Good morning. Welcome to Otello Corporation's Q3 2018 presentation. Today's agenda. I will start doing an executive summary, including an operational review on AdColony. Then our CEO of Bemobi, Pedro Ripper, will do an update on Bemobi, and today at the end we will do a financial update by our CFO, Petter Lade. The quarterly highlights. Revenue came in at NOK 64.9, which was down 10% from Q2 this year, which was NOK 72.2. Revenue was slightly down in AdColony versus Q2 2018 due to focus on margins, where we basically in the middle of the quarter decided to take away some unprofitable and less profitable deals. Revenue was also down in Bemobi, only due to FX, a weak Brazilian real versus the US dollar. Adjusted EBITDA was virtually flat versus Q2 2018 and down versus Q3 2017.
A cost reduction we did in AdColony in the quarter and increased margins was offsetting the lower revenue we saw from Q2 to Q3. If we compare with Q4 compared to Q3 this quarter, we certainly expect higher adjusted EBITDA in Q4 versus Q3 2018 due to lower cost base and more favorable FX and increasing gross margins. Operational review. If you look at AdColony, we divide our business into performance business and a brand business. The brand business was stable compared to Q2 and was profitable, and it has been profitable for the last six months. On the performance side, the business is still struggling, but we did two important things in Q3. We took costs further down, and we decided to focus on increased gross margin, which had a positive impact, which I will come back to later in the presentation.
On cost, we have now, compared to 2017, reduced the cost by more than 50%, and we are now having a run rate, which is a little bit below NOK 70 million. If you look at the general status, gross margin ramp significantly through Q3. This is very important in terms of going to secure profitability in AdColony. We did another cost reduction in AdColony. We merged tech and product organization. This has actually been positive change. We see our ability to prioritize has been better, and our ability to deliver fast on these selected priorities has also improved. In addition, we have reduced cost by moving more support functions, non-client facing support functions to Turkey, where we have our centralized performance hub and publishing hub. This has also taken cost down.
Now with these changes, we now have a current cost base which enable adjusted EBITDA break even at around NOK 50 million in quarterly revenue at 35% gross margin. Let's take a closer look at the global performance business. As you can see in this slide, you can see revenue has been falling from over time. If the first column is Q4 2017, then it fell also in Q1 and Q2 in 2018. As you can see, we saw a significant drop again in 2018 Q3. This drop is particularly due to the second half of the quarter because what really was changing in Q3 was the margin became significant under pressure. We even in July had margins around 12%.
We decided to take action in the middle of the quarter, we decided to increase margins because we cannot run a profitable performance business and a sustainable business at 12% margin. Here you can see what happened on the gross margin in the quarter. The first column is April. You can see we had margins above 20%. In May, it fell to 18%, June it fell down to 16%, and then we saw the low point in July with only 12%. We decided to turn off less profitable deals. The first impact we saw was in August where margins went up to 20% and we had full impact for our changes we did in August, where margins was close to 30% and we just finished October, the first month of Q4, where margins are a little bit above 30%.
This has a big impact on gross margin dollars, meaning gross contribution. In April, we were above NOK 2 million, it fell to NOK 1.6 million in May, NOK 1.2 million in June, and all the way down to NOK 1 million in July. We did the changes. We got some impact already in August. It went up to NOK 1.5 million. September was NOK 1.8 million, October, it's very close to NOK 2 million. As you can see, it has a very positive impact on gross contribution, and it's been increasing since the margin change we did in July-August. We look at how this impacts revenues, we can also see the same trends. The first column is basically showing revenue in July, which was NOK 1.8 million. In August, weekly revenue went down to an average of NOK 1.66 million. In September, the weekly average was NOK 1.37 million.
You can see then by week, starting from October, the first week of October was NOK 1.26 million, it was stable the next following two weeks. We see it start growing again in week four of October, NOK 1.4 million. Last week five of October, was NOK 1.5 million. First week of September, it was NOK 1.55 million. Now the last week, which ended basically yesterday, it was again above NOK 1.6 million. With steady margins above 30%, we're now starting to see early positive signs on weekly growing revenue, which is positive. This is, of course, our focus going forward, to keep the margins and continue to climb weekly revenues. To sum up performance, we took costs further down in performance.
Major change was that we basically took most of the non-client-facing support functions to our centralized performance hub in Turkey, which was the last step to do that. We have moved all these support functions to Turkey over the last nine months, and it's working really well there. The costs have been reduced also due to the changes we did in product tech. The gross margin has been turned around. We expect margins to stay at the current level it is now, around 30%, we are starting to see revenue to be more stabilized, hopefully, we can continue to grow weekly revenues going forward, particularly in 2019. If you look at the performance business, over the last four-eight months, it's clearly been some of the most positive weeks we've had in a very long time.
If you look at the brand business, brands decreased 6% in the quarter from NOK 33 million to NOK 31 million. Most of that small decrease was due to FX outside North America. The decrease came from our IO business, insertion order business, which was as expected. We continue to see a shift towards programmatic and brand performance. Brand performance and programmatic open marketplace now accounts for 50% of the business in the quarter. That will clearly be the growth engine going forward. If we look at our programmatic business here, you can see that with the result in Q3, we have a double revenue compared to Q1 this year.
We see strong growth with our key selected partners, The Trade Desk, MediaMath, Ubimo, OpenX, all grew from 70% to above 30% in the quarter, and it's simply due to we are working closer together with them, and we are optimizing our exchange towards these partners. We have big expectations for this, and we definitely see also growth in Q4 and particularly next year. What we are working on to make that happen is that we continue to improve our programmatic tools and offerings, better dashboard, better reporting tools, and more important, we have a really good roadmap on how we can make more accurate auction dynamics, which can help the business. We also, in our brand business, have invested into better display inventory. More than a year ago, we decided to not prioritize display inventory, clearly a mistake.
We rehired a former AdColony vice president who used to run our display supply, and he's now been working for us for two months. He's very experienced in the industry, and we already have interesting deals in the pipeline, which will have an impact, particularly in 2019. This person will also run our open marketplace exchange going forward. We continue to be very committed to a fraud-free exchange environment. Fraud is clearly the hottest topic in ad tech and programmatic at this time. We are working on renewing our tech certification. We are building new proactive fraud tools in our tech department. We are also integrating with more than one industry-leading fraud monitoring companies to continue to have both first and third-party protection for our clients going forward. Just to sum up on AdColony, obviously we are comfortable with brands going forward.
It's been positive quite some time. We're looking to how we can continue to grow, and that growth will particularly come from our focus on the programmatic open exchange business. On the performance side, we see clearly a positive movement now in the last eight weeks with the cost reduction we did, the focus on the turnaround on margins and also the small growth we are starting to see on weekly revenue. Clearly, the improvement we've seen in gross contribution will help AdColony to be positive, adjust EBITDA in Q4, and this is the first time we achieved that in a very long time. That is positive. I think the last eight weeks after the improvement of margin, the lower cost, has been some of the best months we have had in a very long time in AdColony. Opera TV, just a very quick update.
As previously communicated, we are in a legal dispute with the majority shareholder in Opera TV now called Vewd, MFC. We took legal action the beginning of the year, and we got a favorable verdict granted. This judgment about the sale of our shares in MFC was not appealed by MFC. MFC was also ordered by the court to pay a substantial portion of our legal cost to date, and the cash has been received by Otello from MFC. Now the next step is to ask the court to reopen the proceedings. Basically the court to determine the value of our shares, and so we can receive damages equal to that valuation decided by the court, after we have opened the proceedings. Just a few comments around Bemobi before I give the word to Pedro. Bemobi continues to strong online growth in both revenue and profit.
Otello and Bemobi have had informal discussions for quite some months with several stakeholders. Given the positive performance and global opportunities, Otello is considering strategic alternatives, including exploring a potential IPO or demerger of Bemobi. Obviously, when we entered the agreement with the ex-founders of Bemobi in April, where they converted their earn-out into Bemobi shares, a potential IPO or demerger was discussed in that. Since then, we of course, have discussed with several banks. As a result of that, we're now planning to meet a limited number of international investors during Q4 2018, for the purpose of considering whether an IPO or demerger of Bemobi could represent realistic and value-enhancing opportunities going forward. With that, I give the word to Pedro. Please.
Good morning. Good to see many of you again. Going straight to some of the results of Bemobi. We had an interesting quarter, in nominal results we actually had a negative growth from both a net revenue and an EBITDA perspective. Despite the negative growth from a nominal standpoint, we still consider it to be a pretty solid quarter, and I will explain a little bit further the reasons why we are still comfortable with the long-term projections we did about a year and a half ago. The main kind of adjustment that we think is worthwhile to understand the underlying growth is mostly FX. Although there are two specifics for Q3. Out of the basket of currencies that we operate today, the Brazilian real is still the biggest one.
As some of you have maybe followed Brazil, it's been in turmoil for the past year and a half, especially in the leading quarters to the presidential election. The overall effects against the US dollar, we had a pretty significant devaluation for the past year, which if we exclude of that, we still had a revenue growth of about 7% in the quarter, and almost a slightly flat, slightly negative, EBITDA effect in the same period. In addition to that, as Petter Lade will show in the deep review of the financial results, Q3 last year was a very atypical quarter. We had some non-recurring revenues, and you will actually see very clearly graphically, Q3 of last year was an off quarter where we recognized revenue from the past nine months on a dispute that we had last year.
With that, if we exclude these two things, we actually have a very consistent growth path, aligned with what we've been doing for the past, maybe two years, and aligned with what we kind of forecasted during our Capital Markets Day back in 2017. If we took the underlying kind of a base or the biggest KPI that we use to track, which is the number of our subscribers, we've been having a very healthy curve. Let me walk you through this. This is slightly different from the way that we have presented. Moving forward, we're going to be sharing one additional metric, which I think it's an important metric. Some of you might recall back in 2017 when we did this more lengthy review of Bemobi, we described that we have mainly two dimensions to our business.
We have the dimension of what we call our addressable market, and the way that we measure addressable market is every time that we go within a new country, and that we do a partnership with a mobile carrier, the mobile users of that mobile carrier, once the key service is alive, they become addressable, meaning anyone in that specific segment can buy our services. As part of our growth strategy, especially for the past two or three years, we went to many of the key emerging countries, and we cut deals with most of the key carriers. What that represents is that our addressable market grows a lot, meaning there are more people across all these countries which they have access to our service. That's one dimension.
The second dimension is we need to do marketing campaigns so that these people become aware and eventually they convert to subscribers. These are two dimensions we have talked about, which is first we increase our footprint, and then secondly, we work in finding the appropriate digital channels to convert this addressable market into the subscribers. Moving forward, we'll be sharing this kind of a red line, which shows what is kind of our penetration under our current addressable market. Maybe we could even include the size of our addressable market. As a matter, just to give you a sense, today, our addressable market is already pretty large in the sense that we already reach a big part of the biggest mobile carriers in developing countries. We have about 2.3 billion users that we indirectly touch through our offers. Our penetration now is about 1.1.
The critical thing here is we want to be growing, but at the same time, ideally, we want to be growing also the addressable market, which give us leeway for further growth. Understanding this graph a little bit, why the penetration went down is because if you go back many years, for example, 2014 and 2015, back then the operation was mostly Brazil. The penetration was very high. Actually, in 2014, we already had some of the Latin American countries. The penetration goes down because we expanded to more regions, but we were under-penetrated. The underlying, I think, key metric here is, and I mentioned that before, is in a mature market, and let's use Brazil as the most mature market since this has been the one we've been operating for the longest. We have a penetration about 8%.
If we see everything but Brazil, we have a penetration about 0.3%, 0.4%. The opportunity moving forward, and it's a little bit what we've been executing against is try to keep Brazil at a similar penetration, hopefully even moving a little bit up, but getting some of the countries where we are there the longest and have more distribution channels to a similar level. In that regard, it's clear that we had a pretty strong quarter. Our subscriber base now it's at 23 million. We were able to grow consistently across geographies, which is something also healthy. Just comparing year-over-year on that specific metric, we had a 26% year-over-year addressable. Specifically, I mean, not addressable, but subscribers. In terms of growing our addressable market is we are currently at about 65 operators in a little bit over 30 countries.
Just during last quarter, we launched four new carriers in different geographies. We're starting to expand more in Africa, which is a continent that we lagged a little bit behind, and we see tons of opportunities there. We are also going to other areas of Southeast Asia. We have a very bold expectation for next quarter. We have the plans to launch seven additional mobile carriers. Just understanding again, this means that our addressable will grow even further, and through time, we convert this addressable into more subscribers. That is one of the key metrics that we'll keep sharing with you. I think both are relevant. How much we have in terms of subscribers, that's a more short-term KPI, and having a low penetration, it's a KPI to show how much growth potential we had ahead of us.
The second variable, which I think it's very important for us to track, it has to do with the channels. Again, recapping. Addressable means our capacity to go into different regions and countries and convincing mobile carriers that our offer is suitable and desirable for that specific market. Again, on the second dimension, we have to find effective ways to get new consumers. Some of the overall methods, they are common across countries, but every time if we go to different country, they have slightly different economics, and they sometimes we have to tweak the channel. There is always a learning curve. One of the things that we measure as well, which is also a good underlying KPI, is our ability to generate and launch some of our own channels.
You might recall also from what we mentioned about a year ago, in the Capital Market Day, there are three ways that we acquire consumers. Typically, whenever we partner with a mobile carrier, the carrier also agrees they will promote the service themselves. That's actually a very good client acquisition channel in the sense that it's pretty much free for us. However, we have a strong reliance on the carrier, which can be, let's say, not very consistent. It's a little bit of an up and down. Again, it's good because it's the first gross ads or the first channel that kicks in, and it's a very cheap one. The second type of channels are the channels which we call the paid campaigns.
Pretty much we go out, we map who are the digital marketing players in that market, we buy campaigns, either on a revenue share basis or on a cost-per-acquisition basis. The trick there is always to map this cost against the lifetime value of consumers on that market. There is always a fine-tuning period. The good thing about this one, it typically kicks in fast. The downside of this one from a profitability perspective, is not as good as the first one because there is a cost associated to that. Last but not least, there's what we call the co-own channels. Channels that we develop jointly with the carriers, where we have, let's say, the best of the two worlds. It's, number one, it's very consistent. We know ahead of time how much new gross adds we're going to be getting.
Cost-wise, there's almost zero incremental cost. That's why specific in these slides, we're going to be following mostly this metric, which is our ability to launch in new countries and new carriers, these channels. We have mainly two of these type 2 channels. We've been talking about this before, the no credit, no data portals, which is the ability whenever a mobile user ends up without credit or data, their session is diverted to a smart captive portal that we control together with the carrier, and we use part of the space of that portal to promote our services. The good news there, we have about 10 of these portals already live. We have a very strong pipeline of launching yet another 10 for the next two quarters. We developed something else, which is an equivalent of a version of this portal but with voice.
It sounds very low-tech, but it's actually quite effective. The scenario is a user, again, prepaid, tries to make a phone call, runs out of balance, receives a message saying, "Listen, you just ran out of credit. Please do your top-up. If you stay on the line for a few more seconds, you get to listen about different services that the carriers can offer." Believe it or not, this channel is very effective in Brazil, but recently we were able to start expanding internationally, and since the concept is very applicable to many of the geographies that we have. Whenever we put these things together, the way that we measure if we are actually evolving or not is not so much in the number of launch, but in terms of our channel mix.
For example, today-- or actually going back when we started our international expansion, almost 100% of our gross adds or mainly our new subscribers used to come from our partnership with Opera Mini. Opera Mini, it's a partnership based around revenue share. Every new customer that we get from that channel, we stick about 70% of their lifetime value and Opera Mini gets about 30%. If you see what we managed to do, we reduced from 59% to 25%. There's two pluses on that. Number one, there is less reliance on a single player, which is also always healthy in terms of diversification. Since this is a paid channel, it's actually good because it help us to improve profitability.
The no credit, no data, which we barely started a year and a half ago, if we just look a year ago, we had 14% of our gross adds. Now we're already at 27%, and hopefully we'll be seeing this curve going up in the next few months. Again, why this is important is because that's not only more predictable in terms of understanding how subscribers are coming in, but also that's the best from an economic standpoint because we don't have to pay for that acquisition. This translates into better profit margins. Last is the paid acquisitions, which, as we went into more markets, we're understanding better the environment, we actually moved from 10% to 38%. Overall, we have a more healthy and diversified revenue mix, and the line in the middle is the one that we should keep the closer look.
This is a very key metric not only for continuous growth, but for very profitable growth. Beyond that, a few other updates which I think is worth sharing. Last time that I was here face-to-face, we had just done the new agreement with Otello. This new agreement had to do, as Lars indicated, we swapped pending earn-outs for me and some of the other founders into equity. As a result of that, we used to run Bemobi as almost two different P&Ls because of this historical earn-out agreement. Since this doesn't exist any longer, you might recall that we were very bullish about finding synergies to operate more as a single company. As a side effect of that now, six months after, a lot of things happened.
We announced a new internal organization, which I personally believe is much more streamlined and much more agile, since now we can work as a single P&L. We also agreed on a more integrated set of financials and accounting principles. Let me give you one example which Petter is going to show you whenever he goes into more detail about operating margin. One key example is our cost of acquisition, which has to do with the cost of bringing new users. Historically, for international, we used to count that as OpEx, and we always believed that it's more faithful to the nature of the business to consider that as COGS. Since the nature of this business is to keep bringing users all the time, now pretty much what we consider cost of goods sold are comprised of two cost items.
The licensing costs, which is pretty much the money that we pay back for app and game developers, that's always been there, but we included the cost of acquisition, given the nature of the business. OpEx is pretty much all the other fixed costs. The result of that, you'll see our profit margin went slightly down. Although we'll show you apples with apple comparison to see that it's pretty much the same level as before, but has no effect on EBITDA, because you're pretty much moving a line from OpEx to cost of goods sold. Now we have one criteria across the board and a criteria which we think it makes it easier to understand how the business is evolving. The other thing that we had done now that we're more integrated, we did some really new key hires.
This, as you know by now, it's a very quantitative and statistics-oriented business. Everything is very granular. Every acquisition we do in a very granular level. We have a new business intelligence leader globally so that we will look at our KPIs in a more integrated fashion. We also hired a new products director, which is helping us to push our portfolio further. We are also experimenting with new services and partnerships. This is something which we expect to see results starting early next year. What I mean by that is, whenever we expanded beyond Brazil two, three years ago, we picked one service as a beachhead to test the model. There was Apps Club, and more specifically, there was our games offering.
We've seen that once we have one service in a country, the resistance and the incremental cost to launch additional service is actually fairly low. We are getting some of the initial partnerships that we have tested in Brazil, as well as some new ones around different space, kids services and digital security services. We're starting to experiment across new geographies. Early results, they seem promising. We don't intend to do a breakdown per service, but we will update you more in terms of how this is panning out. The other one, which I briefly mentioned before, is this voice channel. This voice channel is a big chunk of our gross ads capability in Brazil.
Initially, I think we underestimated our ability to replicate this elsewhere, and we were actually positively surprised that we did a roadshow about a quarter ago with including maybe 25 different carriers, and they were very enthusiastic about it. If this really translated into real deployments, which this seems to be, that's going to be a very big push in terms of our incremental machine to get new subscribers. Last but not least, I'm not intending to play a crystal ball here, but presidential elections finally are over in Brazil. Very strange process, to be honest, especially for a Brazilian, which is the first time I've seen something like the country so polarized. The specific impact that we already seen since mid of October, three weeks ago, there was a reversal on the effects of Brazilian reals against the USD.
It's hard to predict how the trend is going to be moving forward, but hopefully by next quarter we're going to have some good news, if nothing happens. Today already, we got maybe a 10% appreciation of the Brazilian real, which most likely will reflect in Q4. Petter is also going to mention a little bit about our outlook moving forward, but without going much into details. Q4 is looking to be a pretty strong quarter. We started the first month as a really strong month, October was very good. We had a fairly solid month-over-month growth. November, again, it's too early, but it's still looking good. December, historically, is a really good month for us, more because of seasonality than anything else.
If nothing strange happens with FX and the trends that we had for the past month and a half, they keep going, we should have a good quarter ahead of us. Moving forward, I'd like to turn to Petter Lade so that we can go deeper on the overall financials.
Thank you, Pedro. Let's walk through the financial review. Revenue in the quarter came in at NOK 64.9 million. It was down around 10% compared to Q2. Some softness due to FX that hit Bemobi and also some of the carve-back we did on margin on the AdColony business took down revenue in the very end of the quarter. What's good to see is that as revenue has gone down, we've been able to offset that by lower costs. That is fully accounted for here. If you look at the adjusted EBITDA, Q3 versus Q2 is actually pretty much the same, NOK 1.4 million versus NOK 1.7 million. We take a restructuring charge of NOK 1.8 million in a quarter. This is all linked to the cost savings that we're doing in AdColony, and we are saving on annualized basis around NOK 20 million with this restructuring.
We also have a small net financial loss that's linked to FX, basically the US dollar versus our other currencies. If we look at the trend, clearly, revenue has been going down over the last year, but it's been relatively stable the last three quarters. If we look on the cost side, we continue to take down costs, and we're going to be relentless and continue to look at costs wherever it's possible. Of course, this is what's enabled the slight profitability you have on adjusted EBITDA to stay positive. Digging into AdColony. Lars addressed this well. The brand business is doing well. It's been profitable and we expect it to be profitable going forward. At the same time, there's a mix shift happening where we're moving revenue from the older business of IO over to more programmatic revenue.
That's something that's going to continue, but it's also going to have a positive impact on cost. It's not a bad thing. The programmatic piece is growing. It grew from Q1 to Q2, from Q2 to Q3, and it's going to grow again from Q3 to Q4. That's also where we expect to get some of the seasonal spend if it arrives late in the quarter. When we look at gross margin, the blended gross margin over the last year has been relatively flat, kind of in the 32.5%-34.5% range, which is really the mix of a brand business that's doing gross margin close to 40% and a performance business that's been going down towards 20%. Brand has basically propped the average up.
When we came into the middle of Q3, we found out that for our performance business, this was not sustainable, and we took corrective action. We're able to increase the margin from the beginning of the quarter around 13% to end the quarter with around 30%. While the blended gross margin for the quarter was 33.5% or 33.4%, we actually ended in September with gross margin over 35%. That over 35% is where we expect to be in Q4 as well. When it comes to OpEx, this has been a long journey. We started over 18 months ago with an OpEx base of $150 million. We took it down in increments. You can see Q4 last year, we were at $120 million, $30 million per quarter. We then had a goal of $90 million, which we achieved in Q1.
We've been continuing to take out cost, and we took out an additional $2 million in cost in the quarter from Q2 to Q3, and we've done more now in Q3 that will have the full impact from Q4, getting us to a new target of around $70 million in OpEx. All in, we've cut more than 50% of the OpEx in AdColony over a span of 18 months. That has meant that the cost savings that we've done have then offset the revenue fall that we've seen, and we've been operating the last four quarters with a small negative on adjusted EBITDA, but something we expect to turn around going from Q3 into Q4. For Bemobi, Pedro addressed all the key points here. I'll just echo what he said. The reported currency being dollars basically masks the fact that the business is doing really well.
We had revenue growth year-on-year if you net out the FX impact. 25% impact on the Brazilian currency when three-quarters of the revenue is coming from LATAM is making a big dent in the reported USD number. The technical point in terms of how we calculate for user acquisition cost, which on the international business before was in OpEx, which now we moved to COGS, which is the same thing we do in the LATAM market. We've actually, for transparency, kept both numbers in here. If you look at the gross margin, if we kept the same reporting structure as in the past, it would have been 71.4%, basically exactly the same as a year ago. Since moving this half a million dollar from OpEx to COGS, gross margin falls, but you also see that the OpEx is down the same amount.
Even if you were to adjust for that and put all the OpEx back, you'll see that the OpEx trend has been pretty good the last year. We also have, and I know Pedro has this, is cost conscious on his business. We want to get out as much margin as we can, as much dollars as we can, as we continue to grow. On adjusted EBITDA, same point as Pedro made. The Q3 last year it was a bit of an anomaly. You had quite a bit of a catch-up revenue in Q3, and you can see that is $1.5 million more in Q3 2017 than Q4 2017. That tells you a little bit about the size of the anomaly.
Underlying adjusted for FX, adjusting for all these one-time costs, we're actually looking at an EBITDA growth of around 30% in Bemobi from Q3 last year to Q3 this year. Over to cash flow. We started the quarter with $47 million in cash. We've been standing here for five-plus years, always having done earn-out payments. This is the last quarter. Q3 was the last quarter we did our final earn-out payment to Bemobi. That's the big red bar, the majority of the big red bar you can see here, which takes the cash position down to $30 million end of quarter. As you can see, this leaves us with no debt and no pending earn-outs. It's kind of a clean slate for us. Finally, the outlook for the remainder of the year. AdColony, this is all compared to Q3.
We do have a few things that's going in our favor now. The brand business we've been very happy with. This is 60% of revenue. It is profitable. Performance business was struggling, but we've taken out cost and we turned around margin and we're starting to see some revenue growth now, even though it is early days, but we are optimistic. We see revenue overall kind of flat to up in the quarter. We're still not halfway through the quarter, so we want to be a little bit careful. Spend typically happens pretty late, particularly in the brand business in Q4. We don't want to overpromise here, but it is encouraging what we see today. At the same time, we see gross margins up and well above where it's been the last year or so. That's good.
At the same time, we see OpEx down by a couple of million almost from Q3. All in with kind of revenue up or flat, gross margin up and cost down, it's a good combination that should lead us to EBITDA being up and also being profitable in Q4, which we haven't been for over a year. For Bemobi, we stand by what we said the entire year. 2018 is going to be better than 2017. It's just on a nice growth path. We have been hit with quite a bit of FX headwind in 2018, but it's good to see now that it's starting to shift. If you look at the average FX rate in Q3 versus where it is today, we have about 5% tailwind. Again, it's on a reported basis, it will just look nicer.
As Pedro say, October is very strong and Q4 looks to be a very solid quarter for Bemobi. That is the last slide. I'm sure you do have some questions. Also, people on the webcast, you're also able to enter your question in and we'll address it.
Thank you. Christopher here from DNB Markets. If we could start on Bemobi. You said underlying growth in EBITDA was 30% year-over-year. If you adjust for the revenue in Q3 last year and FX, what was the underlying growth in revenues?
Almost the same. Let me step back and explain a little bit where this math comes from. Mostly two effects. The first one is actually fairly easy to calculate, is just FX rates, which affects in the same proportion both EBITDA and revenue. The second one, specifically Q3 of last year, we recognize both revenue and therefore EBITDA for nine months, specifically Q3. There was two big disputes that for the sake of being more conservative, we didn't actually recognize until we got a full settlement from the both carriers. Pretty much these two things, they affect both top line as well as bottom line.
Then just going forward, could you elaborate a bit on what you see as kind of the underlying growth of Bemobi going forward as you ramp out to international markets and how much you think the OpEx base will grow? Because I guess you have to grow it somewhat.
Yep. With the new definition of OpEx, and by new definition, I mean, carving out COGS and leaving the fixed cost, the OpEx shouldn't grow very much. Understand it, just going one step deeper. Our OpEx is mainly two lines. The team, data centers/cloud costs, and then the rest is not really material, travels and this type of expenses. We staffed our international team, which is our biggest fixed cost already with the aim to be able to go to more markets and to be able to achieve what we want. Our expectation moving forward is that our OpEx is only going to grow in the hosting side because this is directly correlated to the revenue side, there's no way around that.
In terms of the biggest cost line, which is the team, we should be pretty much staffed, looking forward for the next two years. The side effect of that hopefully we should be seeing EBITDA growing at a slightly faster pace than revenue, because you're going to have a pretty much a fixed-line cost, which is not going to be going up, and then revenue going up. In terms of trying to kind of forecast moving forward, it's a little bit hard because as I mentioned to you, channel mixing affect a lot. Let me just for the sake of illustration, let me give you one example. If some of our own channels, they take longer to kick in, what that means is that they're going to have to have a higher COGS for cost of acquisition to compensate for the lack of our channels.
I think one of the KPIs that we're going to be sharing more and more with you is how well we'll be tracking in terms of deploying our own channels. In other words, if we truly transform this pipeline of 10 no credit no data portals within the next six months, you can expect not only a good revenue growth, but even a profit margin improvement. If this takes longer, most likely the growth will still come, but at the expense of slightly lower profit margin. For now, I think it's too risky to kind of predict what's going to be the mix. Ultimately, the mix is, yes, these portals will come in and we'll get the profit margins what we want.
Since they have a very long sales cycle and they depend mostly on the large carriers and their own internal processes, it's really hard to pinpoint when exactly each one of these will come live.
Just if you look at maybe one to two next years, what kind of growth targets do you have for Bemobi?
I think it's possible to high single digits or really low double digits. That's what we're aiming. Again, too many variables, especially as you talk about emerging markets.
Lastly, on Bemobi, you said you expect Q4 to be a strong quarter. Can you just kind of quantify what kind of growth translates into?
I think for now, let's say that it's still on the single digits, but in the high end of single digits. Again, let's remove FX out of the equation, because FX has been pretty unpredictable. If we remove FX, I think we should be on the high end of single digits.
Great. Thanks. Moving on to AdColony, the NOK 70 million OpEx run rate, when do you expect to hit that?
Four.
Q4. Great. I was wondering also if you could give an update on progress in China, and secondly, progress in terms of LTV optimization.
Yeah. Relatively technical, the second one. In China, business there for us is still relatively limited. You've read that there are some challenges also from the Chinese government in terms of allowing new games into the market. That's been a market where we haven't invested as much directly into the Chinese market, but we do see we get quite a bit of revenue from Chinese companies doing spend outside of China. It's still an important market for us. More so on the performance side. In the big scale of things, it's still fairly limited. When it comes to technical lifetime value calculation of users, I think the best evidence we have is what happened when we carve back margin. You carve back margin because the way you think is that if revenue's going down, how can you make sure that you get more revenue?
Well, you want to lower your price. You will lower our take rate. That's been what we've been doing starting over a year ago. We've had to give away margin in order to keep revenue up. What we then did mid Q3 was to carve back, and basically claim back all the margin that we lost over the last year. We saw initially that revenue went down. Thanks to the development we've done on the tech side, we see that we're actually now better at optimizing the revenue that we have. Basically, you carve out the negative part and you get better yield on the revenue that you have left. I think it's fair to say that if we'd done this 6 months ago, we'd not be able to get the revenue back to where it was.
It is improving, but this is not a black or white. This is incrementally improvement of getting better and better data.
Great. You previously used to share this slide on your SDK footprint in the top 1,000 grossing apps. Do you have an updated figure on that? Last time you shared it, I think it was about 600 of the top 1,000 apps.
Yeah
across App Store and Android.
Yeah. Our footprint is very much the same, but I think it's also fair to say that it's been easier as phones get smarter and bigger, and apps get bigger, it has been easier for others to put more SDKs in. We still have a very strong footprint, but I'd still say that it's less unique and you can do things outside of the SDK. Our footprint continues to be very strong.
Two more on AdColony. Revenue's down 42% year-over-year. Is that representative of your loss of customers as well? Do you still have the same brands in your customer base? If that question makes sense to you.
No, it does. Keep in mind that the revenue is gone both on the performance side and the brand side. On the brand side, it's been basically revenue shifting from IO business over to programmatic. When this revenue is shifting, we're not able to capture all on the programmatic side. Hopefully when we see customers like Disney or Coca-Cola moving their spend from IOs over to programmatic, and we see we make less revenue with them today than yesterday, we know that they don't spend less. It goes to someone else. We still get a portion of the spend, but not as much as we did. That's back to some of the tech issue. Same thing on the performance side. Game developers, they spend about the same today as they did a year ago, but this is also linked to tech.
What's fair to say is that if you look at our top customers, they aren't the same. It's just we get a smaller allocation because we haven't performed as well. Now, when we started to perform as we did now last few weeks on performance, we see we get more revenue in and it becomes this positive cycle where we bring them more users, they give us more inventory, we get more users. It's just turning the tide on that. The short answer is that if we start to perform better, the customers are there. These are not broken relationships. These are not relationships that are gone. It's just we do it at a smaller scale.
Great. Thanks. Just finally on the being breakeven at a NOK 50 million quarterly revenues base, what kind of cash conversion do you expect on that level?
Yeah. If you look at the capitalized R&D, because that's really the only thing that's outside. With the cost savings that we do, that would also go down. Capitalized R&D typically being around NOK 2.5 million per quarter. In Q4, it'd probably be below two. About NOK 2 million is what you can subtract. We need to, again, short answer, we need to be about NOK 2 million adjusted EBITDA positive to be cash positive.
Great. Thanks a lot. That's all.
Okay, let's see. We have some questions online as well. I'll do that first. That's the question on just the competitive landscape. That's from Arctic, the competitive landscape in the mobile advertising space. The question is if it's changed, if it's getting better or worse, I think it's very much the same. I think what we're seeing now is that we're getting better. Competition is always better. Competition will always be tough. It's fair to say that in some parts, we've been too slow. We've been too slow on some of the development of programmatic, too slow on LTV. We're getting better, and it's interesting to see that it works. When you see performance revenue growing, or you see the kind of open marketplace programmatic revenue growing, that's the best testament you can have that it works.
I mean, you will only get revenue there if it works.
Thanks. Christopher from DNB Markets again. Just following up on that, on the competitive situation and kind of the market dynamics. It seems to me there is an increasing trend towards kind of wrapper solutions and header bidding where all bidders can bid on a more similar level without the waterfall structure. Can you just elaborate a bit on how that affects you? Is it like you have already fallen so far down in the waterfall that this is net positive for you because now at least get a chance to bid on the same terms as all the other players? Secondly, more kind of a high-level question, but could you elaborate a bit on what kind of market growth you see for that corner business in performance and brand?
If you kind of fix all your problems, what should be kind of the underlying growth of that business?
Okay. Let's say the first question, that's an excellent question. In this industry, you have two ways you can operate. You do what's called a waterfall, which means that you basically rank your customers, and the first supplier, if it us, we get the first opportunity to serve an ad at a price point. If we don't do it goes to number 2 and it goes to number 3 and number 4. The advantage is that, if we bid in with a high price, we get the number 1 spot. The publisher is happy because they know what they're going to get. Let's say they get $15 eCPM is guaranteed. We can make $10 or we can make $30, but we still pay the same duty to the publisher.
That's why the publishers have, and probably will continue to have a waterfall where they basically rank and they sell the different spots from 1 to 10 or even further. The downside of that is it's not really an auction. It's not so that everyone can bid. If we were to go back, let's say, two years, we would favor the waterfall position because then we had a lot of kind of top waterfall position, meaning that we had first chance. We had first dibs on serving an ad. Now we've been outbid by many of our competitors. On overall, we are fairly low in many places in the waterfall, meaning that we are not even at the table. We have a ton of very good demand coming in from both our brand and performance customers that we're not able to play with.
If you ask me, do I prefer header bidding? Yes. I mean, we have more to gain from header bidding open up, because then we can bid everywhere. We might not keep the kind of small portion where we're number 1, but there's so much bigger market that opens up where we can bid on everything. You can see that in a kind of bidder section of the report, we elaborate a little bit on that. Your second question was?
Market growth.
The way we look at it now, this market is so huge. It's really difficult to measure growth. Like the game developers out there, one thing that happened is that you have quite a few developers that become really big. They do more kind of internal marketing. You're not really sure how to account for that spend. It's certainly a market that is not decreasing. As for brand business, if you believe the partners out there, they say it's about a 20% market growth for the next few years. For us, we should have a vast opportunity to grow, in the market that we're in, not being capped by the market growth.
Great. Thank you.
Any final question? Let's check on line.