Holistic marketing and re-engaging old users | Will Meltzer from Moloco

[Music] it's a new year and a new PGC and we are back once again with William melter from maloko how you doing sir I'm very good thank you good to see you again he likewise we caught up in Helsinki a couple of years ago I think post lockdown yeah and now we're actually back back yeah we had to practice talking to each other again a bit no we were learning we were relearning how to do it but we're here now and it's it's busy as ever and and we're back but I won't make you repeat whole you know big malocco paragraph but just just the on line teaser what what's malocco what you all about we're a performance marketing business and we help clients with their user acquisition and retention beautiful that'll do us nicely so if anyone wants to hear the long version go to the Helsinki interview uh now as we're in the New Year we we're going to start on the slightly downside and go positive as we go but we can't ignore the negatives of of what's happened so a lot of layoffs last year the market feels like it's sort of Contracting a little bit and this year VC marting of VC spend

rather is going to be down we're going to see fewer games being made and people are a little bit cautious and a little bit worried as well like generally speaking um with fewer games means fewer games potentially to Market if that's the case how is it going to affect maloko and the way you guys move forward in 2024 um I don't think hugely um we work with probably about 100 100 of uh developers in air and then more than that gladly and we're we're moving down and I think the beauty of having a DSP is that we have strength and breadth our ability to deliver value to game developers of all sizes and all genres um so really we're we're working to expand where and how we can do the liver value and our biggest I'd say oxide we saw last year was IAA monetized uh with long life times um and hybrid casual tizers um so we really saw a huge growth coming out of those areas and uh we look

we looking to see that continue in 2024 well we spoke to one of your colleagues over in I think on PG beers we talked to them and they talked about the importance for you folks of holistic marketing the idea that it's not just about engaging new users and and you know doing UA through the roof but re-engaging older users do you feel that's an important thing is that something you folks have always been been doing or not something we've always been doing um igin our CEO worked hard to develop uh a DSP that would solve the sort of problem or crack the problem of uh delivering user acquisition to developers in the open web using machine learning um I'd say you know we valued at 2.5 billion um and we valued at that through doing user acquisition in yon web using machine learning um so and we continue to scale so we sort of C that tougher problem right because you got to find the users in the first instance and predict the value for any given title that you're working with whereas when it comes to re engagement um the value is

already being sort of generated and the user is almost being provided um to us in the first instance so the trick really is identifying existing an existing audience Who and the price of which we want to bid and bring them back into the title um which is a slightly simpler problem it's not as huge a market available to it but it's certainly something that's important to us and is's adding value now to a range of our larger CLI and we're looking to bring that to uh to V of portfolio well we spent a lot of last year talking about Ai and it was the generative side of it was kind of new back then and I think the the the public opinion has I think soured a little on the whole thing there's a lot of Suspicion around it and I think the layoffs and then people introducing AI replacement type things has fulfilled a lot of people's worst nightmares in some regard a company like yourselves has been using AI or ml for quite some time and have you felt anyone like suddenly realize in when they say

use machine learning they go hang on a minute that's AI you're you're one of the bad people or do you have to explain to them that it's used in a different way and has been for a number of years prior to this so we we normally get called bad for serving ads not for being machine learning and I have to have to indicate that that's how uh that's how your game gets paid for in the first instance right um we I I use the term AI to describe what we are as a business when I joined M loo in 2019 it was fly sold it's actually machine learning so there's a big differentiation and I'd say the machine learning piece through a deep neural network um and what it does incredibly well is identify patterns um and say inputs to generate specific outputs and they could be many and vared um and it's it's it's that piece that we're we're very effective on U and it's why we continue to deliver incremental value across the open web where clients are already running campaigns on that same inventory directly with the networks and and other

partners um I think with the AI P you've got generative AI you've got uh a lot of that and I think what it is clearly disruptive it is going to continue to be disruptive I don't think there's much can say about that um and I think there'll be positive and negative effects um I think what I'm interested in seeing is the connection of machine learning to Ai and and how you can use the the the patterns that machine learning can sort of identify with the generative examples of uh AI to then build a positive sort of flywheel of performance uh that would be exciting to me um you know it's a new interface AI right well I feel people are less hostile to it when it's doing a billion calculations that would take an army of people a year to do and that's kind of a positive you know good use do the busy work in very Commerce and then when it comes to the creative side of things I think people are more comfortable if there's a person doing it and and a job

going with that person ideally from from your folks what do you guys deliver once you've got all the ml you've got all the data the human aspect is how do you use that data in a meaningful and helpful way how do you go about doing that so so for us it's it's more a case of managing the models so no human really analyzes the data but what we do is we try and work with our existing models to improve the output outputs that they generate for our clients and we do do that with like some some serious brain some of the smartest people I'd say in the world actually who work with Maloo um and then from a from a day-to-day basis it's our growth teams it's our sales teams who work people who've developed great apps great games uh who want to bring them out to Market uh and promote them and increase their Their audience and increase their users in a way that's profitable for them and then in that in that instance it's like helping helping BR our clients understand what to expect

from working with us and from working with the tech that we're offering to them how they what they should it expect over time what they could do to maybe improve performance um to find performance to smooth out uh turbulence whatever it may be and from my perspec from a growth and uh and and a relationship perspective it's it's about translating what we see happening at the bidding end and the performance end into uh real actions that should or should not be taken at any given moment in our partnership with our clients right right okay does that make sense yeah it does make sense I'm interested in like especially with the with the ML and the reaction how reactive and agile is the system for example China has recently changed the way that it works it's decided right we're no longer allowing login rewards for daily users we're basically locking down a lot of the free-to-play model 10 cents Val you got like two billion wiped off it overnight or whatever like it it's and that kind

of seismic shift that suddenly happens can disrupt I imagine a lot of the models that have been built up over time not that that one particularly affects you but in general these big shifts like apple changing you know the way data is used how long does it take to kind of the models to catch up with the new system cuz learning means learning over time yeah it's a great question um so we require inputs to generate outputs um and I I think the philosophy of Maloo has been that at all times machine learning will perform a better job at generating the outputs required by our clients than than human uh human learning um and as inputs change we would expect there to be turbulence but then over time the adaptation would happen um and you know obviously you you youve raised the the Privacy piece we don't know what that's finally going to look like and we have a range of contingencies in play and we're uh and we're sort of continuing to develop

those contingencies based on various predicted uh potential futures um and when the dust settles um we'll see how it lands but we do think whatever the market looks like there's a very very strong place to machum learning in that market and and and that's the faith we hold of our business and our and our ultimately our Tech at the for the business what about the the fun and positive stuff so we're coming into coming into the new year um are there things on the horizon that you folks are looking forward to any changes you're making with the way that you do business or things you're going to be revealing in the next few months yes so I mean I I I talked about our experiments last year with hybrid casual and IAA monetized uh games apps that's growing rapidly so if there are developers out there who haven't previously been able to utilize machine learning in the open web beyond the world Gardens let's say um it's it's probably worth having a conversation with us now now um because things have

been looking really really good and BR in 2023 and we really to double down on that from a plant acquisition strategy um I think from the perspective of most people in the industry the the SDK piece uh with Maloo is really interesting um and we we would be the first 100% demand Advertiser Le uh platform with an SDK so what does that actually look like for advertisers and you know we're the only independent player left in the space um we would be and we are our model is basically looking to pass all of the savings and the efficiencies that come with an SDK onto the advertiser space um and no one's actually doing that um there's margin being driven at the SDK end um and the idea ultimately is to increase scale overall for all of our advertisers and in so doing that should benefit the Publishers as well and we're beginning to see that and that is growing rapidly so it's like we're a DSP

with anate we're absolutely not a network that's not our model we remain 100% Advertiser focus and that's very very exciting to begin to see those performance improvements take shape and how about the show today I imagine you you're booking meetings and that sort of thing do you get a chance to wander around and have a look or is it strictly business have a look right absolutely I love being back here so you meet up with your competitors your clients um your partners obviously the m P the data guys you know you it's a really good moment just to get back into the industry you've had Christmas you got a bit fatter it's time to uh get the braing back into gear again it's uh as you said uh new year new new PGC sounds exhausting but good I like it fantastic well thank you for talking to us as always thanks [Music]