice not only to us, but to the other solutions, including even indexing data from ERP system for cost and price analysis, for example. So that's part of those big plans. It's not a theory anymore. They are -- it has been validated with very well-targeted clients. In fact, last week, there was an amazing presentation from a large engineering company, McDermott, were they show their new generation of engineering process, this is an engineering firm, in oil and gas, where they are using the business platform, the 3DEXPERIENCE platform, on sourcing all data from everything they have; risk management, planning, taking data from Primavera doing dashboarding, risk management, collaborative environment, second to none showcase. That shows why this platform is not a PDM platform but a business platform. So clear path, well identified which is also a criteria for fast adoption -- for faster adoption. François-José Bordonado: Do you want to make a comment on the cloud?

Yes, I missed the question on cloud. We are having a very strong, good experience with cloud. And it's a lot of logos with small number of sites for each of them. But those clients are not evaluating. They are doing real projects, real industrial projects, whether in automotive, even I/O space, AEC. And our solutions, we think is going to be a significant differentiator. It's more to be talked about at SOLIDWORKS World 2018.

It's Charlie Brennan here from Credit Suisse. Just a few questions from me. Firstly, can I just come back to the momentum in the group? I think 12 months ago, you started to articulate 2017 as being a year of consolidation and 2018 as a year of strong growth. Now we're 12 months on and closer to 2018. Outside of Boeing, I'm not sensing that much conviction from you that we're going to see that acceleration in '18. Do you think 2018 still feels like a year of consolidation and now 2019 is the growth year? And just in terms of numbers, just a couple of follow-ups. Can you just break the license and the maintenance down into organic growth for us? I know you've given us organic software. And lastly, share-based payments continues to tick higher in the P&L. Can you just tell us what that share-based payments charge implies for annual dilution in the share count?

On the first part, I think about the conviction on the dynamic. I think in your question, there are 2 parts. There is one which is related to the numbers, and there is one related to the trends. So you see a schizophrenic attitude from our side, very strong conviction, but we are going to win. Okay? Let's make it simple. We are winning on the market. Now that's the conviction and we have the proof points. Who else is in town for aerospace? Who else is going to provide a digital platform, a business platform for even automotive? We have won all EV programs in the world, all of them, not all but 1, all of them. So we are on the lead for that. So that's the conviction from industry adoption, and I could say the same for energy sector or infrastructure. Now there is a fine line, and this is the schizophrenic part of it, is how does it translate to numbers? And that's a fine line where we have the [ proven ] side on my right. Mr. Tersant, who is saying wait a minute, we are building new business model, the adoption rate and so on. So that's why we do a nice cover in the industry actually. And that fine line, we don't want to cross it too quickly, because first of all, we don't want you to overestimate what we will deliver in numbers. I think we have a reputation to be reasonable. But I will not be reasonable on the fact that we are winning on all fronts, because it's a reality. And I challenge all of you to find out in those sectors if there was news flow from the other competitors. There is 0, okay? So the footprint and the conviction over a certain period of time to be the work standard in many of those industries is more than in our brain, it's in our stomach, and we see that turning. But there are a lot of things going on in the business model. Also, the build for clients is changing. We are talking about now build for client of 5 years. Numbers were published on -- or were leaked, depleting for the Boeing side. We never associated such kind of number with Dassault Systèmes in the past, but it shows that things are possible, but they don't come in 1 quarter. So please don't confuse the 2 lines, and I want to make sure because I think in '18, we will see even more of that. But I have no doubt that it has to be translated with numbers and that we have to show how the snowball is expanding, and we will be very open and prudent to make sure that we don't confuse you with those elements, but we are not hiding anything. We are really doing the fundamental work on the job for it, for making this a differentiator. SOLIDWORKS is even winning against Unigraphics. We are replacing UG with something which is 4x less in cost. So I think those things are not published, but it's happening and at some point in time, it should pay off. Thibault, on the other side, like proven side of the...

Yes, that's right. On the question on organic, so if you look at the first 9 months, new licenses growth is 9% reported, 8% organic, and the recurring is 7% growth, a little bit more than 5% in organic terms, with the rentals weighting on it for one good point, the weakness in the rentals which is something that will reach an end, because we know that we are approaching a tough point for rentals, which were actually driven by some reduction of installations, as I mentioned before. But I think after this Q4, we are going to start regaining growth in rentals. And the share-based payment, actually, the -- what we do, our policy is now that we have 2 mechanisms, but we are distributing performance share and these performance shares, we are buying other markets. So the performance shares part is not bringing any dilution in the share count, and we still have stock options for a smaller part, actually, where our objective is to limit the dilution but not fight it completely. And the distribution we are doing for stock options is less than 1% per year. So that's the situation.

How many performance shares last year?

I'm sorry?

How many performance shares do you have to buy back every year?

So what we have to buy back in performance shares is roughly -- let me just -- 800,000 shares. François-José Bordonado: We'll take one question from the conference call, operator. Nicolas, the floor is yours.

I have a quick question on margin. Actually, you are narrowing your full year '17 margin guidance at the lower end of your previous guidance range. While you are -- in Q3, you were at the high end of your guidance and why I think in Q4, your business mix should be more positive with more license and software rather than services. So what's happening there? Is it just due to ForEx? Or what's happening on this guidance adjustment?

Yes, it is right, it's only due to currency. But we felt that we have to open a little bit -- in fact, we were using around 31.5%. We keep -- if we deliver the high end of the guidance, 31.5% as a margin objective, fighting the dilution coming from currency. But you have seen in third quarter, 30 basis points dilution on margin, and the impact of currency in Q4 is actually more acute, so it is bringing more dilution. So we are working hard in order to compensate for it, but we feel safe to be consistent with the low end of the guidance, to have a margin that is going to be between those 2, 31% and 31.5% for the full year.

And just a quick follow-up, if I may. Just a housekeeping question is on the software revenue growth, you reported 8% at constant currency, 5% organic. It happened that we can see a 3 points impact from M&A, which seems quite high compared to just the CST acquisition. So is it just a rounding issue? Or am I missing something on M&A contribution on software revenue forecast, [ 3 ]?

The impact of M&A is certainly not 3 points. The 5% we gave you, and total revenue for the first 9 months of the year. But the impact of acquisitions on software growth in third quarter is slightly less than 2 points.

Okay, because we have seen the press release, and that is the 5% organic growth in software revenue and the 8% constant currency, but maybe it's just a rounding issue.

Yes. 5% is year-to-date. Yes, it's not Q3.

Amit Harchandani from Citigroup. 3, if I may. Firstly, with respect to M&A, you have made your move in the fluid simulation market looking at Exa. Could you maybe give us a sense or comment upon what are the other potential areas in your portfolio where you could see or target future acquisitions? Any particular areas that really appeal to you at this stage? So that would be the first one. The second one is when we think about the cloud, you've touched briefly upon it, but I'm trying to understand, many of your peers are undertaking this transition which is changing the way their margins flow through. Could you tell us for your business, as it stands today, longer term, how do you think about the margin profile as cloud gradually comes and becomes a larger part of your revenue mix? And just maybe finally, as a clarification, and sorry if I missed it, have you commented upon any potential impact from IFRS 15 introduction at the beginning of next year for the way you report your revenues?

Thank you for the question. On the acquisition side, in some way, if I look at what we presented to you in the last 2 years and 3 years in terms of, especially at the beginning of the year when we give you an update on the strategy: the framework, whether it's about industry we cover, 12 industries, 70 segments; whether it's about the framework on the spiral of innovation, from ideation to life cycle; or, more specifically, the framework for multiphysics, multiscale simulation in both physical goods or life science, in some way, you have the answer to your question. Because if you look at those, I've described 3 or 4 pages which are public, you look at who are the best, you look at what we do, you can almost predict what are the next targets. But we follow a lot of targets, about 3,000 in the world to give you -- it's not 200, and including very, very specific technology. I mean No Magic was probably not on your radar. It was on mine for a while. And No Magic is mission-critical and will be a massive differentiator. As a matter of fact, it was so funny to see one of our competitor announcing that both No Magic last week to our big clients when we were signing the contract. So we are going to continue that way, and the framework is clear, but I wish I could tell you more, but I cannot on this side. But there is no fundamental secret. We are really doing what we said for both the strategy, industry coverage, brand. I always said we want all brands to be #1 in the world and we are getting there. Today, SIMULIA is #2. We'll become #1. CATIA is #1. SOLIDWORKS, #1. We are doing the same thing for DELMIA digital manufacturing and the next-generation way of doing, producing things. It's a very stage on focused plan. That's for the first part. For the cloud, I think we have built a very nice model, but I will let Thibault talk about it. The endpoint is we believe we can do a model that is very consistent with the recurring model. And we are building at Power, packaging properly, the engagement model, the way we support clients because all this plays, of course, the way we do even our own cloud. You remember we did -- we sponsored a start-up called Outscale and now we have taken 80% of it. We are performing on our own cloud. We are delivering the same service as also a service commercially available for 20% less. Well, 20% is a good start for a margin of 30% or 35%. So we are very precise in building up a long-term sustainable cloud model that is consistent with the model we love and we know how to operate. Thibault might want to say something more on that?

Maybe. When we look at our peers that you mentioned, they have 2 causes for the margin dilution. 1 is the move to subscription, which may be or not on the cloud. In fact, the majority of it is not on the cloud. And they have also a division that is linked to the cloud, which is coming from, again, the fact that cloud models are subscription, but also the fact that they are, in fact, decreasing prices on the cloud for their offerings. So compared to that situation, we are not forcing our customers into subscription. We are keeping both options open to our customers. We have opened it to SOLIDWORKS as well. And so interestingly, today, we have a relatively stable proportion of customers choosing subscription versus perpetual. And roughly 20% of the new licenses are taken under subscription, 80% are still perpetual, which shows a strong indication, in my view, on the market preference and the fact that there's no reason actually to force a model upon customers. My second comment is the cloud. Yes, the cloud is, in majority, a subscription model. But what is, I think, particular with us is the fact that the cloud is opening us a new market to a large extent. When we look at the pipe we have today on the cloud, it is a true statement then that the majority of the prospects we have are prospects we wouldn't have on-premise. And so the result of that is we don't see a dilution on our top line coming from the cloud. So the top line is not driving actually a margin dilution, and we are keeping our software prices on the cloud unchanged. So we are not doing with a steep price reduction that others are doing. And finally, I would say that if there are years where the cloud is bringing some dilution in our margin, these years are 2017, and, to a lesser impact, 2018, because we are still at the beginning of selling our solution on the cloud but we are not at all at the beginning of having deployed the cloud for our R&D lab to develop the solutions. So we are reaching probably the maximum in terms of internal investment on the cloud, and therefore, probably the maximum dilution we are facing, because of the cloud, is 2017. Sorry, yes, yes. So IFRS '15, just to be clear, is not bringing, in our view, an impact on yearly results. It's bringing an impact on quarterly results, because we will have to recognize rentals. The part of the rental representing the license will have to be recognized up-front. In general, we have 1 year agreement. And so many renewals happen in the first quarter. So in IFRS 15, we will see more license revenue in first quarter and less in the following quarters for this rental part. But also, in 2018, we are going to continue to also produce non-IFRS, with our current revenue recognition such that you can compare apples-to-apples. And so we didn't produce the IFRS 15, yes, of course, but we will continue to give you some comparison apple-to-apple with the old revenue recognition method. So you will be able to see the performance in real terms. François-José Bordonado: We'll take the very last question from the call. Derric?

I've got 2 if I may. The first one, and maybe the 2 for Thibault, the first one is about the EUR 10 million expenses that we are not spending in Q3. What was behind these savings? Is this a question of investment saving? Or is this sort of accrual compensation for your service business -- in your services business? That's my first question; the EUR 10 million savings compared to the mid-point of your guidance? And the second question is about the maintenance revenue. So can you give us a bit more granularity on the drivers tending behind the growth of maintenance revenue, breaking down the price hike, renewals, attach rights of the revenue coming from [ PRSC ]? And do you expect the strength to be sustainable going forward? So i.e., as you said restated from rentals, stable revenues, it was maybe 7%, 8% growth year-to-date.

Okay, Derric. Yes, we did some savings in Q3 to compensate for services, as you noted. We know how to do it, I should say. If you remember -- well, I should not go back to dark years like 2009, but we know how, without restructuring, without reducing staffing, how to reduce expenses through a few measures, and this is exactly what we did. So more controls on travels for not customer-facing, and more attention in purchasing and negotiations and, I mean, a series of measures, actually, in order to save money without impacting our workforce in reality. There is also some automatic adaptation, which is linked to the fact that we have variable compensation, of course, in services, and so there's also an impact here on bonuses. The drivers for maintenance revenue, first of all, I think what's important is that when we look at the churn rate, the churn rate in 2017 are very consistent with our past churn rates, so no change in that regard. And if we divide our software business into 2, artificially, there is SOLIDWORKS and the rest. SOLIDWORKS is very consistent in churn rate, no increase, and it's a churn rate that is, in fact, in accordance with this kind of market, small companies, around 10%. And for the rest of our software business, a churn rate of less than 3%. So slightly less than 3%, very consistent with the past as well. So there is no modification on that front. So in recurring, the main driver was that rentals were relatively flattish, and the rest of the subscription, we have and recurring software revenues were actually very consistent with the increase in the installed base minus churn.

But, Thibault, my point was that you mentioned at the beginning of the year that 2017 will be -- will face tough comps for maintenance revenues because you had excellent performance last year. So I was positively surprised by your 7%, 8% growth that you post in maintenance revenue. And I was wondering if this trend could be sustainable?

Yes, when you project the part which is truly the recurring software, what you call maintenance, we don't like the word maintenance, but recurring software, I would say the rentals, the trajectory for it really is -- should be near 7%. François-José Bordonado: Okay. Thank you very much. With that, let's close this session. And I remind all of you who might be online that we have a conference call this afternoon. Thank you very much, again, and thank you for your attention and interest in Dassault Systèmes.