Poly Bridge and Social Media: A Love Story

[Music] hey thanks for coming so yeah my name is Patrick as you can see up there I'm the developer at dry cactus and we're just so a very small studio based in New Zealand we made a game called poly bridge and we released it at the end of June 2015 and Steam early access and we had a few unexpected and unusual things happen so that's why I'm here today specifically these are the most interesting things so you might have seen some of these pop up and some like social media such as reddit or Twitter in the last kind of six months what we did at the game was kind of integrate a replay sharing tool that made it very easy for players to share your little moments with their social media timelines and post them wherever and it had a very you know welcomed unexpected kind of flooding effect in the sense that we didn't really anticipate the kind of exposure that we got through social media and how that translated to

sales and financial viability of the game so what I'm trying to do today is kind of go over some stats and try and identify the link between exposure and social media and how that translated to cells and how eventually it created this kind of feedback loop between content creators on twitch and YouTube and people sharing on reddit and so on first of all though I just want to have a quick intro what the game is about because I really don't expect most people to know about poly bridge so it's a very simple bridge building game at core I'm a big fan of bridge building games as one of my favorite kind of genres when I was a kids particularly the pontificating ones familiar with that from chronic logic was like the best bridge building game ever still so we took you know a number of ideas from those games we did try and take in a different direction though so as you can see from some of the images and also these screenshots specifically we try to kind of take a step back away from the seriousness they are bridge building and kind of physics games in general and it's you know it's a train that's been

going on you can see it like in the hugely successful Kerbal space program for example is while you know super realistic with all the physics and so on it has a more playful approach and makes it look less hired and less kind of engineering which is a big barrier I think to adopt in these kind of games so one of the first choices yet was kind of have a simple aesthetic style limited color palette make it a bit childish and just when you see the game I didn't want anyone to say oh this is like it's an engineering game it's not my kind of thing you know so we try to break down that wall first of all along with that came some other choices like we tried to make a safe space for players as we were calling it so we don't punish failure in any way if anything we try and encourage kind of trying this trying that it's very quick to try a bridge it collapses I'll try another status draw it and move this thing around and kind of make it very easy to iterate on your bridge design instead of feeling punished for failing so we don't you know don't know failure along with that Adriane talents who's a very talented composer based in Canada

did a really nice soundtrack which goes very well with creating the mood of we call the anti rage soundtrack because the game does really kind of push the button sometimes if you have to spend 45 minutes tweaking the brute design and it keeps collapse and it's not much fun so to try and counter that feeling of I want to kill the game developer we put this kind of soft mellow guitar instrumental soundtrack which is really nice the niche is this word that is tossed around a lot there might be some other developers here who kind of focus on niche design it's a bit of a broad word doesn't really mean too much but the bridge building game is kind of a sub Anish of the physics building ish in itself and what I did when I started this project a couple years ago and I wasn't alone it was shortly after my partner became pregnant and I had to kind of decide what to do in my life so it was either find a job and do that kind of thing or try and do another game and the only successful game I done previously was a bridge-building game so

taking the lessons learnt from that I really thought I'll just do a kind of small niche game eventually the scope kind of got bigger over time but it's something that many other developers kind of feel the same is if you're going to make a game and you don't have much money try and do a niche game so that's kind of what I did and they paid out in a way along with that the sandbox is a big thing you know very trendy as well like low poly straining so we try to identify train so to speak in indie games in what players want so sandbox was a big thing is like so you're not limited in content to what the game developer gives you but you can kind of do your own thing and it's integrated with Steam Workshop and so on and the other big word community which means nothing and everything at the same time we kept tossing this word around as we were designing the game developing features and deciding what to do what direction to take in we kind of thought communities has to be kind of the

foundation of this game for it to have a long-term linearity and viability and so on so throughout the game design we always try to keep in mind how can this feature specifically work and integrate itself with the community forming and building that community and kind of keeping people happy at the same time yeah okay so again just to get an idea of what the kind of style is very talented artists whose name is Xavier is based in Spain and he's here with us today came up with this kind of low poly style very limited color palette so hopefully it's inviting to the eye you know it's not like an eyesore and when you see it you might think oh it's cute it's you know it's something a bit different at least for a bridge-building game so I think that contributed a lot to kind of the exposure we got and going you know moving forward so community you know we kept tossing this word around not really knowing what to do with it and how to try and build a community and have people actually positively interact

with off as game developers in between them and kind of take the game in a new direction so this is just a short recap of some of the choices we did as we were developing the game the main one being the replay recording and sharing which is kind of why I'm here today so I'll talk about a bit more about that shortly along with that we kind of integrated some other tools the galleries so when you share a replay in the game it goes to an on one anonymous online gallery where people can then log in through Steam and Ray to comment and kind of filter through submissions for specific levels so it kind of gives the content creators that give the replays to our gallery a bit of feedback because you get kind of your little star rating and comments occasionally and it just allows kind of other players to see what people are doing and the most important thing for us as developers is that we're able to see very very quickly what kind of ideas people are coming up with when playing the game and this has been a huge factor because as we moved forward with development shortly after darryl access

launched we were able to see all these new ways kind of a new mechanics to play the game that we weren't aware of and by doing so we kind of then started designing levels and design features of the game that brought that into the game so a very good kind of feedback loop so to speak was created between the early adopters of the game and nasa's game developers in trying to kind of you know everybody's happy at the end of the day another thing we did was the feedback tool i attended GDC last year and there was a talk from one of the developers of subnautica which is that kind of underwater diving game rhetorical and they had a whole talk about how they developed a feedback tool that allows players to send the little item of feedback without ever leaving the game so a panel just pops up you can say for happy side or medium you can give them a comment and it will also send a screenshot and the d'bari log of what's going on behind the scenes so when a player identifies or a bug happens hopefully they will send us a little feedback item saying this bug happened

I'm not sure why and I get to see the debug log and I can hopefully identify why ending bug kind of happens so it's semi bug tracking semi kind of feedback and again the loop between the player and the developer kind of trying to narrow that gap and make it a bit more a bit more kind of meaningful fitting for players to interact with the developers again the sandbox and the workshop a huge part of building the community because you know that's where the community forms is when people get to design levels and send them off to the Steam Workshop specifically in this case and players can download them so along with that we also created the link between the Steam Workshop submissions and the replay gallery so there's it's time to kind of pay off now so to speak seen lots of replays so as a player of the game if you make a sandbox level you submit to scene workshop for other people to solve you can then see when a replay is submitted for your level you can see what players are doing with your own level so this kind of gives the this level designer a type of feedback that

is rarely seen in other games which i think is a bit of a shame because we design a level you just send it out and often it kind you can see maybe some comments but in this way you can actually see the bridge designs that are being made with your kind of layout and it makes a really kind of nice feedback in tying in the community and replays so just a little bit talking about replay and kind of the technical implementation without going too deeply into it because there's lots of stuff out there and I'd be more than happy to answer any questions later about the more technical stuff but what we're doing is we're always recording footage as the simulation happens we record that at 15 frames per second for technical reasons and also because animated gifs or gifs the pinyon way from kind of a standard of 15 frames a second it's you know the animated gif is a very ancient file format it's like from 1990 I think so it's not very well compressed it has a very limited color palette which is 256 colors the files get very large very

quickly and obviously the challenge there was as we're recording footage it goes into the vram so we can't obviously take up too much of it around otherwise lower end devices can't play the game and also when you go and upload the file itself you want it to be quite small because it doesn't you don't want it to take like you know a minute to upload a replay it kind of breaks the flow so we you know keep those quite small so we're recording for three at all times this kind of was a choice to try and you know kind of break down that wall I was talking to many other the game developers who are doing similar things and some of them are saying you know we put these tools in but we're not seeing the kind of interaction that we were hoping for and it's often because they require you to log into your YouTube account you have to do all these extra steps which players such as they're not going to leave the game to go and upload you know something to YouTube or to the social media platform some of them will of course but you're just like a good 90% making it up obviously if people will not just bother pretty much so what we try to do is you never have to leave

the game with a single click it generates an animated gif of the buffer that has been recorded of the last 10 seconds and then it saves it locally or it can upload it to our online gallery which I talked about and social media we have Twitter integration at the moment and are working on some other ones so once the authenticate the first time in game you just place a single button share it saves you replay uploads it to your Twitter timeline to the online gallery to your local hard drive so then you can repost it wherever you want and that created the whole kind of feedback loop which we weren't expecting if you're interested in the tick it's open source it's actually Zack Bart who's a developer on Zack Tronics and recently released t is one hundred and infinifactory he actually wrote the encoder and he's very kind to give it to me and then it was released open source in a bit market repo so it's it's a software for everyone to use it's C sharp obviously obviously a bit C sharp and it comes with an example to integrate into unity very quickly so definitely you know if your game allows

for this kind of feedback number one recommendation would be to integrate something like this and think about it you know so ya covering this what I'd like to then go on to see is the more financial kind of aspect of the game so this is a bit of a confusing slide so I'll give you a few seconds to try and kind of make sense of it it's a bit hard to kind of it was hard for me to identify the timeline of events so it's even harder to try and communicate it in a meaningful way but this is the best I could kind of come up with at the time so you can see on the left you have like a linear time line and on the right is just a kind of meaningless diagram of how things kind of feed back into each other so it's very hard to say oh this first happened on Twitter feed and then it fade into this Twitter stream which then means somebody posted it on Reddit and then the YouTube guy came along and so it's just you know I just thought it's just a feedback loop between all the different platforms and the content

creators they use those platforms but what did happen though is the first kind of couple month is kind of what you're seeing in the left timeline so we launched entirely axis that was a 30th of June we had some positive reviews because we had worked with a small PR company to get some people playing the game early on and do some YouTube videos pre-launch which helped hugely I'm pretty sure but so on launch we had 15 positive reviews I think and a bunch of YouTube videos were already out there on launch day which was a very big thing and kind of creating the positive image of the game so along with that shortly after launch we start to see in a very steady stream of of content going on the Twitter timelines lots of people sharing the replays on Twitter which were then getting retweeted and kind of starting the snowball effect of exposure of the game which is ultimately what obviously we're trying to get because with exposure comes potential sale but without exposure you probably know where unfortunately so then we started seeing more YouTube content coming through the first twitch stream started happening

and kind of generating pretty sharp spikes of sales when the twitch stream were happening reddit was the main one as we're going to cover shortly and then shortly after the rative kind of taught page we started climbing up the steam charts and we're in the top 10 for a couple of days which obviously created more interest and you can see quite a while down the timeline we then started getting mainstream media coverage so it's a bit unusual but you know it's something that's been shifting so much in the last couple of years is you no longer aim as much for getting that coverage from kotaku of his mode or IGN personally you know what happened for us was seeing the coverage from mainstream media would bring virtually like it would it was impossible to identify when an article and IGN how that would translate to exposure in sales it just it wasn't even noticeable on the sales graph while if you see a twitch stream or like a reddit post it gets up both as a lot you see a very very clear spike in sales so as an

indie developer I guess you're aiming more these days for getting the exposure on social media and with content creators on these new platforms and not so much with the traditional media so to speak just because the level of exposure you get is is nothing compared to what you can get you know a reddit post so to try and kind of bridge the gap so to speak funny in bridge pun between sales and exposure on social media what I thought I was going to do well on doing is looking at the first 30 days of sales and so with these 30 L sales as you can see here so quick sides and the first 30 days we sold just over 40,000 copies at full price which is 1199 and it made a bit over a half million gross in revenue this is aggregated data from the two platforms they were selling on which are steam and humble and so what I did here was identified these kind of little events that happened the first month you

can see so RL is the launch you can see it kind of goes up a tiny bit and it starts coming down and until we hit t1 that was kind of what launch was gonna be if we hadn't integrated all the kind of social media and replay sharing in his community tools I am really positive this graph would have just kind of stopped that VL and kind of slowly gone down and just tailed there what happened instead are these other events that you can see obviously kind of crushed the launch held and so going to yeah dive a bit more deeply into these so we're going to ignore launch but you can kind of launch we had the maybe on launch day about 300 sales and then 250 200 and so on so it's going down pretty pretty quickly and so we have t1 t1 is the first major event it's a twitch stream we peaked about 2,000 viewers as you can see from the graph so while these are daily sales so first 30 days these are

now hourly sells so you can see on the Left we're tailing at about 5 to 10 sales per hour and that day and then we get a huge spike it's about a team and it was a the first spike and about 100 and the second spike what's quite interesting with twitch which might be obvious of Salman was not obvious to me because it was quite new to me the whole twitch stream in YouTube and so on and with a twitch stream you if you often get these spikes of sales during the twitch dream and as soon as the stream ends you kind of just go back to where you were so it generates a small and short-lived spike of sales potentially obviously it generates a longer-term interest in the game because people talk about it to their friends and cover it on YouTube and so on but as far as immediate visibility of increased sales there is virtually none you just see the spike and it goes back down so that was a first too extreme and you can yeah pretty decent spikes and you know we were very happy at the time was like or

this is very unexpected wasn't really anticipating the game picked up by twitch streamers but it was pretty cool yes so the second spike t2 it's just here that's another twitch stream this was a larger two extreme it had about 16,000 concurrent viewers at its peak and similar effect we were tailing at about twenty up to fifty sales per hour because of other er like YouTube was starting us to pick up so daily or hourly sales were kind of increasing slowly on their own but then the twitch stream happens there was the one spike because you can see in two hundred there shortly after that twitch stream another twitch stream with slightly fewer viewers so that's the second kind of hump you see after the 200 mark but again interesting to see how the end of the sales graph kind of matches the beginning of it so we had the spikes and then we went back down a couple days after when I was working on the game so we released the game and I was somewhat obsessively looking at the stats what

you can do on a steam store page is link a Google Analytics tracking code so then you can go on your Google Analytics dashboard and see what's going on and you can see how many people are viewing from which platform to come in there's some issues around tracking where traffic traffic is coming from but overall it gives a very good idea and what I was doing I was keeping a little window open with the live stats so you can Google Analytics tries to tell you how many people are currently on your page and that was a way for me to try and identify very quickly when something happened because like these two extremes started to happen and another event was going to happen and it was very important for us to try and keep track of what's generating a spike in sales and so we can kind of leverage on that in the future and try make features that encourage that kind of thing to be shared or whatever so a few days after the second to extreme I was looking at in my life stats which were generally hovering between 30 and 40 live concurrent people on the store page and I just turned around and it was

about 1,500 and I was like there's something suspicious going on and because I'm not too much of the digital generation I was aware of Reddit but I did not frequent read it as such so I wasn't too familiar with how it works and what goes on so a friend of mine sent me a text saying you should check out read it and then I saw at the top of the front page and read it and it was a poly bridge gif which is the one you actually see in a top right incidentally that's as you can probably tell kind of a physics glitch what's happening there is that the monster truck has infinite friction on its wheels so when it hits the edge it just kind of has to go somewhere and it flips wildly back somehow it actually redoes a thing and hits the flag and wins the level so it's kind of a you know a very happy glitch so to speak it was the first poly bridge gif they got lots and lots of traction and got voted so much it was top front page and read it for any person subscribed to the gaming sub or also just anonymously browsing read it so you can see the spike there pretty much these are our

early cells we showed up to a maximum 350 cells for an hour and interestingly enough when it goes back down it doesn't go as down as it was at the start so rate it has because it's not obviously a twitch stream it doesn't kind of end suddenly once you drop down the front page you're still have a certain amount of visibility and that just created the kind of phenomenon of the meme so to speak so somebody actually published a page on the website called know your meme calm and poly bridge gif sharing became kind of a very you snitch meme where people would play and say here's how an engineer plays poly bridge here's how a student plays poly bridge here's how whatever my job title is plays poly bridge so we tried to see in like some areas things and it just translated to huge sales and it kind of just kept growing and when I read a feature would drop off front page it was still like holding our resale will still hold in very well it wouldn't just drop out of nowhere so

definitely the fact that it got picked up by reddit players was the main thing that drove sales running a bit short time so very small stats might be relevant depending what you're kind of trying to accomplish on reddit these are stats that are new to me and a bit hard to find so that's why I thought it'd be maybe useful to share these so on the left you see the top ten poly bridge gifts they are they come up and read it they've accumulated roughly twelve million views probably a few more if you want to go you know if you're trying a mean for the front page of course you have to get at least two thousand votes up to five thousand if you want to be like actually in the front page and not only in the top of the gaming subreddit when we did get a front page gift it would easily generate one two two even three million views so if you think about like just you know we're independent developers we're trying to make a small game and suddenly we're getting four or five million views every couple days just through rate it was just you know insane for us what is hard

though because on radio you just share it'll give luckily most people would at least put the name of the game sometimes but if you want to then find out more about the game most likely you're just going to go in Google search for poly bridge so it's very hard for us to track and direct traffic from reddit to the store page so it's it's kind of guesswork but it's pretty obvious when you look at the sales stats and how you know you see a rated thing and then you see the spike in sales so but it's very hard to say track the long-term life of a Reddit kind of feature the other thing being YouTube lots of videos on YouTube for some reason or another very big in Brazil in some other South American countries so you did top nine are from non-english speaking countries and then PewDiePie did kind of a very short video and misspelled the name of the game thank you but still they had a few million views you know the top ten videos accumulated about 70 million views there's 20 out whatever number of

thousand videos have been posted and YouTube does have a very long-term impact on sales but again a bit hard to track because generally players don't seem to click on YouTube and then go and buy the game they might be interested but later on they'll Google it and so it's hard to track the direct correlation between a YouTube exposure and how that translated to sales and how many people you know what percentage of people actually bought the games on and this is the dancing banana which you might be familiar where this is the top result if you do a google image search of animated gif so I thought it was relevant and yeah if anybody has any questions there's a few mics or you can just shout out a null repeater and also because we're all kind of running out of time I'll be across the room in the wrap-up room if anybody wants to have a chat or whatever I'm very open yep hey um so other than having in-game videos and gifs making it easy to share other than that what to get the the twitch players to stream

your game or their editors to post and upload your game what can you do to to enable what what can you do to make the twitch players play your game and the writers to post it and upload yes it's a bit of a hard question I guess like so just I'll repeat the question real quick in case you know we've heard over for the video he was asking like what can we do to try and encourage twitch streamers or editors to actually share the content so what worked for us was trying to knock down his barriers with the sharing so within the game it's literally a single click and half a second and then you can keep on playing and then it saves your gift locally and uploads it so I think that was what created the initial interest in the game and as far as twitch streamers go we really didn't do anything though - we didn't reach out to twitch streamers it was just it was natural discoverability through the social media platforms they did everything pretty much it was largely not anticipated in our part and mostly mistakes of the owners that just kind of

happened and it was a very happy accident so to speak but we did contact two weeks prior to launch we worked with evolve PR and did they contact some YouTube and some twitch streamer as well which kind of smaller channels the ones that are covered as far as the events in the sales graph that we looked at those were completely organic and there were larger streamers that normally covered minecraft specifically and for some they decided to coverage our old friends I think and they just covered it and yeah so kind of organic in that sense just I think just because so many people were sharing these gifts and then doing on radio as well just because they felt like doing it like there was not an encouragement on our my power necessary to do so yeah sorry I think I was gonna ask something similar but uh how do you think this will affect like the next game that you haven't development knowing like that this gift stuff and this twitch is really a really effective marketing strategy yeah so the question being how does this going to affect our

future development on other games with the gift sharing and so on specifically we're not working out of the game we're a very small team so we're still in the early access so it's just working full-time on this if I had to take a gift though I would try in any future game I work on the core design will be keeping in mind how a player can own the content that they're making this worked really well in our case because you're building a bridge and every Bridge is a bit different so when a player sees something funny or it's quite ingenious if they feel like they own in that content they came up with that bridge so it's going to motivate them to share these these moments and that's a huge factor which other games had tried to tap into but maybe because they're quite linear and the end result is going to be the same for everyone it's not a personal experience so I'm not going to be motivated as a player to share so you're definitely kind of keeping in mind the game design with the player owning what they're doing kind of in a sense hey Patrick great talk

I'm curious if the analytics you put in to track your spikes and correlate them with social media stuff yeah was all of that or most of that in ahead of time when you launched or did it become apparent that oh the stuff is feeding into each other we better start to investigate it more yes a question being what we did with the analytics pretty much and whether we had in from launch and so on well I'm not the biggest adopter or I don't really know much about analytics so luckily when you set up a steam store page so all the analytics come from sea pretty much so you can set up a Google Analytics code in your steam store page and then that's pretty much all I did as far as timing I set up pre-launch because it seemed to make sense and it was recommended by some other developers otherwise I probably wouldn't have been aware of it and yes so that's kind of all I did I just put up the Google i/os tracking code and then I would keep a close eye on live start specifically and

then also getting an idea of how many views that we have yes in a store page and what percentage of those views came from a specific platform the problem though being that is very hard to track direct traffic because of the way you discover a game on Steam specifically it's very very vague where the traffic came from so for example I could be tracked traffic from reddit it says it's like 3% but it's obviously way more than that but because people go through Google search engine to actually find a store page there's a disconnect which is very hard to kind of do you know to make sense of it but that's what we did really wasn't too much in the analytic space it was kind of just keeping an eye what's going on it was useful for that but yeah cool ok thanks [Music]