Build or Buy: Gaming Backends

foreign [Music] gaming podcast where it is our mission to explore the business and future of video games we bring together the industry's brightest Builders investors and thinkers to keep a full sun current events dissect emerging Trends and games share Lessons Learned and have a great time this podcast is also part of novik's growing ecosystem which ranges from free and premium research to Consulting and advisory services for more information visit www.navic.com now let's jump into the episode greeting citizens welcome to the novik gaming podcast I'm your host Alex Takei and this is the interview segment today we're going to chat about one of the most critical parts of building online games gaming back-ends it's one of the most important choices a game developer can make switching costs and get very high and from The Venture side is the classic infrastructure play that mimics

the SAS based investing of the tech world the lingo if anybody is wondering is stated as bass backends as a service um and gaming back ends can make all the difference between an exceptionally smooth scaling and the rickety scrape together growth if you're not prepared joining me today is Eden Chen the founder of pragma a back-end gaming engine geared specifically towards multiplayer games they are backed by investors like braylock upfront Ventures and inside Venture Partners welcome to the podcast Eden thanks for having me um and I know it's been a while since I last saw you I caught you at the friends of met you and caught you at the friends of bitcraft event in Santa Monica that's right summer yeah so rooftop uh barbecues with Moritz I know and then he moved so now there's no more Santa Monica um Malibu now yeah just leveled up in life on further west yeah um so before we uh you know continue on to talk about the you know the meat of

what we're going to discuss today I'd love to hear a little bit about yourself and you know intro yourself to the audience and tell us about your background in games and and how you came to found pragma yeah so always been a gamer since a very young age uh for a lot of people that means that I was socially awkward when I was uh young and so a lot of the ways that I made friends was tracks yeah playing uh playing games uh my parents actually didn't let me get like the consoles uh because I have you know Asian parents that are like video games are the death of my children so uh I always played computer games and you know hid that very well until uh I started playing competitively uh and that was Starcraft and into Warcraft 3 and so uh became joined a pro team and you know just like total nightmare nightmare scenario for my parents um and then I I uh from there went from there to just kind of like spinning kind of uh time apart from games that did

hedge fund trading for a little while uh and then started a company that was working really closely with snap on uh VR AR uh work and then um started pragma about uh three years ago uh with Chris awesome yeah well here's the strategy I you have to have a one white parent I had a white mom and an Asian Dad and I got all the gaming consoles so it worked together yeah that's it keep that in mind keep that in mind nice yeah two Asian parents and then uh you will not get one console oh to this day like struggle I I play uh I try to keep some separation between the work and and uh and gaming and so I play on Console nowadays but I'm like just eternally gonna be bad at like using the controller to some extent I play with the same pro team I played with in high school but we just played like console FPS now um and it's just like I just like don't

have the muscle because I didn't grow up playing no it's it's super true and even if you do have the muscle it expires anyway so um it's uh you got a very narrow window to cash in on your uh animal too yeah so older for a game nice okay so it sounds so you basically moved on to found pragma and how long have you been having how long have you been doing that and you know what made you interested in building a back end versus a gaming project itself yeah so we started in uh like end of 2019 early 2020 uh the so Chris my co-founder he had started a Ed Tech game right before this uh they ended up getting acquired by Phoenix Labs um uh you know who's been in the news recently uh but they so he stayed over there for a little bit helped scale dauntless and then we we had talked about working on a game uh so it was definitely a serious consideration uh both of us are just you know been friends for a long time super passionate about games I had seen uh several

Studios that I've been talking to at the time that had asked us both of us particularly about backends that's both of us are back-end Engineers by trade and once they started asking us about that um one of them had just said hey if you build us a back end we'll just kind of let you guys take the tech and make a company out of it and so there was some interest around that in terms of just there was demand coming in from us around lots of people that asked us you know can you build us a back end and so I started doing a lot of like digging on the history of backends and um there were lots of them out there there was you know playfab and game Sparks and there's probably a dozen other ones that started sort of in the 2010 to 2015 period so a lot of it me was as a student tried to learn like okay why did why are people still looking for back ends if these companies have already been started like isn't this problem already kind of solved and uh what I kind of learned from a lot of these Studios is that like if you're making a sort of cross-platform I call it like the

aspirations of a AAA game these Services tend to be extremely uh unique uh particularly like player data and matchmaking like if you just think about um you know it's just something like a battle pass and if you even look at like one battle pass like the fortnite battle pass the evolution of that battle pass is that it's been tuned so many times that when you're looking at like a play Fab or game Sparks and you're an experienced person that's built these systems before it becomes very difficult to rely on a black box because those systems are inherently part of the game design uh and not not being able to tune those is highly restrictive for a game that wants to sort of be this aspirational AAA game so once we started seeing that Trend play out with uh a lot of Studios coming to us and saying hey can you look at this or help us solve this problem we were like oh maybe there's like the different way to think about building these systems that's not the way that it had been done in the past so that that was sort of the the

beginning of us like exploring it and then um you know we found like sort of three design Partners right outside of the gate and um then we sort of just went down the path of building this this this company so we stuck with those three same Studios for our first couple years and then last year we started kind of broadening out more awesome that's such a cool story I think like even it sounds like you like started with the game but then you discovered that there was this tooling problem you know you're reminding me of the slack story of course um you know you're building a tool to help you build games and then that tool is in itself becomes a business and you know just the other day we had an nfx investor um come and talk at Stanford and we talked about sort of like the four aspects of competitive Advantage um you know Network effects brand scale and then the embedding problem and betting is like what Oracle does it like just like hitches itself into an organization and then like it never lets go right and so I think back ends are super fascinating because you're are saying that you know like hey there's all these problems that maybe game

Sparks or playfab didn't solve but these games are kind of stuck on this technology so I'd love to um like you know later on talk about like you know what makes pragma different from maybe some of the back ends of of old but before we kind of get down there let's just maybe you know for the for the noobs in the room you know let's talk about like aiming back ends and what they are and the services they typically provide you know how would you describe a gaming back end to like somebody who has no idea what that is yeah I guess like historically speaking in the games industry there were like a lot of things that were stored on clients that like be like the hardware device that you're playing on um or just like you know the the everything was run sort of on the client itself and as time uh passed on and there was more competitive games and there was things that needed to be validated in some way because there was dollars attached to it uh a lot of this stuff needed to start moving to servers it's also just like easier to update things from a server than it is like having a client update so like the easiest way to think about back end is

like something that's interacting with servers and databases the the suite of services that we typically think about are things like accounts and login it's social features um it's player data player data is uh and game data are are pretty broad terms but that can include things like progression um your monetization your battle pass um and then uh you know it Go class crafting the list kind of goes on and on it's a very broad based group of things but it's effectively the metagame um inside of a game uh and then the last area is sort of matchmaking uh which is like how do you get players in the same room together so it's a very Broad um sort of Suite of services but uh namely it's not things that live on the client like that's the actual networking itself like what unreal and unity are doing the physics the rendering uh those things are not like part of effectively the back end got it okay so if I'm if I'm a player right whatever like basically all Commerce login accounts

maybe the matchmaking maybe like the social systems um those are things that are included in a back end right yeah and I say I think the confusing thing is that um the same language can be used for things that used to live on a client um which is why I tried to make that distinguish distinguish that earlier in other words like there are purchases that in the past um may have been an in-game purchase that may have been on a client and that's like usually when things you know you might be playing um a single player hyper casual game where the cheating yourself doesn't make any sense like the points that you get you're just like sort of cheating yourself out of the game it's when games get competitive and when there's real money exchanged where it's really important to have some security and that security typically happens in sort of the validation of the have to set a server so like that's where like um in the past you might have lots of things that looked the same whether it be even like client-based matchmaking

for example back in the day with like blizzard um all this lobby-based scrolling where you would go and you would you know go and your user map setting games and you would scroll down and find like the 3v3 game and just trying to join it that's a client there's no server that's sort of acting as the sort of middleman between the the two players there's that's just like some maybe peer-to-peer hosting or that's a client-based sort of hosted Matchmaker and so that's where the terms can get sort of confusing where there's not necessarily a back end that's associated sometimes with a similar feature set or something that looks similar it's just when a game becomes like cross-platform for example you can't really have things live on a client anymore because if you log on Xbox and then you log on your computer some the server needs to actually sync those two things together what are you gonna do yeah yeah that's right got it got it got it and so then so you're providing the example this this huge Suite of services it might be security Commerce matchmaking social systems you know when have you seen in the past this is a developer that might be wanting to

make a cross-platform game choose and commit to their back end choice you know like where in the development cycle would I decide oh I'm going to go with playfab where I'm going to go with Photon yeah I mean it's typically fairly early in the development cycle um and uh the the reason why is because once you start Play testing or if you want to run real by real play test I mean like you're actually going through a real matchmaking Loop and that's not like sometimes people test locally um like in the same office on a VPN and you can kind of just play that way but once you start doing any time to outside play tests particularly even with like remote you know work you can't really just do the same things that you used to do when you started getting in those real motions that's typically when people start looking to sign up with the back end so it's like when you want to do like the first two things that we typically people see people do is like and maybe like an account integration with like a steam or epic um and then like a a sort of game Loop or a matchmaking sort of Loop and that's

just so you can start actually play testing um so like the earliest phase of game development might be like you know you're doing the concept art you're doing like very very early kind of like game design and then the next step might be like even at the most like sort of grayscale kind of art you might start wanting to start testing some of these mechanics with like either a small group of players even if it's like intern Journal testing or testing with some friends so that's like the first step some people I'd say might do some of that work early themselves like they might create you know go into AWS and start like you know building some very light platform work so that they can start facilitating some of this themselves and then they might push the platform decision or the back end decision until you know like six months before launch or whatnot and that's usually like a pretty bad idea and I could I could talk about why but uh but a lot of Studios will do that too as well they'll sort of like put together something janky just so they can like test it and then like kind of punt the decision until later

got it and I think the idea of punting it until later is sometimes people don't know exactly what they need out of in terms of the services that their back end might provide right so maybe let's just say like I don't know whether or not I'm going to have PVP yet I haven't decided that so I don't know if I need matchmaking so I'm just going to hold off onto this decision yeah yeah and I mean a lot of times like when you're trying to find the fun you don't you don't want to make that many decisions and you don't uh you know it's like it all makes sense it's rational it's just kind of like one of those things where like um typically where we see some students get in troubles when they start baking a lot of their player data into the client um because it's it's easier and it's what people have done for a long time and then when they actually choose a back end they have to sort of redo a lot of that work in terms of moving that all over to like a completely different architecture so that's where and then typically their their launch deadlines don't really adjust to account for the fact that they have to do a lot of that rework so they may

be like oh we have six months until we're launching and um you know all this stuff has been baked into the client and they and they can't ship that way they have to move all that to a server um and then they make that decision then but then it's like crazy crunch to try to move all that and change all that and that's where you see some some kind of issues when it's not sort of like built into the actual like development plan got it fascinating yeah so it's so it sounds like the best course of action but you know best practice would be to sort of commit to your to your back end earlier in your Dev cycle maybe around the time that you start doing play testing um that's right yeah and so in that's in that regard it sounds like play testing is like the critical step right you know it's not Commerce it's less about Telemetry but it's more about play testing and well I guess my point there was that that's typically when the developer starts saying I need this um some of the more complex areas for like live at least live service games

are around content um and that I mean my belief uh is that you know that player data you really want to to really think about very early on uh because there's things like data versioning that get very complicated when you like a lot of people launch a game and they haven't even thought about how they are gonna version you know X1 X2 X3 and therefore they have all these like weird database conflicts and things when they go live and the game starts crashing because it's like you'll see these errors like you know you know wrong database version or something like that the game will crash and then you know things like that start to happen because um there's not a lot of like pre-planning so that's around more of like the player data side but from a game developer standpoint the way that they think about it is sort of like I have a need which is I need to start Play testing I don't want to right basically roll all this technology myself so I'll I'll use a third party and then typically if we would meet with a studio at that point we might also start making suggestions on like how are you there how are you guys thinking

about progression or how you guys thinking about your quests or your crafting or your inventory or skins or it depends on the game but we start you know asking those questions to start getting them to think about doing that earlier sure sure um and so I'd love to dive actually a little bit deeper into like the three big or some of the three big services that you mentioned in a back end right that being matchmaking the Commerce side and then maybe the player data the Telemetry setup um you know on the matchmaking side you know what makes a good like matchmaking program um I have personally worked on a few multiplayer competitive games OverWatch and Hearthstone being two of them and on the League of Legends side the matchmaking is just bad because I'm not my my teammates are bad so I'm totally totally the thing is everybody complains about matchmaking like that's that's the like you know joke obviously or it's like you know it kind of doesn't really matter how good your matchmaker is everyone

complains right about the match shaking uh I I mean I think fundamentally speaking I think matchmaking and game design are kind of inextricably linked in that like if you look at like a you know I guess what we would I I would consider like a good Matchmaker like a league even though you were making the joke um the uh you know there's a lot of decisions that were made uh on games like that um even like with DOTA around like you know Banning players where that the whole matchmaking process is actually part of the fun of playing the game um in in League that you can select you know multiple you know uh roles for example if you're if you want to play mid or and you want to play like a backup as a support or something like that there that's uh added to the hold of whole complexity of the Matchmaker which means like part of the game design is allowing people to make bands and make alternate selections for their um you know for their role and even the

team compositions you know part of like the game itself like in some you know if you're like a really good team and you could lose the game before you even start playing if you have a bad decomposition so so much of that is like really important to the Matchmaker so I guess I I'd say a few things to answer the beginning part of your question what's important Matchmaker one I think is just like scale is is one thing that a lot of people think about because typically when games don't scale they fail in the login or they fail in the Matchmaker um and so you know if if you can't get into a game um that's a huge problem so and matchmakers are kind of like Infamous for not being able to scale um so I'd say like first thing is like sort of like are you stress testing your Matchmaker are you love testing um you know do you have a Matchmaker that that can't actually scale and you know it's so hard with games because it's very difficult to predict you know especially like on a launch like you know how many players are going to get and whatnot but but that that's one aspect I think the second one is much more of a broad statement and it's uh

one that's tied into why we started pragma and that's a Matchmaker needs to be flexible enough for you to tune the Matchmaker to and make it evolve as you learn new things about your players as you learn new things about the game and as the game evolves over time particularly in a live service game and that's I think our Insight was like particularly with matchmaking um folks that have built live service games for a long time they really want to own the game Loop when you when you sort of lose access to to making those changes you really lose a part of the fundamental part of the game and so a really good matchmaker is one that's flexible enough to allow you to to make the changes that you need to make whether that be like like I just saw like Apex maybe it was two or three weeks ago they're getting [ __ ] on yeah yeah I mean and again like that's gonna happen like any time you make changes everyone's gonna complain but I mean the point is that even after you might have made a mistake you have to be able to

adjust that based off of the you know maybe some of that feedback is is inevitable because again everyone's going to complain but you have to be able to evolve that very very quickly and make changes so it's it's one of those like never ending there's no perfect matchmaker I'd say because it's something that needs to constantly evolve but the systems that you have need to allow you to even make those changes and that's like fundamental to having a good Matchmaker sure sure and then for the some of the biggest online games that we know of right and Apex and OverWatch Hearthstone right um are these basically using back ends that they built in-house in terms of their matchmaking or okay got it yeah almost all of the biggest games are um in-house games so Chris worked on the League of Legends my co-founder work on the legal and flashmaker I think that was you know at any given time 20 30 40 people and they were working on it for you know six seven eight yeah I mean it's forever basically right it's that's not like

that team gets any smaller it's just it continues going and they continue making changes and tweaks but um you know that's a team that I think started actually as the player Behavior team at Riot um because they're and that was like how do we address toxicity um and then what they realize is that the only way that you really can't address toxicity is actually like incorporating those choices inside of the game itself it's inside of how you match into a game um in in terms of how you uh get positions for and I think like part of the learning there was like you'd be pissed off at your teammates in the beginning if you got a position that they wanted or something like that it would you know it's like a lot of that has to do with like how you enter a game um and so uh that's how it started and then but but yeah these teams end up being very very big and they're long lived in many cases got it so before I ask you on the existential question of Builder by um because it sounds like you know there's a major in-house requirements and and you know Staffing and overhead

for for building um a matchmaking system just want to touch back on the on Commerce and and Telemetry you know I think something that I've always been super curious about is to a do like all gaming back ends come with modestation tool assistance um like stuff like virtual Goods or analytics or currency management um to manage maybe even like uh currencies across different regions and then how does the actual point of sale operator someone like a steam or an Xbox or Sony like actually like play a role in this dynamic yeah so um there's yeah this is such like a complex topic um the the whole monetization as a whole I'd say they it's it's very nuanced into which back-ends do what because everyone can sort of again like I mentioned before anyone could say there's a Matchmaker um and but it could be a clients based Matchmaker anyone could say there's player data but that could also be cloud-based but so in terms of I think

there's a whole swath of different sort of monetization type products out there um some are um you know much more complex and some are very simple the most simple like database it's just like a key Value Store um and that's just like you can you put stuff in a database and take it out and that's like your your most most most basic uh sort of uh back end it's just database and then there's uh ones that are going to do a lot uh more you know a lot a lot of different things so for example if today uh you purchase something on steam or epic or Xbox or Playstation those people act as the actual payment facilitator so they they take the you know the dollars in and then the back end itself is going to handle the entitlement of the item so steam will communicate with the backend and say hey this person has purchased this item and then that entitlement uh is a cross-platform entitlement so that

might have to crap in across like the PC their Xbox account Etc so that's that's sort of like the the base layer ID that deals with um the the actual fulfillment itself so at the most base level on the monetization side it's like players purchase things um on PlayStation and then the uh the actual back-end entitles it where it gets more complicated is that's just like the actual real money exchange and typically um in most games what what people do is the the real money exchange just turns into some in-game currency um and then the in-game currency is typically the the complex part of what the back end needs to deal with um because once you have the in-game currency then that can be used to buy more items inside of the actual game which is what the back end's dealing with and then dealing with how to to do the crafting and the bundles and the discounting and all that and so that that's part of the so like so like if you think about like an apex you might purchase

something that gives you some kind of gold Apex currency um but once that currency lives inside of the game then that's all the back ends like responsibility to handle all the rest of that post like platform purchase I guess so all the all the actual like platforms they're just handling like sort of the real money exchange but the back ends have to have all like the complexity that happens afterwards got it got it oh that's fascinating yeah so it's like I'm in my mind and picturing someone's just like Sony and Xbox and steam are basically just like couriers that like run to your back end with a letter that says you know bill paid twenty dollars give him his 200 gold and then it's Off to the Races to the on the back end to basically say okay now bill has 200 gold where is he syncing that in our system and you know how do we surface him the kind of offers that he wants to have so even that kind of like shop personalization that would be handled on the back end side that's right yeah it offers bundles you know there's Dynamic bundling which is like if you've purchased one item in the bundle how do

you adjust the pricing to to take into account the fact that that users you know already has one item in the bundle because they don't want to buy double purchase you know uh something like that there's all sorts of complexity uh uh there as well and and then you have to deal with things like if the player asks for a refund on the platform how do you deal with that what if they spent their money already um there's just all sorts of things and a lot of that is just like up to the developer also to decide like how they want to treat those things because you know it also feels bad let's say you ask for a refund but you've already fought an item like you claw back the item the purchase purchase like there's all sorts of decisions that need to happen there super fun um and so then I guess this is maybe this is a little bit tangential but you know we just talked about Xbox and Sony and steam right those are some of the big you know platform and Hardware providers as well as you know but there are other you know ways to link into Tech right like you know msoth owns playfab um you know I think Google owns Firebase right what's up

um oh sorry I pick up MSA from the Activision Blizzard lingo um yeah he's just so he always abbreviated the tickers so like I just started we just started saying that um yeah but you know so what are the pros and cons of maybe picking a back end that's affiliated with like a larger scale server like you know within Xbox right like is it is there some Advantage right if you know I just told that analogy of The Courier running with the letter that says I you know bill paid 200 give him his 200 gold is that faster if it's on playfab and this is an Xbox game and he made the store on Xbox you know on the Xbox store yeah I'd say I mean there's not really in my mind like a pro oricon um I think it really just has to do with the feature set that uh a studio really wants and that um yeah I I play five and Xbox they have very distinct teams like the folks that are working on Xbox or

not uh the same folks that are working on playfab um and there's no real advantage and then I think it would do a a disservice to Xbox developers if they gave special preference to you know developers that were only using playfab versus uh people that were building in-house or using pragma or whatever else so I I don't think I think it's intentionally sort of separate there's no real Pro or con to any I guess kind of first party systems the ones I can think about are Steamworks is another one that's the it's sort of like steams uh in-house uh you know set of services that deal with some of this uh playfabs the sort of Microsoft version Amazon used to have game Sparks but they deprecated that um but then those are the two ones I could think of but there's no real Advantage I'd say outside of just like the product itself I'd say got it got it okay and even that's true even in the you know Telemetry setup as well you know I remember at the Indie Studio that I worked at we basically had a playfab plus snowflake setup right so this is

again me potentially butchering this process but I've seemingly collected the data they said that's the snowflake the warehouse and then we use stuff like Tableau to surface that data on the front end to like an analyst or something like that yeah I mean that's exactly right and that that's exactly how we treat it yeah we would treat it the same as well and that like you don't really want to get in the business of trying to compete with snowflake or datadog or some of these services like what you want to do is make sure that uh as a back end you're collecting data and you're packaging it well and then you're passing it to the provider in a simple way um and that provider could be snowflake that could be datadog there's there's like multiple levels of analytics and Telemetry there's like sort of the live metrics data which is about like how many players are currently playing my game you know how many crashes are there it's the things that you want to know about the live health of your your game and there's lots of standards for that like open Telemetry and different things

like that and then there's the sort of um metrics that you might want to be able to query at any point and that's like your your data your data warehouses like snowflake where you might want to say like um you know what's the balance of this gun and like you know are people winning with this gun or that gun and you want to that that's a lot more about like being able to query and ask questions about what's happening um is there a character that's winning way more than another character and those kinds of questions and so does the back end provider and the developer work together in tandem to establish some of the criteria that I may want to eventually query or is there a standard off-the-shelf kind of boilerplate metrics right like yep we're going to give you Dow yeah on the metric side there are definitely sort of Standards like concurrent users and um your crashes and and you're so your health Telemetry is definitely something that's more standardized it's more of the the data warehousing side that's not standardized and I don't think that's really I mean I

think that is never going to change and that like game teams always need to Define what's important to them and what they want to query and what they want to store um and uh even for us a lot of it is about helping teams ask the right questions and helping them make sure that they have a good data pipeline not necessarily saying like hey pragma off the shelf is going to give you because a lot of people are like oh do you have analytics and the question is like I guess like yes but it's like what does that mean what does that actually mean what's that mean for you and your team yeah yeah it's like that you still need a team that's going to go out and write a lot of you know do do the work in terms of actually capturing the information that's important to you because the important the information that's important to you versus someone else is fairly different but the things that can be helpful are like um actually getting that getting those uh getting that data over to the data providers is something that isn't that trivial sometimes and that sometimes collecting that data can take time and so that's what we what we really do is

really just like collect and package data so that the provider can actually read from that in a in a fast and efficient way got it yeah I mean I think that asking the right questions part is also so critical I had the wonderful honor of discovering that some of the Dows of Blizzard's games are defined differently so like in one game a dow is someone who hits play in another game a Dao is somebody who sits in the menu for 30 seconds so asking those right questions I think is so important because it has super big implications especially if you you know if you make it past launch and your game becomes a hit like then you're doing all these live Ops situations and you're trying to your your PM teams are asking you all these questions and you're like how about like the data like and so totally yeah it's just fascinating um and so I guess that brings me to the question right like it sounds like you know a back end should really be a partner to a game development studio and I think in that regard you know I would

love for you to sell me on like you know why should why should I buy this service you know why should I but get a playfab or get a photon or get a pragma versus building it all for myself in-house if I need so many custom tools and I need everything to be so specific to me why would I not just build it myself yeah so um I guess I'll quickly sort of delineate a group of services from another group of services so there are services that tend to be more generic it doesn't mean that they're less complex but it just means that um they can be used by a broad range of Studios so one example of that would be like the account system so like accounts are basically like the the guys who basically link up with steam and epic and Xbox and PlayStation and iOS and Play Store and everybody and then sort of like create that sort of ID that player ID that can basically facilitate logins across all those different platforms so no matter what game you're working on um you're gonna have like a pretty generic account system that works across

all those different uh providers and the work that you're doing is also just like just like creating facilities to interact with each of those providers it can get extremely complex when you start getting into cross progression across social and how those work across different platforms Etc so I'm not saying that those are easy services but they're more generic um the player Data Systems and the matchmaking are hyper uh uh they're both complex and they tend to be unique per game and so I think those are the ones that are probably addressing your what what you mentioned in terms of just like hey if these things are going to be pretty different per the game that I'm writing why not just sort of like build it on my own um versus like using a third party so I think it kind of depends so like I think like playfab for example if you're if you're just sort of like building an account system play five is going to be sort of it's going to be amazing for that they've done it for a long time a lot of people use uh playfab's account

system works great for them um and and that's that but when you start getting it into the more I'd say like extends like the need to extend out services like matchmaking and player data that's where um these sort of Black Box SAS providers tend to to hit their limitations um and uh that's kind of how we tailor-made our uh our services to address more of the the guys that need to tune um the like the player data and matchmaking services so uh again like I think part of the rash so you have like sort of three choices you have sort of the playfabs of the world are going to be more black box oriented off the shelf very easy to get started easy to use you have sort of like build it myself which is like I gotta roll everything from the beginning and build it myself and then you have someone like us who's like we're gonna you know try to solve those problems but also give you the ability to sort of like extend out and customize the actual things because we don't think those services are really able to be generic um or not even generic they're not able

to they're needing real tuning uh to make them uh work the way that you they want to work so one is like the actual actual feature itself in terms of why you would want to use a third party versus build it yourself that's like you know when you're looking at like a player data system there's all these things that you don't want to deal with like data versioning database migrations in a live environment that's like if you rebalance your economy you rebalance an ax because it's um it's it's imbalanced it can ruin the game for a period of time you don't want to do a client patch you want to actually have a server ability to sort of update that ax live but then if you do that and your systems don't have the ability to do database vibrations there's no data versioning you start to deal with all these conflicts your game starts to crash Etc so these are things that are that are generic problems within highly customized areas so I think I people can go down the path and think like the the feature itself might be relatively simple to build but it's not just to feature itself but all the things around it that you have to worry about

um like the data versioning stuff that I mentioned uh the other part of it is also like even after you've built all your features you have to actually operate those features live globally um and that's uh 24 7 lights on all the time someone needs to be monitoring those servers um you know for a big game team that has three 400 people that are dealing with live service live operations and monitoring the graphs and doing all that kind of stuff that's feasible but if you're getting started and you're you know having to put a 10 20 person team on building out your player Data Systems your matchmaking your accounts all these Services that's one thing and then you have to staff the whole live operations team to actually like manage it long term so it's just a question of just like time risk uh you know cost to Hot and Hot to hire the team and so those are the questions that I think game teams need to ask um you know in some cases it does make sense for a team to to build that in-house I think the our bent is actually like we we really think games game teams really do need a lot of

platform expertise in-house even if they're using pragma that's always been our our push uh to game teams we are not trying to say we're going to solve all your problems and uh I think it's unrealistic uh to assume if you want to ship these complex cross-platform live service games that we could solve every problem for you um what's much more realistic is to say we can maybe solve 80 90 of your problems and then they'll last 10 20 your higher in-house and have some expertise there so that's the the sort of the trade-offs that that folks are thinking about got it got it so obviously feature complexity in-house hiring you know paying I don't have to pay my back end infra engineer I pay you guys a fee maybe potentially reduce time to Market since I'm not building and building and building all these back-end tools and get my game out faster um you know I don't have to do any kind of like live server management um so those are some of the some of the memories well I say the the long long

term dream for us is that most Studios they sort of get uh they just get to building the core Services the core Services being all the things that we've talked about you know can I log in can I you know get into a game those kinds of things but they never really get into like the really interesting platform metagame features which are just like much more asking like how do these social features actually like encourage people to play my game more um like whether that be Guild and Clan management whether that be like um how you interact with your friends um whether that be even just the Matchmaker being something that's a a good Matchmaker for your game um and that's where I think like most Studios even the ones that have large teams they'll hit this sort of cap where there's still to guacamole like platform problems all the time or they'll start working on their next game and they'll never really get to the point where they're working on these really interesting social features so for us it's kind of like yes like our hope is that of course like we can save you money save you time and like de-risk your launch and things like that like

that's just the core services but the long term is sort of like we want to actually start working on services that game teams never get to work on and that's just like how do you make these meta game systems like actually make your game like more social more fun like give you higher attention and those kinds of things are really what's interesting like we're still working on the core services so we're not there yet but that's that's sort of the dream got it and I think this is like a perfect segue to actually just you know shift over and talk you know about pragma's kind of value proposition you know you guys are a new entrant into the back end space you know I think your seed round debuted in like 2020 and I think you recently raised a b round of like 20 million or so in October um so in the light of that what we've discussed so far you know how is pragma different than the current incumbents like you mentioned that there's these three pillars you can take your playfab pretty inflexible good at doing the old stuff you could build it in-house you could choose us you know so you know how are you guys built to scale with the game studios that you guys on board or you know provide that kind of custom flexibility yeah like we've tried to

focus on a particular Persona as a good startup does in terms of just like trying to identify exactly like who is our who is our customer and like what does that mean and for us this sort of like double A to AAA sort of Lies cross-platform live service game like building that match based game for the most part is sort of like our sort of ICP initial customer profile that like we're we're focused on this type of persona and that requires a whole different set of services that uh then like a off-the-shelf play Fab that's really gearing much more towards a mass audience does like all of our stuff is like single tenants there's no Noisy Neighbor issue or like if one game takes off it'll just knocks down the whole service it's like every um every kind of I love that noise I've never heard that before that's amazing yeah so like everything is sort of like built for the one studio um and each one of those Studios is fairly unique in how they're set up they might have different Dev environments stage environments um there's some of them have like

designer shards where they can use to really like uh for their designers to kind of play around with and the engineering shards where they're dealing with a very a variety of things we do like a lot more white glove kind of like service and some like every single one of our customers we like have a slack integration with where we're basically like hand-holding and working with them you know very closely so it's kind of like a different kind of offering I'd say we we're trying to look a lot closer to like a Central Tech team that you would have at a big studio um versus like um you know play Fabs trying to be like this SAS provider which is like you know more off the shelf like easy to set up easy to use it's a different Persona um type of uh a studio so what we've looked at uh very closely is that like we're trying to look more similar the the analogy breaks down in many ways but it's good in some ways um we we try to look more similar to like the unreal of back ends in that like Unity was sort of this more

off-the-shelf engine and uh you know you could use the artist tools that it provided to create your game whereas unreal the expectation with unreal is that people were getting inside of the engine making like pretty heavy modifications to it our kind of thesis on the sort of cross-platform live service side is that in order to do like some of these really complex um you know platform problems particularly ones I mentioned uh people need to make those modifications and so our our core value proposition is that like for those big teams that really want to make those really you know to build those really big systems like Prague was really like the only place you can do that without just like rolling your own which is that option and and it's option that makes sense for some people but for you know some studios that don't want to basically like start from scratch and get a lot of value out that's like kind of where we come in and again we're not trying to solve like the hey we're like working with like a hundred thousand Studios or whatever it's like we're really focused on like this narrow group

of like it's all these like cross-platform live service games and it's like that's an ambitious effort uh it's expensive it you know it's generally like a pretty big team even without us it's still like a pretty you know big effort so that that's like kind of who we're focused on our goal over time is like of course we want to like democratize these services for a broader range of people but like what we had seen was that like the playfabs of the world they would have like these these rocket ships that that sort of grew out of the the solution or sort of got to the point where it was no longer an appropriate solution and what we want to do is sort of like work on the the sort of end state so that once we start going down the path of sort of democratizing these Solutions people are then like oh like I can actually like continue evolving my game with these services I don't have to like rip it out and like start from scratch right right you like trying to grow up with the studio in a way um that's right yeah so we sort of started at the it's the most complicated sort of use cases in many cases and then we're gonna sort of go down and make it more generic got it and so do you guys have it sounds like you said your Target demographic is going to

be like a double a triple a studio with you know live Services cross-platform what games have you or what Studios have you currently partnered with and how do you actually like on board and increase that pipeline from the BD side right you know one of the things I ask myself when it comes to kind of growth is like you know for you guys you can only onboard a certain number of Studios that you want to Ally with and make sure that you do do right by them right and so totally you know how does pragma itself as an organization scale up like in head count with the number of Studios that it has yeah yeah that's a good question I think when we got started originally like I mentioned we we work with three Studios to start this wonderstorm uh mountain top and one more game when we when we first uh started going as like sort of our sort of design partners and that was very much a very different relationship than um the studios that we've brought on more recently so studios are you know Venture back Studios uh you know spinning out of AAA games they've you

know work with these Central Tech teams in the past they they know what they're doing when it comes to building these services but again like maybe like not wanting to start from scratch so that relationship was very much like sort of um build we'll build along with you kind of thing um you know once we had much more of a mature product after like a couple years of doing that we started onboarding um you know we like started working with like uh students like Mountaintop and um you know free range games which is a much bigger studio um and that that one they're working on the the Lord of The Lord of the Rings Mario game that's coming out in uh June 6 survival crafting game um and then there's a bunch of other bigger Studios that we are aren't allowed to name yet uh that we're working with as well so we started like moving up the like chain of like starting off with like sort of the Venture back Studios much more in the design partner relationship moving more to like uh like a real product and then moving up to like working with teams that are like you know the beginning was like 20 30 40 people to teams that are now like 100 plus people um the in terms of how we think about like

scaling our own uh business I mean the benefit to us is that like because we have an extensible uh platform uh Studios themselves can go and actually build on top of our platform using like plugins extensions these custom Services uh the difficulty of like taking on you know hundreds of Studios is that all of a sudden they have like all have different needs um and I bet and those are that becomes a major issue if the studio can't go in and actually write some of their own services so it's like yeah no you know we could work on these services for 20 years and not be done with them it's just there's an unlimited set of like things that we could be building and so like no Solutions ever gonna like meet a hundred percent of any Studios needs that's just like not the reality of the industry that we work in unfortunately I wish it was um so then the question is sort of just like what do you do when the solution doesn't meet the thing that you need and the question is you can wait you can ask for a road map or you

can build it yourself and what and then the question becomes like what does building yourself look like so building yourself mean like I have to just rip out everything I've done and start it from scratch does everything just doing yourself being like I could do it within the ecosystem that I've already Chosen and so those are like kind of the questions that we try to be like okay well how can we give Studios like a lot of value but then like not block them from doing the thing that they're inevitably going to be missing right right right yeah that was actually I'm glad you you brought that up because that was one of my questions I was like what if you guys have a feature that the game needs that's not available you know how do you let them make it so for themselves within your own platform right yeah it's challenging it's like you know in the perfect case like 20 studios are all asking us for the same thing and we're like okay that clearly needs to be the next thing we work on got it um you know it's usually not the case though it's like usually five Studios want you know more you know they want like a hook into our Matchmaker in some way you know another five Studios

want some social feature like friends that cross-platform friends and you know it's like we have to sort of prioritize that but at the end of the day our answer is like you know we're trying to do our best at addressing the most critical issues for the broadest number of Studios but if we don't you also can you know solve your own problems as well and I think that's comforting to the people that know like the people that have done this in the past know that there isn't going to be a one-size-fits-all solution that's just not um you know if you if someone believes that they're probably they probably just haven't done this before and they just don't see the complexity yeah yeah no absolutely I mean that's super true it's true with products like a product for everybody is a product for No One game for everybody is a game for no one yeah in the financial world of an Excel model to end all Excel models never works um uh so it's definitely a challenge I think it's really interesting to see kind of how you min max and you decide you know it sounds like you you know take a lot of feedback from the studios that you're partnered with but if I could just ask sort of like you know

since you guys are really new you know how did you convince your first Studio to come build on the pragma back end when they could go with the safer game for app or playfab or anything else like what was that story like for you guys yeah it's a good question I think like yeah early on like I said like the folks that uh could use playfab were not the appropriate people for us it was actually the people that wanted to build it themselves and saying that like we know that playfab's too restrictive for uh what we want to do and so our approach was hey like you're gonna go and try to hire a bunch of platform Engineers like in the beginning it was like let us be your that that team for you it'll be a lot cheaper for us to do it because we have BC money and you know we we can build a a business that works for multiple Studios obviously there are some trade-offs with that and that like maybe you don't get the exact thing that you need uh you know that that you need worked on today but it was like we had a limited number of Studios like I said we

only took three on the beginning um and they had similar game types so we tried to be pretty um choosy about uh you know the crossover of features between the three Studios that we picked um and and so I think it's just a value proposition for them and that like hey I can go and hire a 10 person team out or I can hire pragma for cheaper uh and these are very difficult people to hire it's just platform engineers in the space so that was how we got started and then once we had a more mature solution it was a different that was a different conversation because we could actually show people like oh now we actually have like some services that you can take a look at got it and then on the app like you know what what I say models have you guys chosen to go with you know usage or fixed or a combination of of both yeah it's it's a combination of both like we have basically like a fixed license model which is like an annual or it's break basically broke broken down to a monthly fee um and then we have like a variable rate that's attached to concurrent users

um and then the concurrent user uh is used because it doesn't deal with it deals with sort of the spikes that can happen like of a streamer you know logs on and you get like you know five million people that hit your game um you're not charged for the entire month so if you're if you did like a MAO obviously that would be like charging them for the entire month versus the CCU is just like snapshotting every single hour and so like it deals that burst and then it comes down and so um the diff the difference between us and like sort of like an unreal from a technology perspective is that we're a live Service Tool like unreal is like much more of like a authoring tool I'd say and that it's like an engine to create something not the problem but our costs are higher because we have to operate uh in many most cases the platform uh globally so there's a need for us to have some fixed-speed deal with sort of just like the day-to-day operations of supporting and um operating the platform and then the variable rate is more just about like the upside of like you know

participating in some growth of the games as well yeah exactly yeah yeah got it interesting yeah I think those um the different types of incentive models and sort of scaling and sort of ratcheting ratchet tiers of fees are very interesting I think um you know you can have one of those situations where you know all of a sudden now game growth is a tax right on the studio and so I think it's an interesting um Dynamic of having the back end player have a variable rate on users yeah I mean we've tried to make it like reasonable enough such that like it ratchets down over time and and all these things and you know like for us like if a studio is not successful like it really doesn't matter for taking you know some crazy fee it's just like we have to it has to be a reasonable amount and you know we're a startup we're like it really matter like the most important thing to us is successful launches it's like it that that's what like it doesn't we're not trying to like monetize the perfect and have the perfect pricing model like we would adjust all that just

to make sure that like we had the good relationship with the studio and make sure that it's like a win-win it has to be that way or else like we can't be successful at this stage sure sure and then who would you say like right now like there are I think there are other new entrants to the back end space like a beamable right or would you consider them to be your competitor and how do you gain that Competitive Edge over a new you know it sounds like you have a clear value proposition against like a playfab right but what about the new guys that are doing something similar maybe to you yeah everybody's sort of approaching it a different way they're so I'd say like we were the first ones that raised like serious Venture money I'd say and then uh there's been lots of new players that have launched that have uh subsequently raised on different sort of thesis or just different focuses like there's like beamable and loot Locker loot Locker has been much more of a Content system uh that uh you know

focuses on uh it seems like more of like a mobile audience uh beamable is focused primarily in our minds on like web3 and um and and sort of like you like Unity games they started as a as a mobile Unity uh type of tool whereas for us we've specifically focused on primarily PC console like live service like games that you know our team comes out of like like uh like a half our team comes out of like Riot and I have our team comes out of like uh the the the the folks at PopCap that worked on plants for zombies and Patrol Blitz um so we have like a little bit of mobile and PC console DNA but like the still the vast majority of our team is sort of just like geared in the sort of more like kind of League of Legends type uh game genre so it's just a different genre like I think the beamable guys come out of the Star Trek mobile game and it's just like so I think all that has plays into the the focus like you're gonna understand your the customers that you

understand the most like you're gonna serve the people that like that you've dealt with the problem the most for so I think like our guys coming in for that background are gonna deal with the problem differently than like folks that have dealt with it in other ways so like we're glad to see you know more entrance of the space because we just think the the number one competitor for us is people building it themselves and like the more attention that's around to um like back ends and like the the the push to have Studios like sort of offload that is the better it is for us yeah well it sounds like you guys have a fantastic product Market fit um which is awesome and um I have like I could probably go on a thousand years and pestering with questions but I know that we're um running up on we're running up on the hour and so I wanted to say like you know thank you for joining today I learned a ton um and hopefully our audience does too and thanks for hopping on the podcast um I know that you also are gonna be starting podcasting you're kicking off a Channel or something what's that all

about yeah so um we're starting a new podcast called the gaming Founders podcast uh over the last kind of like two three years uh one of the areas that I've been focused on is helping um game designers and folks raise money um and uh I've had literally hundreds of conversations about helping them navigate that whole Space so part of it was just like I was getting so many questions on the space and wanting to just come up with a podcast that deals with sort of origin stories how do people raise money how do they hire their first people where they find their co-founders from um and I I did see a product out there that focused particularly on those origin stories so uh Kevin Zhang who's that up front and me uh we've already recorded a bunch of episodes and we'll be uh yeah launch that really soon so super excited to to tell those stories yeah I think by the time this comes out maybe uh there'll be a couple on air so you can you can go and check it out um if if there are developers who are interested in listening to your podcast or interested in pragma or investors that are interested in pragmang you know

how can they get in touch how can they how can they listen to your to your new content yeah so the new podcast is uh pragmat.gg gaming Founders podcast and then I'm on Twitter at uh it's just at Eden Chen um or uh our website's obviously pragma.gg and go there and and find find out uh more their contacts us there as well that's awesome okay so on that note uh we'll be concluding big thank you Eden again for coming um thank you to our listeners and I'll be back in two weeks and so until next time friends feel free to hit me up with um any questions comments or concerns at alexandra.co would love to hear your feedback and with that we're out our war if you enjoyed today's episode whether on YouTube or your favorite podcast app make sure to like subscribe comment or give a five star review and if you want to reach out or provide feedback shoot us a note at podcast navic dot Co or find us on Twitter and Linkedin plus if

you want to learn more about what novik has to offer make sure to check out our website www.navic.com there you can sign up for the number one games industry newsletter navic digest level up your insights with our premium research platform novik Pro or contact us to learn about our wide-ranging Consulting and advisory Services again that is www.navic.com thanks for listening and we'll catch you in the next episode