Qt Group Oyj

Good afternoon, and welcome to the Qt Group's Third Quarter 2025 Results Presentation. My name is Heli Jamsa, IR Lead.

And with me today are CEO, Juha Varelius; and CFO, Jouni Lintunen, to present the results. After the presentations, we will have Q&A first in the room.

And if we have time left, we will move on to questions from the lines. Without further ado, please, Juha, the floor is yours.

Thank you. Good afternoon, everyone.

My name is Juha Varelius, CEO of the company. And as Heli was already saying, I'm going to go through the business performance on Q3 first.

Well, our quarterly sales was EUR 40.7 million and the decrease on 3.4% comparable currencies, it was flat basically. And year-to-date, we've been growing 1% on comparable currencies.

And our EBITA margin on Q3 was 10.5%. What has actually led into this and why are we below on our expectation is the fact that the market has been a softer longer than we've been anticipating.

So if we look at the -- what we've been missing is basically larger deals. The number of the deals we've been making this year and on a Q3 has been pretty steady and growing.

And so we've been doing more deals than we've been doing before and we've not been losing any customers. So the churn rate per se is at the same level that it has traditionally been, but the average deal size has been lower.

So what we are experiencing on a few segments, particularly where our customers are suffering, we are doing smaller deals. Our customers are reviewing the number of licenses they require and they try to go forward with the, let's say, kind of a -- with a minimum investment.

So there is still life in the market. We see the activity in the market.

We don't see any competing technologies. So we're dealing with the same customers.

We're getting new customers, but the deal sizes are lower. We've also experienced some shift from 3-year deal to 1-year deals.

And we were expecting that this -- on the second half, this market condition would get better. Well, it hasn't.

And we gave a profit warning because we anticipate that basically this same development will continue in the fourth quarter as well. So we were looking for that market demand would be stronger.

Now we don't anticipate that anymore. Of course, we could have been waiting and see what's going to happen on big deals on a Q4.

I'm going to talk about that in the future outlook more. But as we concluded that the -- we had a bit of same thing on the second quarter and now in the third quarter that we saw the bigger deals moving forward and the overall deal size being smaller.

We think that this outlook that we've now given is more on a realistic size. The distribution license revenue on third quarter was on the previous year level.

Overall, this year, the distribution license revenue has been developing favorably, and that's been on our expectations. So -- but on our license sales, we've been suffering both on a QA and Qt side.

Personnel-wise, we had 922 people on September 30 and year-on-year increase is 64. We are, of course, cautious on the cost side, but on the long term, we are still continuing our investment as planned because we don't see on a long-term vision any changes in that sense.

We did complete the IAR acquisition a couple of weeks ago. And like we've said before, where do we see IAR is that the -- we do have a more comprehensive product portfolio.

We see our strategy as that when we look at the development process of our customers, we want to be on the whole process. And with IAR, we are now with their compiler.

It's in the very beginning of this development process, so to say, so that when customers are starting a new project, the first thing they will do is that they will choose the hardware and then they start looking for a compiler. And after that comes actually how to develop software and so on.

So it will give us a benefit, of course, to be aware of customer projects on an earlier phase. It also gives us a benefit that we can be yet even a more one-stop shop for our customers.

They don't have to go and shop various things from various places. And specifically on our QA, our testing offering, it's a complementary or we can do cross-sell.

So when people buy the IAR product, also that needs to be tested. So we have a cross-sell opportunity over there.

IAR is well positioned in safety critical systems, which is also an area where Qt works. So we do have safety critical, you can see in the automotive, for example, quite a lot and on medical.

So they are typically the same segments where we work. We do have coincidentally also offices pretty much on the same locations throughout the world.

So we are operating in the same segment and this strengthens our position in embedded world quite a lot. So we are becoming a Nordic powerhouse going global.

IAR is selling perpetual licenses. They have started the subscription change on licensing model, which we did a couple of years back.

We are now reviewing -- we're doing a bit different scenarios that on what scale and on what speed we're going to be doing that transition going into next year. As of now, I don't have an info to give that what that's going to look like.

But of course, more aggressive you're going to be the effect on revenue is going to be greater on the short term and then on later, it will grow faster. But what is the kind of a speed of change, we haven't yet decided, and we are doing that study as we speak as well as we're doing the next year's budgeting and so on and so forth.

So we do look that the IAR is going to be a very complementary product for our portfolio, and we've started the integration work now. When we made a public offer, it was on a due diligence -- on a light due diligence.

Now we are going through the processes. We've started the integration work.

And like I said, we started the planning for next year budget. We started the planning for the subscription change.

And once we have those ready, then we're going to share more on that information to you on a later stage. And with these words, I hand it over to Jouni.

All right. Thank you, Juha, and welcome from my behalf as well to the earnings call of Q3.

I will dig into a little bit more details on P&L, income statement and balance sheet as well. Juha already discussed quite in detail already about the top line net sales.

We reported negative 3.4% net sales growth. And we see that happening driven by the customers' kind of cautiousness for most parts.

We are seeing the headwind from the FX, specific from U.S. and the magnitude of that was a negative 1.4% in the Q3.

So in other words, in comparable currencies, the net sales were flat year-on-year. For first 9 months, we are reporting a negative 1% reported net sales growth.

With the constant currencies, we are around 1% positive, so flat all in all. We did some flattening on the materials and services part.

There's still an increase of roughly EUR 100,000. That's the resources -- external resources that we are using for our customer consulting projects.

So kind of insignificant in any means, though. Our headcount, as Juha described, was up by 64 year-on-year.

And we have been adding resources into R&D, product management and also customer-facing organization during this period. And these are specifically the growth areas we see to be contributing going forward.

This headcount increase, it reflects very much in line to personnel expenses growth, 10% in Q3 or 9% for the first 3 quarters. Some increase in depreciation.

We have extended our -- in some -- extended the premises in some of the locations of ours -- in our locations and also in Finland during this year. So this shows a slight increase in that line.

The other operating expenses, the expense side, it's up roughly by EUR 2 million. That's for most part driven by the IAR-related acquisition costs.

And that impact is EUR 1.7 million now in Q3 or roughly 4 points in the EBITA margin, if you will. So run rate EBITA margin, excluding the one-off, would be somewhere 15% level, still close to 10% or 9% down from last year's.

The amortization, specifically from froglogic and Axivion acquisitions back in '21 and '22 remains unchanged, EUR 2 million a quarter, EUR 6 million for year-to-date. And this leads us to the EBIT of EUR 2.3 million or 5.6%, down by 13% points from last year's.

And the year-to-date EBIT percent is 13.2%. The financial items did not play that big a role now in Q3.

There was not that much fluctuation in the exchange rates. We are suffering from the headwind from the first half year from USD fluctuation specifically by EUR 1.8 million.

Our income tax was for third quarter, EUR 650,000, for first 3 quarters EUR 3.4 million, which equals to roughly 21% effective tax rate, which is our run rate and a good scenario going forward as well. And then this leads us to a net profit of EUR 1.4 million for the period -- for the quarter or EUR 13 million for the year-to-date numbers.

On the balance sheet side, we see a significant increase in cash balance. I mean that's the reason of the seasonality of the business and that shows as well in the accounts receivable, trade receivables bucket, which went down by roughly EUR 16 million from end of last year.

And this is driven by the seasonality of the business we execute. I mean, fourth quarter is always the busiest one with highest number of invoicing.

And then the cash will be collected in the first half year time. And then again, fourth quarter will be the busiest one.

We also see a reduction in the contract assets by EUR 3.9 million, which is a reflection that we have not been booking any major significant deals recently with multiyear deals with extended payment terms. So this is kind of contributing to cash flow, which is EUR 32.4 million for year-to-date.

When it comes to the equity and liabilities, there's very little movement on that in accounts payable or any other items. And I mean, this balance sheet obviously will be subject to change now quite significantly because of the acquisition of IAR and then that will be taken into account into Q4 finances then in February.

With these words, I will hand it back to Juha to go through the outlook and guidance for this year.

Yes. Thank you.

Well, we don't see any changes on our long-term growth prospects in a sense that the -- we do see all our customers planning for new products. They're going to be launching new products.

They are designing new products. We do see graphical user interfaces coming more and more into play.

We see on testing that the more and more software is being developed that needs to be tested to be robust. So in that sense, we don't see a -- on a long term, we don't see a whole lot of change on that.

However, we do see that the -- on the short term, what we see in our customers, there has been lay-offs in our customer base on different regions and segments. Basically, on all our regions, we see that our customers are on many cases on a saving mode, if you like.

And we do see that there is -- on embedded market, specifically, we see on consumer electronics, we see in automotive that there is a bit of a downturn on our customers on that. Do we see that, that's going to continue in the long term?

No, we don't. And do we see, like I've said before, that the -- is there a need to develop further new products, new product launches?

Definitely. So the number of devices will be growing.

The software will be growing. AI will be generating a lot of software.

And whatever software AI develops, all of it needs to be tested because we never know what the AI does. The market uncertainty, this is a, as I say, a great question that how long do we think that this is going to last.

And as a matter of fact, I was thinking and I was -- we were kind of hopeful and we were -- well, not hopeful, we were pretty certain that the second half would be better. Well, that's not been the case.

And we see that this market uncertainty on the embedded segment will definitely continue. How long?

At this point, I don't want to make that estimation. But the -- let me put it this way.

I don't see it getting any worse. So we don't -- I think that the cost savings that we're seeing, companies are doing it and I don't expect it to get any more challenging than it is as of today.

So we estimate that the -- we gave a profit warning and we gave the new estimation for this year, 3% to 10% year-on-year comparable exchange rates and margin between 20% to 30%. And as you know, the large part of that delivery will come on the fourth quarter.

We took a very -- well, if we were on a positive side, now we are -- our estimations, we've been on a conservative side on the -- that the -- how do we see on 2025. As we go forward into -- if we look into the next year, like I said, the basis what we have on our -- how do we prospect market going forward, we do expect this market to get better.

And we're kind of on a low end of this turn as we speak now. Well, it's the usual I already mentioned that basically on our segments, the automotive, consumer electronics are suffering the most, defense and medical, maybe the least.

So it's a good thing that we are on multiple different industries. If I look on the regions, maybe U.S.

been for us -- kind of varies that which region is the best. Probably U.S.

was suffering a bit more than Europe, apart from the -- on APAC, we're doing better. And well, of course, when you think of it, it's kind of no surprise that the Chinese automotive is doing pretty well.

But it doesn't kind of offset that how we're suffering in the other parts of the world. So that's basically the outlook we have.

And now if you have questions, please.

Felix Henriksson, Nordea. 3 questions, if I may.

It sounds like your customers are reducing the number of licenses that they have in use. What is the reason for that?

What do they tell you? Is it merely because of cost savings or is there anything to do with structural matters with developers becoming more efficient and companies seeing a lower number of licenses and that sort of thing?

Well, they have less developers. They're downsizing, right?

So if we look on the IT market, I think it's 2 years back, there was a shortage of developers. I mean, everybody were anxious to get developers.

It was very hard to find them. And that was kind of a bottleneck for IT company growth.

If you look now at the job market, I mean, there are developers unemployed basically at this point of time. So now it's kind of the opposite.

If you look at the big companies in the U.S., for example, that how big lay-offs there's been during the course of the, let's say, 1.5 years now. So that's one of the reasons.

Then the other is the overall cost awareness, let me put it this way. So it was very typical for our customers that whatever they had when they renewed, they renewed the same amount with the same deal like a 3-year deal and so and so many developers.

Now they are calculating exactly that how many do we need and they try to survive with the least amount of licenses. And then when they start new projects, when they are starting a project, they started with as small amount of developers as possible and try to go forward like that, whereas before, they started in a bigger scale.

So that's where it comes from. But like I said, the churn has not increased.

So we still have the same customers. They continue their development.

They are just more cautious on the spending. And also the number of the deals we do, so the number of new customers, that is actually even increasing than what we've been doing before.

And secondly, what about quality assurance? Did that grow in Q3?

Because it sounds like you seem to think that there's a bit of a structural tailwind from AI in that area. Is the demand on that front any better than for traditional Qt developers?

I would say that our QA business is the license sales is suffering a bit same things than on Qt. So the growth on QA has been slow as well.

We've also -- well, testing is kind of -- it's -- development is something that you either do development or you don't. Testing, you can always not to test and hope for the best.

So you don't have to test everything and completely and so on and so forth. So that is -- for customers, it's easier to adjust on a testing bit than on development bit.

But I would say that our license sales has been sluggish, both on QA and Qt.

Were quality assurance sales down year-on-year?

No, it's not down, but it's – yes, same roughly -- follows pretty much closely to what Qt is doing.

And then regarding the one-off costs relating to the IAR acquisition, they were for the full year, at least a bit higher than what I had anticipated. Will there be any one-off costs in 2026 from that?

2026 on IAR?

Yes, these one-time costs.

I don't think so, but you never know if there are surprises that we need to close down something or do something extraordinary that we are not anticipating. But I mean, these one-off costs are -- well, this money so far has been flowing mainly to bankers.

So what can I say? It's a big amount.

But the -- so do I anticipate any one-off costs on 2026? Well, at this point, I'm not aware of.

But of course, if there would be something that we would totally write-off, then there would be, but we don't see that as of now, no.

Antti Luiro from Inderes. I could ask on the lack of large license deals and kind of the drivers behind that.

What -- do you have any sort of idea where that comes from? Why are larger deals not coming in?

Well, they are being postponed. Yes.

So they've been pushed forward. The big projects, they are waiting to start.

So a bigger deal usually comes. So in our business, the first deal is always a smaller one.

Then there is -- the people start developing, then there is the expansion and that comes a bigger deal. And there have been postponements on those projects.

They've been kind of -- well, put on hold is a wrong word, but they continue with a smaller amount of developers, they don't scale up. That's the -- so the projects are not going away, but they go on a lower flame, so to speak.

Right. So does that mean that they are basically extending the time lines for getting those products out or...

Yes, they're doing -- yes, basically, they're doing with less, yes.

Okay. I could also go back to the discussion around having less licenses sold to the same customers and then optimizing the amount.

Drilling down to the AI effect because you could assume that developers are getting more efficient every year. You could see a recurring effect that companies downsize every year because they can do more with less.

Do you see that as a realistic risk for the market and your license sales volumes or do you think that the customers might just, at some point, expand the scope of their products because the AI can help them do more?

Well, if I look at AI as of now, where I see that -- you can use it is that on the web technologies or mobile technologies. So if you want to do a simple mobile app, for example, you can have the AI helping on that.

If you want to do kind of simple things that are very easy to verify that what they are, yes, you can do that with an AI. If you're doing any safety critical functional safety type of things, you can't -- or let's say, that the infotainment system on a car it's a very complicated system.

AI can't do that. Will it be able to do that some day?

Well, of course, you can take both views. Some people say that in a few years, we don't have to work anymore because AI is doing everything and other people are saying that, well, maybe not.

So I think that on embedded before the AI starts doing so much work that there is really less need for developers, that's kind of down the road and let's see how that goes. On top of that, AI is not very reliable, as you know.

So as of today, you can do simple things with AI. On testing, for example, you can do -- you can have test scripts written by AI.

And if that's not complete, well, then the test is not complete, but it's not end of the world. Those type of things you can do.

And you can use -- it can be a helper. But do I see that developers being so much more efficient that there'd be need for less on embedded side?

Not really. Do I see that what's really affecting our customers is the lack of demand and their profitability is under pressure and they need to do less.

That's more of the reason as of now.

[Indiscernible] Private Investor. So my question is about the competitive landscape.

What's going on in there? And are you seeing any sort of advances in the competitive technologies that might be impacting the license volumes?

No. Yes, that's -- and I can elaborate on that.

So we do have the usual suspects. We do actually see the -- I don't know if you've heard me speak before, but so we have Android on the IVI on the automotive, but there is not a whole lot of change.

There is Flutter that the Flutter was coming and that was kind of the recent emerging technology came from mobile and web and they were kind of making inroads into embedded. We don't see them that much anymore.

And as far as I know, they are more in a maintenance mode nowadays and they've cut back on their development. We do see Unity.

Unity is very good on the 3D. And on advanced 3D, if you want to have very nice-looking 3D, then Unity is -- it's a good choice.

However, it consumes more hardware. So you need to have more powerful hardware, more expensive hardware.

And it's fairly expensive on the -- per item cost on Unity is much higher than on our pricing, for example. So what we now see is that on kind of good times, we saw Unity being used also on kind of a middle tier automotive or middle tier cars, whereas now we see that customers are looking at cheaper offerings for low and middle tier and Unity can be used only on a high tier vehicles.

So basically, this cost pressure is, on that sense, it's working on our benefit rather than and more against Unity. So we don't see a change over there.

And we don't see -- at this point of time, we don't see any new technologies that would be coming into our territory. And like I said before, we don't have any customer -- our customer churn is the same that it's been for years.

And that's kind of a natural, I would say, natural churn that the project ending and whatnot. We don't -- we haven't seen that and we haven't seen any reduction on the number of the deals we're making.

So we're selling as well as before, even a bit better on a number of deals, but the actual sizes are smaller.

Waltteri Rossi from Danske Bank. First, on Q4, as the problem this year has been especially related to the large deals.

Do you expect Q4 sales to be under more pressure actually compared to Q2 and Q3? Because I would assume that there is even more of those large deals.

Yes. And so, yes -- well, yes.

And when we gave our estimation for our full year guidance, we kind of took that into account that there will be less, yes. So we were more conservative on that one, especially for that particular reason.

All right. Then about the license maturity mix once again.

Would you say that the 3-year license lower-than-expected renewal rate has had over or under 5% impact on this year's sales?

How much is 5%? I would say that on -- it's somewhere between EUR 3 million to EUR 5 million on the third quarter is the effect, yes.

On the third quarter?

Third quarter. I was trying to calculate what percentage, but yes, somewhere between that.

And how much would you say year-to-date?

That figure I don't have out of my head. But I knew that you're going to ask, so I looked at Q3 specifically.

All right. Last one about the underlying market conditions.

Do you expect the market to improve still this year or are we going to have to wait until next year for that to happen?

Well, yes, I was more hopeful when we were here on the beginning of the third quarter, I was expecting a -- obviously, I was expecting a better third quarter. That's for sure.

I was expecting the -- and that didn't happen, right? And so we are now more conservative on that and we don't expect much of a change on the fourth quarter and hence our guidance.

Are we going to see better next year? Well, at some point, this starts turning for sure.

And so yes, we are expecting -- we're kind of seeing that it doesn't get any worse than this, but when do we see it turn to better, on what particular quarter that will happen, it's -- well, I can't say that. And as you can see from our fourth quarter guidance, it's fairly conservative.

So we don't expect any big turn this year.

Jaakko Tyrvainen from SEB. I could continue on the AI and the related productivity gains on software development.

Are you seeing such kind of a trend or pattern that proprietary development would become, again, a bit more appealing for the clients or do they still need to trust in some sort of tools when developing the embedded solutions?

Yes, they are definitely going to be using tools. That's for sure, yes.

And I think that the -- like I said, on embedded development, you can use AI for writing test scripts on embedded development. You can use AI on design phase to give you trade ideas that what could be different kind of different kind of solutions and ideas, creativity ideas.

But the actual coding on embedded, I don't see that the AI will be there for anytime soon to replace the developers. No, we don't see that risk.

But on simple tasks, you can -- I've used the AI like that it's a great buddy -- it's your best work buddy. It can help you out on many – automating many simple tasks and whatnot.

But the actual coding, I don't see that on embedded for the foreseeable -- in many, many years that would change. I mean you can use AI doing simple mobile apps, for example, now.

But -- and of course, there is also the other side of the room saying that it's going to advance so quickly that we're going to all be surprised. Well, usually on these new things, as you know, is that when the change starts happening, it takes many, many years and people kind of even forget it and nothing happens and then the change comes later on.

But on this embedded coding, not in the -- well, foreseeable future is always kind of a scary word, but not in the coming years, let's put it that way.

Okay. And still using the word of AI, have you included any kind of AI features in your own products?

And has that improved the customers' productivity so much that they need less licenses perhaps? So are you basically cannibalizing the renewals by including such features?

No, no. I mean these embedded systems that people build using Qt, they are very complicated systems, very big platforms and whatnot.

So no, that's not the case. I think what we see is that the -- well, first to your question, yes, we utilize AI in many aspects in our products.

Is that downscaling the number? Is that affecting less license sales?

Definitely not. We do see that there is -- our customers are feeling the pain that they are not selling their products as much as they would like to and they have cost pressures, and that's where -- that's what we are seeing.

And those cost pressures are not only that they're selling less. Many of our customers are having high tariffs, for example, selling stuff in the U.S.

I mean, like the -- well, I don't know what's going to be the South Korean car manufacturers' tariff, but it used to be 25% before. Trum now visited and them, they've made a deal.

I don't know if it's now 15%. But I mean many, many of our customers are having a 15% cost increase on stuff they are selling to U.S.

So -- and then there is a bit of an oversupply on some industries and whatnot, and this is causing the overall friction in the -- on embedded business.

Okay. Then finally, on the license maturity mix, you mentioned that customers are perhaps now choosing a bit more on the 1-year licenses.

Doesn't this imply that you should have a pretty nice growth in your 1-year license base for '26? And could you elaborate a bit what type of a growth you are seeing in renewing 1-year licenses when going to '26?

Well, yes. Of course, yes.

I mean, on a short term, it affects us specifically. As you know, that our monetization model is that if you buy a 3-year license, we book it as revenue at that point of time as one goes.

So if people are buying more 1-year licenses, we book it at that point, which is obviously less than a 3-year license, right? But then the good thing is that the 1-year is going to renew next year.

So obviously, it's going to help us, absolutely.

And how much larger is the base now versus a year ago?

Well, I can't answer that. But I mean, the logic is right that it will help us next year, of course.

It's Matti Riikonen, DNB Carnegie. A couple of questions.

First, regarding your cost base at the moment. It's now clearly more elevated because you have done growth investments, but you haven't got the growth.

So you are going with a pretty heavy cost load into 2026. So how are you going to tackle that?

Should we expect lower margins in '26 because of that or do you think that just operating leverage would work in your favor in '26 to basically set it to the right path?

Yes. So you should not expect lower margins because of that and the operating leverage will fix that.

And in the case that, let's say, that this would be a permanent situation that the revenue will never ever grow, obviously, then we would not have growth investments and we would get -- we would still get the profitability, right? Now the one thing that will affect our profit margin next year is obviously IAR.

And that effect, I'm not fully aware yet and it depends on how aggressive subscription change we take. If we take very aggressive subscription change to IAR, then the revenue might be flat or even decreasing, which would mean that the IAR profitability would be diluting our group profitability.

I will give guidance to that once we've made those decisions and I know that what the effect will be. But the IAR profitability traditionally has been lower than Qt.

So obviously, there is potentially an effect. Having said that, IAR profitability will obviously improve because they are not any more listed company and whatnot.

So we're going to get some savings out of there. We have some ideas over there that how can we improve some of the performance on revenue even if the subscription is over there.

So that remains to be seen. But overall, that is the moving part over there.

I would not be worried about the Qt profitability. And by the way, of course, we're going to have one-offs the same type on the fourth quarter than we had on the third quarter.

Okay. Now regarding Q4, you have a fairly big hockey stick model for Q4 to meet the full year numbers because in the first 3 quarters you haven't grown at all basically.

So when the customers know that and they kind of want you to give them discounts at the end of Q4 to close the deals this year, not next year, so usually that creates a psychological kind of challenge. So is there a greater risk that if you stick to your discount policy and don't give any discounts then there would be a bigger share of those deals being postponed to '26?

Yes, that is a risk, yes. Of course.

Right. Then a question of your forecast model.

Throughout this year, we have basically been disappointing in each quarter. And your sales forecast model looks to be kind of broken or it hasn't worked like it did in the previous years.

So have you scrutinized what's wrong? And how can you improve the accuracy so that going forward your forecasts would be a bit closer to reality?

Yes. So we have 2 ways of looking into the forecast.

The one is that actually starts from the bottom up. So the sales -- each salesperson, they have their pipelines and they make the forecast.

They make what is their best case and what is the most likely case. And it's been built upwards from the pipeline and the sales makes their forecast through that.

And then we have through finance, which is more like a scientific model that they've been looking at the pipeline over the history and they've been -- they have forecasting model that this pipeline is likely to get into the sales, kind of an AI approach. And both -- basically both have been broken this year.

So what we did -- so if you look on the third quarter, for example, when I was here telling you what are my expectations on the third quarter, we had a pipeline and we had a forecast model done by the sales that this is the most likely out of this pipeline. And we do have -- and the same thing from finance.

So we know that if this is the pipeline most likely with these multiples, this is how it's going to turn into sales. And that was not the case, right?

And so when we look into more detail on the big bulk of things, that's how it's been moving around roughly there, but less. And then these bigger deals being missing over there.

So if we look at the end of the day on these numbers, it doesn't have to be -- the deviation doesn't have to be that many millions, right? So if you're missing some of the bigger deals over there, then all of a sudden, you are on the -- out of the scale what you were forecasting.

And that's basically been the -- what's been misleading us, say a bit. So we haven't been closing the pipeline as we did in the history.

And when you look what's been the reason behind on that on the pipeline, we've been closing the deals on the pipeline, but the deal has been smaller. So the average deal size has been smaller.

So have we been able to forecast that we have this amount of deals, are we going to close this amount of deals? Yes, we have.

We've done even a bit better than we've been expecting. But the deal sizes on those pipelines, they've been smaller.

So the deal has been closed, but on a smaller amount that's been expected. And that we need to adjust going forward in our forecasting.

So our customers have been closing the deals, but smaller than we've been anticipating. And then your follow-up question is that have we been giving discounts so that the deal has been shrunk?

No, we haven't. It's been less licenses basically.

And that's where we've gone wrong. So now on a Q4 or at the -- when we saw what happened on Q3, we took a more conservative look for our Q4, because in our old world, if we look at what's our pipeline for Q4, it's big enough for a bigger sales than we have, but we took a more conservative view how that pipeline is going to be closing.

All right. So that was actually partly an answer to my next question, which was that, did I really hear you correctly saying that you think that your guidance for Q4, and of course, this year is conservative?

We think that if it goes like the Q3, then this is the best guidance we can give. If we look on the pipeline and if we look at the -- we take the older history kind of the multiples, then it would be conservative.

But now we've seen 3 quarters that the pipeline doesn't close as it used to be. So I think that this guidance that we are now giving is very best we can give.

And I think that that's going to happen given the fact that we are using now the multiples that be into reality on the second and third quarter. So for the old world, it's conservative.

For this world, I think it's spot on.

Jaakko Tyrvainen from SEB still continuing. In the aftermath of the, let's say, Q3 and perhaps the year-to-date performance, which has been the most kind of a disappointing revenue stream for you?

Has it been the renewals or the new license sales or the quality assurance tools or the distribution license?

Well, new sales, definitely. So new sales has been the -- that's been lower than we anticipated.

We've had our challenges on renewals, but I'm very happy with our renewals team. They are doing a magnificent job.

And of course, they do have this challenge that people, when they renew, they're going to go through each licensees and there are reductions on some cases, but our customers are renewing. So the projects are continuing.

They are not resigning. They are not churning.

They do have a pressure on the renewals, but less so. But new sales being the biggest challenge for this year for sure.

And if I look on Qt and QA, I would roughly say that the same challenge.

And then finally, I know it's a bit difficult to have the apples-to-apples comparison in your case, but could you elaborate a bit what is the magnitude of average price hikes during the year?

Average price hikes?

I believe you have hiked prices.

Yes. We have -- I would say that not significant.

We've increased our distribution license. It's kind of -- it's a -- there are different buckets in our distribution licenses and we've changed the pricing on the different buckets.

But I would not say that not a huge impact on that, no.

There are no more questions in the room. I think we can check if there is anybody on the line.

No. So we can conclude the Q3 results and maybe some final remarks.

Yes. Thank you, everybody, for great questions.

I think we kind of covered pretty much everything. Like I said, we do have -- I want to emphasize the fact that we're doing the number -- the number of the deals we're doing is looking good and promising.

It's actually bigger than we've been experiencing so far. What we do see is that our existing customers and new customers are very cautious on buying the number of licenses, and we've seen deal sizes decreasing.

We haven't seen any decrease on our churn. And so we continue with the same customers we've had.

We don't see any new technologies or competition coming into the market on that effect. How long do we think that this embedded market downturn will continue?

Well, definitely, it will continue into Q4 and going into the next year. Do we think that we're kind of on the bottom of the downturn here?

Definitely. And do we see that we're going to be going forward upward from here?

Yes, but the timing is a bit of a question. Do we think that -- are we concerned about the next year profitability because we've been investing on a long-term growth for next year?

No, we are not. And we do expect the return on the normal profitability that you've been expecting to see from us.

And like I said, we're very thrilled about the IAR acquisition. We are now going through with them, different customers, integration facts and whatnot.

We are preparing a budget for next year. And depending on how aggressive we are going to go into the subscription change, that depends on what's going to be the IAR profitability next year and how that will effect on group profitability.

So that is the moving part and we're going to get -- give you more info on later once we've concluded that work. I don't expect that to take a very long time because we need to get going in the early next year.

So all in all, disappointing Q2, but we are very -- we think that the future looks better and we are in a good move to execute in -- towards better performance on the top line. Thank you.