Now it’s possible the ratio is a lot higher, which we’re also very happy because, yeah, that has direct impact on positive uplift because they’re simply such more and they have a chance to expose to end users more. So all in all, all that very positive. And I think the only thing that had we say, of course, it’s only at the beginning because now it’s only EU and it’s also a bit fragmented in many countries manufactures. So we feel that — we rather feel this is probably the pioneer of where we’re heading to and some even bigger markets, like U.S., which we are highly anticipating. So let’s see how this develops. And in any case, one is coming with growth because it’s — the incentive is large for us to move forward to.
Naved Khan: Okay. And then maybe a quick follow-up for Frode. Maybe just talk about the opportunity they’re seeing for marketing spend. It came in a little bit lighter, I think, in the first quarter. How should we kind of think about that going out? And do you see some new channels you can deploy ad dollars into?
Frode Jacobsen: So we had expected Q1 marketing spend to be very similar to Q4. It came in 2% below. So I think it was quite on expectation. And for the rest of the year, we continue to expect like $1 million to $2 million of incremental spend quarter-by-quarter. So keep ticking up as the year progresses. I think in terms of the channel mix, we expect to continue what we do today, which is to work with influencers and to turn, keep turning more of the spend into Western markets and other high-value populations, and then in particular for Opera GX and its global growth.
Naved Khan: Understood. And then for Opera GX specifically, maybe is the growth you’re seeing in some of these new emerging markets, is that primarily word of mouth, or how much of that is advertising and how does the monetization there kind of change over time? Do you think that it’s early days and it might actually improve as you become a bigger presence in a certain geo or does that — just give us some thoughts there.
Frode Jacobsen: Sure. I mean in terms of word of mouth and distribution mix, we do benefit from a strong brand once we have started to build presence. So that we also show a chart in our investor presentation to sort of illustrate the size of the organic inflow of users. But it does require investment to start building a presence and in the example of GX, raise awareness of that product. So we do — we definitely do that. I think the second part of the question was around monetization opportunities. We are in process of expanding the monetization of Opera GX. I think the Q4 to Q1 ARPU was almost flat, down a couple of percentage points, something like that, which is very strong given the seasonally strong fourth quarter. The product does benefit that the users it attracts outside Western markets are relatively more affluent than the average user in those regions.
But let’s say, if ARPU was constant and it continues to grow internationally, that of course would be a headwind. But fortunately, we’ve been able to grow the underlying ARPU faster than the, let’s say, the geo mix impact.
Naved Khan: Understood. Thanks a lot guys.
Operator: Thank you. Our next question will come from Vicky Wei with Citi. Your line is now open.
Vicky Wei: Thanks for taking my questions and congrats on the solid quarter. I’ve got one question. Since the announcement of adding experimental support for 150 local LM driven from 50 families to Opera One browser in developer stream, any color you could share which are the top five favorite models from the users you see on your Opera One? Any indication how some of these usage of models could help on future financial growth or will this be more enhancement tools with limited monetization? Thank you.
Song Lin: Yeah. It’s Song here. So I’ll comment. So first of all, I would say I think it’s more like — number one, I would say, this vision is rather a reflection of way in, but also many ones like us, many European leaders in the industry feel, well, the AI is heading to that it is always will be a combination of you have this very powerful large language models which are hosted by the guys who have OpenAI and the likes. But then also in combination of, I would say, local models, more like — well, I think a typical example is Llama and a few others. I think the reason just because — really just because there’s always the case that you want to give some help very fast. And then for privacy and other considerations, you do not want this to be uploaded to a cloud.
And also there are some cost considerations among those. So I think all the mix there for some actions that will be taken by remotely, and those are the actions which will have to be done locally. So I think that’s where we are heading to, and we feel that we’re quite happy and proud almost to be the major ones to go to that path I think really just because we are rather independent. We don’t have any more like possible other players, I understand there’s some other considerations, right? They want some other cloud or whatever. But for us, we just tell what is user need, right? So I think that’s quite important for us. And then in terms of models, I think the typical ones, Llama, of course, will always be very popular. It’s one of the most powerful ones from Facebook, from Meta.
And then you also have — Google has Gamma, which now they have — and even Microsoft has announced some local models. So we’re happy that this proving to be the right trend. And also if you — yeah, if you want, you can also try Llama 3, which is the latest model we just had. So we’re quite proud because it’s only coming out last week, and now you can actually try that on Opera browser and use it very easily, just a few steps. You don’t have to develop or whatever. So I think that’s the role we’re playing, and we’re quite proud of it. Yeah. So more like those are things that we’re happy to do. It doesn’t like — unlike Meta and a few others, we’re trying to do this in a very smart way. So we will not bring up a heavy cost, and it’s almost mostly upside just because essentially this makes our product more differentiated and hopefully retention will be high.
And if we solve the user loyalty, we hope that they will stick to us longer. So I think it’s a win-win situation because in end of day, what’s been consuming is users CPU and the electricity, his own laptop, which I’m sure he’s happy to take, and it also doesn’t bring additional cost to us. So I think all those are good things and good features, and we hope to remain the leading roles in that field within the browser.
Vicky Wei: Thank you, Song.
Frode Jacobsen: I can maybe chime in from the monetization point. I think being able to take these types of AI features mainstream, that can attract to us exactly the type of user profile that we monetize the best, right? And that we expect that these offline models will exist in combination with the online models like Aria because of the needs that require essentially updated information, access to product, access to other live things that won’t be captured in an offline version. So slightly different use cases.
Vicky Wei: Thank you.
Song Lin : Okay. So if there’s no other questions, I’ll just say that thank you all for joining us today. We have been looking forward to share this quarterly update with you. It is ultimately exciting times. And as you see, we innovate and evolve quickly to seize the opportunities ahead. There will be lots to talk about as the year progresses, so stay tuned, and thank you for your time.
