Reinventing Successful Game Launches - Jesse Lempiäinen, Geeklab.

[Music] [Music] [Music] first up I'd like to invite Jesse from Geek lab who's going to be talking about Reinventing uh successful game launches all right hello everyone uh let's kick it off with a Leverage The Power of Loy not that one this one a mandatory introduction slide so uh I'm yes

L um I used to work at roo as a product marketing manager primarily launching games uh new games uh and and and um trying to figure out what would be our our new strategies for a few few new Studios after which uh I I co-founded a geek lab dur at which we actually then again help help different game companies to launch games which is why today we're going to talk about Reinventing successful game launches and to get to that well I'm going to quickly share why we need to change so to to get a bit of context around the topic uh then what are the actual three main conflicts that that that I see in this uh this current way that we are launching games how can we fix it and then hopefully if we have time I'm trying to make this all into into 15 minutes or so so also practical examples on how you can actually do it to today or tomorrow cool why do we need to change so we all know making games is super fun and not only fun but it's a very very lucrative business

uh very solid solid business but unfortunately uh there's a ton of mobile games out there 700,000 more or less so uh which is why majority of the new games actually the launches fail so if if you look at top 10 uh like uh top Crossing in United States only one game was made in 2023 more or so monopol go and uh if we take a look at all these successful launches post at they they tend to have like two things in common one they're either from a billion dollar company or two they have very strong IPS or actually the third one would be that they have both of these so if you don't have a billion- dollar company or you don't have access to a very strong IP what what can you actually do we we kind of need to rethink how to launch games then and to get back uh to to to year 2018 while I was still working at Roo we were working around the this game called

Angry burst dream blast and and that's when we actually released it and it made it top Crossing to to USA 100 which is why we can call it a successful Game launch as well uh but what I I hope majority of you guys you guys know that when it comes to game development that wasn't our our our immediate success uh we had a we had a budget for for new games that we were doing we tried out different things and they failed and they failed and they failed until we had no money left and then we did dream blast and then money came pouring in obviously that's not truth we robio is a big company so this is just to make it a bit more dramatic for you guys but yeah in in short again as you as you saw we we failed multiple multiple times before I actually dream blast dream blast came out um but then uh as we we started geig laab we were in a bit of a different situation we had no money nor have we any any features so we we had to f figure out what what to actually do now

so what we did we actually built directly with um with all the developers that we work with uh we we tried out different things we build them and put them directly live uh and and a lot of these things they failed as well but uh but we were building them together with the developers directly for the for the product which kind of then um got me into thinking as we were building this practice got me into thinking like what are the things that actually are currently a bit of in in the way we are are uh launching games so what are the three main conflicts the first one is we say that we do it for the players but we're actually developing the games for ourselves uh what I mean by this is that this is sort of the traditional way of of building games that I I hope that most of you would recognize we have this typical flight path or whatever you you call it with different stages where the initial idea is the is the first one on the left and the Milestone to get

forward is kind of is this game fun and and does it have great Market potential you go to your studio leads or or if you are the studio lead you make the decision that okay this is this this ticks all the boxes that that we we've set and and we can move to production and you give it a production budget uh then you start to uh develop the game after which at the end of the production then you do some early validation with with um with the marketing assets and and then technical tests to see if it works and if it's if it's good you move the soft launch and at soft launch it's practically to to build up a profitable use acquisition uh machine so so that the lifetime value is higher than the cost per install and if that's good so then you kind of press the button and and you call it a global launch and then the infinite continuous live optimization kicks off but when we look at this this Frame the issue here is that actually this entire beginning is is all internal this is very internal process uh the initial idea again like that's

that's typically might have some surveys with the audience or whatever but like it's it's very very internal whereas just in here you actually start to develop it with the users themselves um and on top of this uh obviously the cost accelerates uh over time which means that if you if you only get to know like kind of what are the things that uh are not working and what what are working at during the soft launch uh so it's it's it's going to cost you a lot where as if you find those things earlier uh you can actually try twice so so fail late fail expensively fail early you can try try twice uh as an example uh at Roo with Roo we we tried this this uh puzzle RPG game called puzzle Phoenix Rangers uh and this was the current art style at that time but we thought that okay maybe uh because the characters were such a crucial part of the the game maybe there could be other other styles that could improve the marketability but thus also the engagement of the of the game itself

so we tried out like different concepts uh typically I do a boting thing here but I'm out of time so I'm just going to jump right into the results so out of these uh uh like um art styles uh we found out that this particular one was actually performing way way better than the the the original one so in terms of how much better for every 100,000 installs that the original art style got this new art style we got 100 like 38,000 installs more the problem was we ran this experiment at this time it was too late for us to then redo all the art on the on the actual uh game itself uh and yeah it was we already like uh ran ran through all the all the budget that we had for this particular game another issue is is that we want to collaborate uh but we have two separate separate units without connection and what I mean with with this is that specifically as you get into bigger teams so uh the game team is obviously

there uh the core game team uh development team is there from the very very beginning uh but the central unit kind of comes in with these different moments of time uh throughout like the process but it's not there continuously and it brings some issues uh as sort of the game team has this very internal Vision uh that's Guided by the external data points that that that U are there and the central unit is is typically then the one that then makes sure that the internal vision is communic get it with the external audience there's an audience match and and then delivers these external data points but as you can see it's it's scattered and only after or during soft launch it typically gets so that these two teams collaborate continuously as as one single team and lastly the smallest decisions have the biggest impacts um another another example we were working with this this game studio called funcraft uh and um they they wanted to build a word crosswords puzzle game uh and it was a new company so small team uh a lot of big question marks so they wanted to to

find out what how how to make it work so we did some market research and we found out that actually there were no crossword puzzle games that had like a very very strong meta in it so we thought okay this is this is a good good uh thing to test and good thing to build as well so we did more tests and we found out that in United States which was the target market people prefer dogs way more than cats so we we thought about like uh we build up this nice looking uh dog character that would be the guiding uh voice for this particular game so we had this these two different versions of this game uh we had other other uh things that we tested along as well but but between these two um with our internal data points that we our internal decision making and and external data points that we use to to gather up the data and figure out that okay this dock would work work better in paper we would have gone with that decision on the spot but then we actually uh if we if we take a look at this so so this decision has quite a big

impact so if you would go with this dog one so then the G focus on that game would be more with sound effects and all these other things that that that meta would actually require you to have uh it would mean that then that the small Studio would now have to hire a 2d visual artist for this game and maybe a narrative designer to make the character really flourish and then about how much this this decision would cost this would actually be something we found out find out during soft launch more or less whereas with uh this one the focus would be just on the cross wordss themselves and and maybe adding some social features in it which would mean that then the the team composition would be like different you would hire an additional level designer level designer maybe social features developer and so on and again you would find find out like what the what's the impact uh during soft launch but instead of waiting till soft launch or just going with our decision here we actually run this test before this game was even out so we built all the assets we we buil all the advert ements and and and run a

test to to see what actually what's the impact and we found out that this one worked twice as good as the dog one uh and and that's actually now the game that's out out there and there we can call it a successful Game launch as well because they're spending quite a bit money on marketing and making money as well these days with that game so to sum it up uh we say we do it for the developers but we're really developing it for ourselves uh we want to collaborate internally but have two separate units without connection and the small decisions that we make have have the biggest impacts so how do we how do we fix it honestly there's no secret worth hiding from your audience 700,000 mobile games almost all ideas have to some extent already being out there as well so what we truly can say just show it immediately test and learn um so what you want to do is always deze the market uh actually test it with the audience immediately so you don't have to wait

until you go to the end of soft launch you can do this this can be your your like original milestone in the initial idea test it with the real audience ask questions run surveys use tools like uh to to actually test with the with the players immediately to to find out what works and what doesn't and uh yeah and and when you do that you actually start to build a community of very engaged players already um and to to yeah and and on top of that instead of thinking it uh as a flight path with sort of accelerating touch points with the audience you should think of it as a loop so figure out what you want to test you test it and you learn you validate we had an idea that okay pule plus plus dogs would would equal money found out that it's not actually true so we went the other way around so you continuously do this and think of it as a loop rather than like a linear linear flight path and to give you give you some examples on how this could actually be

applied in in practice so so it makes more sense um I've been in love with balers gate uh since it came out I lost my sleep to it so we could now think that we would we would make a new game where we would take some of the the very well-made mobile mechanics of of Diablo Immortal uh but but take like sort of the rich story environments uh from from balers gate where all the decisions matter and huge trees and build it in Unreal Engine 5 but now instead of us building it traditionally with with like uh shooting in the dark so to say with this flight path we would actually do it um with with uh this this Lube mentality instead and what that means um it's not free it requires effort but it's 100% worth it so in the beginning what we would do uh it would be as simple as us uh building up an ad of this this game that we have in mind in this case uh as you can see it looks

like this this sort of well gets the point point across to to extent that it should this is the only AI thing that I will say now because I I bet a lot of people will will do that regardless but this is a perfect example where you can use AI as well during these early tests uh all these assets that I'm going to show you are are 100% built by by uh AI so uh yeah you can it it helps you cut the cost a bit during this test phase as well but yeah you show the advertisement on on Facebook or Instagram or both and then you drive users to to a survey where you ask more about you communicate this game isn't out yet you ask more about like uh what would be the things that you would want to see developed here and so on and after they they complete the survey you could direct them into into Discord already to build up your your engaged Community as simple as this um what you would get with this is actually already insights of of the audience you could see if there's any Market traction for your idea it

wouldn't cost you too much either uh 1,250 bucks would be more than enough for you to get the early insights here needed um and in terms of actual content you would need copy uh a few visuals and so on but quite standard to be fair all right then we figured out like that we actually have Market traction so the next step we could extend this a bit would be that we would run the test again on on meta and Instagram Direct the users to tools like ours where you can create like a lookalike store page so it feels in act just like the actual App Store page but it's actually an web page that you that you can build as you want to after which you direct the users to a survey that would in this case be hosted with us as well and again then go to to Discord but now we would want to figure out what would be the best art style so you would build up one with realistic art art style one with pixel style one with cartoony art style one with this sort of 2.5d style and then to to spice things up you would have one which is text based uh um content as well again you would you would run this

test 5,000 bucks would be more or less something that you would want to invest in in a test like this to have have enough enough users um to to validate the insights and with this you would know which which art style would actually have the best best impact we could continue instead of art style we could actually try out with the with the different themes so you could have fantasy you could have sci-fi you could have modern and again this this exact same thing you would have the ad then go to the app store or the look like app store Page survey and then Discord uh same thing you would still invest more or less 5,000 bucks to have enough users to to validate these insights um but with this um like actually instead of traditionally what could have happened is that we had we would have had our idea and we would have just gone with it and and built something that based on our external data points that we've gathered through through different platforms and and so on we could have made a very well educated decision and and and build up a

game like this which yeah which which would end up it being like that whereas with this this Lo mentality as we've tested and validated these insights along the way um we would have first find out that okay this art style is the best okay this theme is the best which would then end up us building completely different game but something that's actually in line with what the audience would would want and expect and lastly uh I like to say as well that you don't necessarily always have to pick the one that's uh sort of um from Market insights point of view the most most um effective one uh but what it means when you run these tests is that if your team is very passionate about like uh Absol they absolutely hate sci-fi and don't want to build a Sci-Fi game but you would see that okay this sci-fi team would cut the cost by by 20% so when you actually run these tests so then you will know that the decision that you're making has an impact of of of you having to spend 20 20% more on Market marketing meaning then that you have to take that into

account when you design the monetization and these other things uh to your game U so that so that then you can add up to the to bigger cost on on UA yeah that was it thank you uh we have two minutes left for Q&A uh try to be as fast as possible anyone wa one or two yeah uh first of all Jesse thank you so much for the amazing presentation my name is k I work at Google and just to W sure Jess see uh what are the main languages that are available in the gig laab platform uh everything in Gig laab as well and and that's again good point like uh when you when you design like this test and and you run these test so

having your target audience in mind is extremely crucial and then going all in so if you if you if you want to do it in States English is good but uh but really really going all in with with with um translations and the and the local languages is is extremely important I think there was one question there hello okay thank you so much Jesse for this presentation um I head in the into this meeting like 10 minutes when you were already talking so maybe I missed it uh when deciding developing a game what do you have in mind first the game you want to build or the target audience you want to make it for what comes first it's that's a good question and I uh we work with a lot of different game

developers so I would say that that also varies a lot between the developers so some actually go for the market first and then try to figure out what to build for that market but then sometimes you can also with and that's exactly what you this this methodology allows you to do is to also come up with your idea and see if if there is an audience and what what that audience actually looks like what are the main motivations what are the the well demographics don't matter that much but like the motivation and these these uh things behind the behind the audience so both both work