Thriving in Steam Early Access: Turning 20XX's Slow Launch into Success

[Music] my name is Christopher king over the last couple of years it made a game called 20xx which we'll talk about in a bit we were in early access for a really long time on 20xx for about two and a half years this is a talk about community building and how to leverage your community during your games development specifically during access but I think the lessons kind of apply outside of that too we're going to talk about how to build goodwill and build trust in the early access community and then use that to make your game better so let's dive right into it with a couple of quick facts about 20xx 20xx is a co-op action platformer it's very heavily inspired by Mega Man X so the very similar kind of old-school action platforming gameplay it's multiplayer so you can play it online or on the couch and it's a roguelike so it features high arcade style replayability through random game generation and permanent death and it's out on every current generation platform on day one of early access we sold a whopping 64 copies of 20xx I was very thrilled by this I thought this was outstanding the idea

that this for this first game I had made it's 64 whole people were like yeah I think I can pay my money for this thing that you've made was outstanding obviously in hindsight that was kind of terrible and counts as a bit of a slow launch and here today we are going to talk about one of the factors that got us towards selling over 220,000 copies or 20xx as of today an awful lot of factors contributed to our success but we're here today to really dig into one of them which is our community management strategies just a quick outline I'm not gonna dig into too much here but there's gonna be a brief introduction I'm gonna hit on these six points about specific things we did with our community to either build goodwill build trust build loyalty or use it to get better feedback and in the event that I haven't run out of time I'm gonna take a couple questions so let's uh let's get into it what is this talk when you properly nurture your early access community you can get an a fantastic source of design feedback and testing and if you treat your fans right they

will treat you better and they will help you make a better game and they will love doing it so today's goal is actionable second level information about how to build community goodwill and use it effectively and when I say second level information I'm referring to action stuff it's one level of detail past what's blatantly obvious hopefully stuff you can go ahead and use today first things first though before we get into the meat of this talk it's point zero it's not one through six early access is not free this is a chart of my time spent on 20xx that 17% in the middle is my time spent managing early access community over our hundred and forty two weeks in early access that's two and three quarters years or so I spent about eight hours a week interacting with our community and as our games primary designer or creative tech guy level designer all that stuff that's a very very large cost it's about twenty eight full-time weeks or seven full-time months of community management and if you're not interested in really integrating player feedback into your game or helping your players help you design your game this is a huge cost and

you obviously have to account for this if you just assume that community management is free or effortless or doesn't take any time you're gonna be real disappointed so one establish a schedule and keep your word the section is about loyalty building as a developer you want to build loyalty because it increases your player attention it makes your players more likely to give you better more detailed feedback it engenders kinder communities and I really can't overstate how nice it is to have people not be dicks to you over the Internet especially when Steam communities can be very toxic places and thankfully ours was not and it makes players more likely to be your fans going forward so when you work on more projects the people who have played your games before will remember your positive interactions and be like oh I kind of have some some basic default trust with that guy I'm gonna I'm going to check that out so first thing a set and update schedule and stick to it this is a very concrete thing on 20xx we updated the game every other Wednesday at 6 p.m. Eastern throughout our 142 weeks in early access so that's 72 bi-weekly updates they were very very very on time

one of them was off by a single hour one of them's off by a day and the other 70 were smack on time and it really builds a sense of reliability in your fan base even if you don't have a lot of players immediately the players you do have will come to rely on the schedule that you've set and they'll plan play sessions like 20xx has to player online multiplayer so we'd have people like organizing sessions with their friends oh we're gonna play at 6:05 on Wednesday because I am confident that the game will have updated and we can rely on that plan and it really builds a lot of loyalty in your fans and they know exactly when they can come back and see the cool new stuff and if you take absolutely nothing else from this talk take that add an update time use your main menu here is an example of that so this is a picture of 20 X X's main menu at one point during development you can see that very professionally drawn green circle around next update in and then a timer very simple just look at the time that you're gonna update compute the difference between now and then and put it right on your app right on your front

page much harder than this is actually adhering to your schedule you know sometimes you're gonna have to make cuts to features you intended to push on time being reliable here is important and it's much better to release a small update now and then then be late and it's much better to not set a schedule at all then set a schedule that you reliably cannot adhere to if you're gonna do this make sure you actually can and finally on this make sure that the time you set is life friendly for you this is actually a mistake we made on Tony xx I set our update timer at 6:00 p.m. Eastern every other Wednesday and I lived on the East Coast so we would be pushing regular patches right before I was supposed to either go out to dinner or go to an evening social thing or do something right after I was off of work right and instead because of my very poorly chosen timer time I reasonably like I very frequently either have to push a patch and then just leave and not bring a computer and potentially have crashes or other bad bugs and life build for a few hours or I have to cancel going out or I'd be like hey friends I brought my laptop because also I have to

work pick a good time preferably right in the middle of your workday if we had done 2 p.m. instead of 6 p.m. we would have been much better off and we will do that going forward lastly on this guy never promised anything unless you're 100% sure it's going to happen I am sure you guys can think of more than one game community that's been kind of ticked off by a developer making lofty promises or specifically are promising specific release dates for features pretty far in advance and then letting their community down and this is really just about oops e1 back this is really about expectation management announcing features too early really doesn't win you a lot especially when you're on a slow launch or a relatively small team and don't have a Hugh playerbase and it risks our roading community faith and trust if those dates slip when you do tell your fans what you're working on pick better language instead of saying feature X is going to be available on date y say we are actively working on feature X say feature X is a really high priority for us and then don't announce a ship data until it is tested and ready to go

Oh actually lastly on this guy be open about your mistakes in a lot of cases making a mistake admitting it and then fixing it leaves you with a lot more goodwill than just trying to cover it up if you goof in a publicly visible way admit it and say we goofed our bad explain how you plan to fix it this is a high priority fix for us we understand there's a crash in the game we're working on it right now if you can't fix it explain why this is a relatively small issue and we've got a lot of hot irons in the fire we're gonna take care of it at this point ish down the line but of course note the publicly visible part there's no reason that you have to sort of make up fake mistakes or mistakes that nobody has noticed because they're relatively invisible you don't need to tell anybody that you did those sorts of things so when you patch a crash out announce it when you how to fix a balance issue announce it pretty straightforward stuff and it's not some kind of life hack people just like honest developers a lot of developers don't seem terribly human in the way that they interact with their fan bases and when you can appear human it's nice for you to maintain an active visible

relationship this section is about making sure that your presence interactions are known and felt in your community when you make changes nope that's one slide too many when you make changes that players have suggested do so very visibly call them out your patch notes post in the forums like if there's a steam community post where you've got a really good suggestion post after you've made it post when it's made it to the backlog and say hey this is a fantastic idea thank you closing that loop with a player and making sure that they're aware that their awesome suggestion actually resonated with you makes them a lot more likely to give you stuff like that in the future similarly take advantage of low-hanging fruit here if a user suggests something to you that is already in your backlog so it's not actually a new suggestion feel free to say that's fantastic it's in our backlog anyway keep early access sexy Armstrong gets a fade transition is only one be ready to shuffle around your backlog be ready to shuffle around your backlog yeah sure yeah be ready to

shuffle around your backlog once you're developing your game in front of what's basically a live studio audience you suddenly have a major priority that you weren't dealing with before the immediate now customer experience and your backlog has to take that into account you're gonna have to address player pain points sooner than you intend it if you know there's like a really bad quality of life if you in your game or if you know there's an awful crash you're gonna have to come back and fix those things earlier than you intended and you need to be aware that whatever your regular development process was isn't going to be exactly that way once you also have a whole bunch of users like telling you about their experience on a day to day basis similarly you might want to ration out exciting things you're working on if you have a whole lot of once like for example maybe you have 10 new levels that you ready to drop to new modes and you know that the next month of early access is a pretty dry time consider putting the modes in one patch and then one mode in another patch and then the other mode in the third patch so that your users coming back always have something exciting about your constant updating finally though don't don't ration bug fixes don't do that

basically if the thing you're working on the rule of thumb if the thing you're working on is a detrimental to player experience kind of thing don't ration that out fix it as soon as you can hot fix it if you have to don't wait for a patch cycle make sure that anything that might be turning your users off that is an unintended part of the experience gets out right away never ration that as an aside here's a quick example of our community forcing us to shuffle around our backlog so it's August of 2015 we're getting ready for PAX West and a major competitor of ours screws the pooch big-time and gives us a press opportunity this is another game pretty heavily inspired by Mega Man that had built a big hype wave and ended up kind of disappointing their fanbase and they gave us a press opportunity so we got together and we said okay what do we have what do we have that's ready if we want to sort of set an announcement if we want to set a milestone what can we put in this milestone to be kind of major and we said okay we've got a we've got a couple character redesigns if you guys have followed the development of 20xx at all you know that we we changed

the are we iterated on the art style numerous times I think we had six prynt main character designs between the very first steam launch and 1.0 a couple years later so we said okay we've got these new character designs are really excited about we've got a couple of new modes we can throw all that in there that'll be great that's perfect let's go with that so eleven days before this big beta launch which was sort of the day that our competitor was supposed to have had a major event we pushed that stuff to our steam community first to early access users to say hey can you guys go check this for us let us know if anything here is super broken or if there any like immediate feedback bits you have or anything like that and this was our response these are a little bit cherry picked but this was pretty much the response of our entire Steam community it was overwhelmingly the most negative feedback we've gotten about anything at all we've done with 20xx we're 11 days before our biggest press push of the entire development cycle and we're getting this about our main character so we scrambled to fix it we canceled the vacation we build new character models using a lot of the

lessons that we learned from the you know sort of failed character designs we had and of course our characters feature very prominently in all of our promotional media they're in our trailers they're in our screenshots so a couple of days before this major event we have to remake all of our trailers and all of our screenshots and we we end up doing all of this we loop in the community with every single step the moment it is clear that we have you know something approaching outrage on our hands we say oh we we goofed we're working on this you know please bear with us trust in us we're gonna work on this and every single day we would post our new concepts we would post models in progress we would post animations we would basically just show our entire community day by day this is the new stuff we're building does this fit your expectations better and if we hadn't done this we would have just gotten toasted honestly this was a really big press push for us there's a really solid chance that if we had pushed that beta with unappealing main character designs that we just would have been dead in the water right there like this moment is one of the major sparks that made us successful and we

almost wasted it this was some of the feedback we got after having done this so keep in mind that these were community members who were downright pissed we had turned their beloved characters into something they found so hideous that they couldn't even play with they didn't want to look at these characters while playing the game and even though we had made mistake of that magnitude once we sat down with them when we fixed it they loved us and we got a lot of this we got a lot of I can't believe you guys were willing to go that far for us like thank you for listening to us and I like thank you for telling us we need that stuff three develop and maintain a group of super friends these are these are early early access players they're sort of a side group from your basic EA community that you're gonna use to test content that isn't ready for your entire user base they're gonna get to see cool new content before it's ready for release or maybe it's work in progress or experimental stuff and you're gonna get last line of defense QA before stuff goes live you're gonna get an awesome source of feedback before you're really ready to show your greater player base

something you're working on feel free to handle membership in this group however you like you can opt in you can have people send you a nice email or say I like your video game that you're making you can you know just have it be invite only if you have a big enough community that you can afford to cherry-pick as long as you're getting the feedback you need the way you handle membership the size of your group doesn't really matter set your group's expectations accordingly if you're superfriends content is going to be full of garbage or pushed relatively and frequently it's very important that your fans know that you can afford to push less frequently or less consistently you can afford to push relatively lower quality stuff as long as the people you are pushing to understand that that's what you're doing if they are expecting the same level of quality in your regular releases and then you push them kind of garbage you stuff they're gonna feel bad about it respect the shared value relationship here remember that your super friends are a two-way street you're getting something really valuable out of them because you're getting sort of an earlier line of feedback you're getting awesome testing from them and they're getting to see a lot of cool new stuff

that you haven't actually pushed yet before it's ready accordingly don't push Super Friends content that are only full of relatively boring stuff if you have performance enhancements if you have engine changes if you have things that are not really impactful to the average user experience like content really don't push these things to Superfriends at least don't push them alone if you have to test this stuff maybe you have engine changes that you're really excited about or you have optimizations you've written in that you're not sure are safe because they might break other things that you can't really test on your own either wrap the with other stuff wait until you have other content to test to push them or push them to just a another public-facing build that isn't superfriends say this is an optional test build on Steam this will go live in a month whatever like that lastly your Super Friends are not contractors they're not employees they won't have signed NDA's unless you make them which we did not and if you are gonna have something that you could possibly show that would ruin your plans if it gets out like these people are in your Super Friends group because they love you but

you can't just explicitly trust them a hundred percent don't show anything in this group that will absolutely ruin your plans if it gets out it's very likely that if you're you know kind of a slow early access launch this doesn't apply to you and if if the special stuff you're working on gets out to whoever actually cares about it at the time that's probably fine but just be aware for interact efficiently sections about cutting down on the time that it takes to do this job right encourage public space feedback so when you're communicating with your players encourage the use of forums encourage the use of discord encourage the way are encouraged the use of any kind of public space where your words have a force multiplier where when you respond to something you can reach a whole lot of players at once instead of any one-to-one mediums like like email or Twitter or calling me at 3 in the morning reducing the number of times that you personally need to deliver a given message saves a lot more time for other stuff just makes this take less time similarly once your fans are using these public spaces once you have veteran

players in your Steam Community in your discord chat stuff like that they're gonna answer a lot of questions for you they're gonna take over a lot of the softballs if they've seen you answer a question once they're gonna come in the next time somebody else asks and either directly link to your comment or say here is the answer and it's a huge win win win you don't have to spend any time on this at all your veteran player gets to feel like they contributed to the community and they get to feel like they helped somebody which they did and the fan gets their question answered everybody wins and it's awesome here is an example of how we tried to encourage this with 20xx using the same awesome main menu screenshot you can see on the bottom there we've circled our community button which of the time linked directly to our steam community cuz we hadn't done discord yet and we had a similar circle over there linking people to the subreddit in hindsight I don't know why we didn't just make that a set button we probably could have gotten away with that pretty easily this is also a perfect example of a mistake I have an email address on my main menu screen and I have no idea where people are getting this email address but to this day I still receive support emails for 20xx even though this has been gone

for literally two plus years somehow this email address is circulated and now I'm doomed and stuck with it forever always encourage public spaces never encourage one-to-one mediums of course there will be some users that are not comfortable speaking out in the public space and they will find your email or they will tweet at you anyway and that's totally fine but in general make it clear that your preferred method is anything public like this use a public bug database you're gonna get fewer bug threads any other communities if you're not already using a public database please do we use mantis but there are a ton of really good ones lead all of your bug complaints there get them out of your other communities it turns them into more positive places you know how many times have you loved playing a game and then the subreddit or the community for it after a big patch just turns into a big mess of bug complaints and nothing else it really it really drags you down strategy to does apply here you have to make sure you're visible in your bug tracker if you have people reporting a lot of bugs you need to make sure that you're spending at least some amount of periodic time looking your bug tracker visibly fixing bugs marking them

complete submitting notes for feedback stuff like that otherwise people are gonna come to your bug tracker they're gonna see a whole mess of open bugs that haven't been interacted with in a long time assume the bug tracker is dead and either not report the bug or reported on one of the other forums pick your battles don't respond to everything over time you're gonna get a really good sense of the kinds of messages that your veterans can handle for you ignore them entirely let your veterans do it take the time to respond to a player when they just make up a love thread if they say thank you developers so much of this game you know put yourself in their shoes if you are telling the develop your developer of your absolute favorite game how much it's meant to you and how much you've enjoyed playing it there's really nothing that would make you happier than that developer coming right around and saying god thank you this is the kind of thing like this is what I love this is what keeps me going and closing that loop really cements them as a fan pretty much permanently this is a very easy way to convert somebody who likes your game into a permanent and respond to popular feedback that you disagree with if there's a long thread

of comments or a big discussion about something in your game that people want changed that isn't aligned with your vision but you don't intend to do your fanbase is going to understand if you explain this is where we are and here's why instead of just completely ignoring them lastly you can respond to feedback you should respond to feedback if it's not very popular but is a good idea you know the whole point of doing all of this is to get as much good feedback as possible and it only takes one person with one idea sometimes even in a single sentence to get a point across and responding to those sorts of people makes them a lot more likely to give you good feedback in the future ah don't respond to everything unless they're negative steam reviews negative reviews can be a really good solid source of actionable information for example if multiple users post the same technical issue I crashed when doing X and this is my negative review you have no idea how many players are experiencing the exact same issue and just not telling you they're either walking away from your game they're leaving a negative review

either way they're not your fans and they don't like your game because of this thing if multiple users post the same misconception and we'll talk about this in a bit you probably have a messaging problem if a lot of people are posting negative reviews in your Steam review section and saying I didn't like certain thing about your game that you didn't actually intend for people to think you probably have some way that your message in your game or some way that you're appealing to users that doesn't align with what your game actually is and if users have a pain point about the game's design that is intentional for us we were very heavily inspired by Megaman X the game is very challenging the action is very precise and for some people it's just real hard and they want to get into that kind of thing and it's just not their cup of tea because it's real difficult you may or may not have actionable information like sometimes your vision is very solidly on the side of well you know this game just isn't for these players and that's kind of how it is but sometimes you may realize that you can do something relatively small to alleviate those pain points as an aside just a short story about how we learned something from our steam reviews in this super great

pie-chart v I guess the the green is the 31% of oh I didn't want to rogue complaints so just as a brief refresher 20xx is an action-platformer where everything is randomly generated every time you play it so you go through a set of levels you die you retry and we a lot a lot a lot of negative views 31% of them where it basically boiled down to I wanted to play this style of game but I didn't want a roguelike and we were very sure that we had covered our bases on this we said roguelike all over the place on our steam page pretty much all of our press stuff all of our marketing sort of showcase to this this game loop and for some reason a lot of our fans are a lot of our unfinished more really getting that and we just sort of figured you know some people aren't going to read your game store page some people are just gonna look at your game they're just gonna buy it and there's nothing you can do about it and we eventually had a fan point out hey maybe you know a games are really these days they're implementing a lot of different kinds of sets of roguelike elements like some of them don't do permanent death some of them do permanent fresh and some don't if you laid out exactly what you're doing not only would fans of roguelike games

understand exactly what about your game is a roguelike but people who don't understand roguelike games would also know exactly what those terms mean I'm really oh jesus they're right oh man so we did that and those reviews basically vanished we stopped seeing that influx almost entirely and this was kind of a painful point for us you know we stopped the bleeding but if we had done this off the break if we had messaged this properly from the get-go there's a really good chance we'd still be over that 95 percent review threshold on Steam that overwhelmingly positive and we believe that is an impact on sales but we don't really know 5 I develop a motion agnostic evaluation you have to learn to read feedback out of context when you get a suggestion when you get feedback strip it of any context whatsoever ignore the way it's written ignore the tone of the writer ignore your relationship with the writer ignore the fact that the fan might be angry with you is the suggestion good for the game and is it within scope it's all the matters angry fans and players are often gonna have good points and they're just mad because they really care about your

game and something has happened that is not the way they wanted the game to go obviously this has its limits if you have and have said are angry user who is resorting to either harassing you harassing other players hate speech anything like that ban him straight away don't deal with that but somebody who is upset is not somebody you should inherently ignore you know try and forgive their undiplomatic approach to contacting you similarly don't reject feedback just because it's poorly written or poorly worded we have a user who speaks in some of the most broken English I've ever seen it is as though I can't tell if it's it seems like he's kind of speaking in memes all the time but he's still like one of the best players of the game he has really sharp suggestions he it's like specific design critiques and taking the time to like try and actually understand what he's saying and sometimes it takes a really long time has been entirely worth it for us on the other hand don't favor suggestions just because the community is clamoring for them or they are popular or because they are from your favorite players and yes we have favorite players popular suggestions aren't necessarily good ones and lastly hone your evaluation skills

all of this is about getting your community to trust you enough to give you all the feedback you want and if you can't decide what good feedback is from bad all of this is kind of worthless unfortunately this is a it's a little bit subjective you have to define the ways that you judge feedback for us how well does it fit with a game how well does it fit with our current community because once you're making a game on early access you have to take that into account what's the scope cost of this and what's the impact is this gonna be good for my game is it gonna make my fans happy can I implement it quickly enough to be worth it do I have time and is it actually impactful you might have things that are really good fits but end up when you think about them being kind of very minor or maybe only 1% of your fans are going to see them and so they're nowhere near as important each dimension obviously is weights that are going to shift over the course of development when you are right at the beginning of production you have plenty of scope probably as long as you have money to deal with it as you get closer and closer to shipping you know you're probably gonna have to cut stuff and you have to be looking at great feedback you've already taken that you can no longer deal with by 1.0 so when

evaluating suggestions try and maximize you know positive community positive game stuff per unit scope and as a final example of some of the concrete value we've gotten from our community net play testing sucks net play testing sucks really hard setting up networked bug conditions even if you get full repro steps setting it up and testing it and trying to figure out what the hell is going on is so much worse and so much more time-consuming than almost anything else we've dealt with and as we built the games online community we got hundreds and hundreds and hundreds of hours of co-op footage of people playing the game with a friend online with fraps open or something like that and sort of bug hunting they'll play the game and then they will go through their hours and hours and hours of footage and they will clip out all of the bugs and they will say here is a bug real and we they'll give us a ton of information on every single last little thing so not only will they post all of that stuff in our tracker very neatly for us but they even have video footage for everything and without getting the respect of our community if those fans

didn't know that I was gonna go over every single inch of that footage and fix as much of stuff as I possibly could they wouldn't have done it we got you know thousands and thousands and thousands of dollars of free value out of them submitting these bug reports and there's no way we could do Shep 20 xx's online multiplayer at one point Oh on time without their help . on top of that most of your fans that are reporting these bugs do so with a really intimate knowledge of how the game works they've played for hundreds of hours they know exactly what this looks like they have a good feel for why something might be occurring because they know how your systems interlock better than like third-party QA might obviously internal QA should know the game just as well as anybody else in the team but if you try and use external QA that hasn't really seen your game a lot before instead they're not gonna do as good of a job at least in our experience this is a blank slide so a little bit of additional reading two years ago after we finished 20xx one point on PC I wrote a really really really detailed two post-mortem on gamasutra that has all of this stuff and an awful lot more less any kind of

stuff that I learned over the course of making 20xx for four years and I have two additional articles here I linked an article about community building on discord from Andy Winn in case you wanted more concrete stuff on how to actually get more eyeballs and build awareness of your community while you're making it and I also linked Tanya shorts excellent medium article about building boyfriend dungeons community pre Kickstarter I just sort of see those as supplemental information for the kinds of strategies that I used while making 20xx because they're much more about getting the people there in the first place and my stuff is much more about keeping them engaged keeping them loving you and making the best use out of their feedback and that's my talk thank you [Applause] we probably have time for one question if you have a question for Chris please come to one of the mics and in the aisles and while we wait to see if anybody else has a question I have one for you so how did you keep your update schedules consistent even when you were doing things like packs because it's

pretty dang disruptive so how did you manage to actually release something two weeks later we would just plan all of that stuff out far enough in advance if we knew that there was gonna be a week where we weren't going to get anything done we would basically finish an entire patch way ahead of schedule and then just release it when it was time um it was just it's all about planning right hi I was wondering how many people you had and if each person kind of took a different median like one person did read it one person did steam we had two full timers sorry but the question was how many people did we have working I guess on the project or just the community aspect of it there were two full timers on 20xx and then a big army of part-timers doing things like sound and music and stuff like that in terms of community engagement it was basically just me . this might be something very particular about me as a person but I am very anal when it comes to messaging and making sure that my way of communicating

with the community is is upheld and as a result I just kind of did all of that work myself you know every so often somebody else working on the project would interface for the fan to sort of you know happenstance ly but it wouldn't be like this is their job to do it they would just see it every so often so it's really all me is that Thanks all right all right so please make sure you fill out your email surveys to give Chris feedback on his talk thank you so much Chris [Applause]