Designing Episode's Interactive Fiction in Three Phases

[Music] hi there thanks everybody for coming to the talk I'm Michael Dawson gonna talk about designing interactive fiction on episode in three phases so I wanted to start out okay but what does it mean to design great interactive fiction that people love and to do it on this device it means putting together two very distinct worlds one world is I think well represented by Hollywood which is really best on earth storytelling the other world is mobile which is both the largest games platform on earth right now as well as maybe the largest platform overall that we have to work on so today I really want to walk through what are the parts we on Episode have taken from both of these worlds to put together hit interactive fiction this is way of background not kind of me and episode for those of you that might not have it I'm Michael Dawson I leave the studio at episode which was created by pocket gems in 2014 it was created by seeing this real opportunity to

unknowable to say what does it mean to tell a story on a mobile device and so we've launched since we've launched we opened up our toolset we have over five and a half million creators using our tool set today we also have an internal team that's kind of working on episode originals that use that tool set as well we've launched over 50,000 stories um including both these great things created by our user community as well as things we created internally like Mean Girls Demi Lovato and Pretty Little Liars and all of which have seen over three billion episodes viewed which is like fifty-seven thousand years or so I'm spent watching it so I give this only as background just to say this is kind of I want to give this talk to talk about what we've learned is we've done all of this and I want to talk about how it works for us in three phases so the first phase is really creating a great story and because episodes on mobile I'm going to talk about creating a great story that works on mobile the second phase is then

you created that great story how do you iterate on it with your internal team to make sure that it goes from a great story idea and concept and outline to a fully fleshed out great story and then the third phase and kind of anyone here who kind of has built mobile games and launched an iterate of them knows this but how do you iterate on it with viewers in advance of its worldwide launch so let's dive in first to this first phase of a creating a great story that works on mobile the place you start this is for episode I think Freddy was really knowing and understanding your audience and so for episode our audiences women who are 13 to 25 on mobile and this audience is interesting this is truly a mobile only audience not even a mobile first audience so some of the stats we found when we talked to people 14 percent of our audience use their phone for 12 hours a day or more 11 percent check their phone when they're in the shower twelve percent especially those 13 to 17 check their phone every few seconds when we ask them how often do you check your device so this is very different than Hollywood

Hollywood does not design for this audience they design for an audience that can that goes into the movie theater that sits in front of a TV for long interrupted time not an audience where they're native things built for the mobile device it's much closer to gaming they can another part of this audience and this is great for all of us gamers is they want to be active participants that's what they've grown up doing that's what they've always known they've grown up with social media they've grown up with smartphones they've grown up beginning to be the center of their stories but for them these stories are like what are they for lunch what are their friends doing the stories aren't the great sort of Hollywood caliber stories so the question is how do we take those great stories of great cover stories but weave in thoughtfully the interactive elements that are so natural to gamers how do we make sure that the audience is at the center of that story how do we give them agency as they read the story and and how do we design it then to fit inside

mobile and how we're doing this so the last thing about mobile is episodes audience but Mobile out in general wants bite-sized stories so there are two parts of this so one part is for all of us have mobile phones our mobile phone is something we use when we're in line at Starbucks and we're on the bus when we're kind of walking when we have these small bite-sized increments and so I think that's true in general but then episodes audience there's a second layer to this if it's an audience if you guys all use snapchat or you see if anyone's seen someone who's 13 to 17 use snapchat in the piece at which they do it and the attention span that's associated it is even shorter 10 seconds is way too long a three to seven second video is the right length of time for short form content so thinking of these two factors both mobile to begin with and then this mobile only audience together we said the right for long-form storytelling it's no longer the 129 minute feature film it's not even the 22 minutes sitcom or 44 minute drama it's a five minute episode length is the right long-form

story length so if this is a giant move away from what is normal in Hollywood in typical formats it's more like kind of session times normal to mobile games but instead of it being a single compact size adjust how do you tell a long-form story episodically still by doing this bite-sized way that's kind of the audience overview so now the question is how do you pick a concept this audience will love well here's a place where Hollywood's pretty great at picking concepts they have a lot of things that they do they've taken up a genre the audience loves and do a really original take on it so you can think of like what Amy Schumer's done recently to the romance comedy and their find exceptional creators exceptional writers and have them make something so James Cameron Steven Spielberg standout in this dimension but there are lots of others where you really trust in a creator and they also can license a great story that was started in another medium so Harry Potter and Twilight both started at his books before they came to Hollywood so these are all great things

to start but they're ways that you can if you have kind of a platform on mobile you can augment this in ways that Hollywood can't really or can't very easily so one thing we did when we were deciding which partners in this licensing group who to partner with is we just asked our audience we're like what do you like what TV shows do you like what movies do you like what stars do you like and they answered they told us pretty little liars they told us Mean Girls they told us about Demi Lovato among a set of others so we went out and partnered with them and it's been great we've had a really wonderful response we didn't have to guess we just had to ask I think a second thing is because we have people searching throughout 50,000 stories we can look at what they're searching for and it's maybe not that unexpected but at the same time it can really zero you in and help you focus and so you know white this audience loved Twilight a decade ago doesn't it make sense vampires and werewolves were high up on here today still there's love on here whether summer or otherwise their crushes there's even cheating there's a lot about babies and pregnancy to talk

about princess princesses here in royalty did that inform some story ideas and so one story that came out of this for us was one we called the royal baby the Royal baby it's a story where you meet a guy you end up with him you end up getting pregnant turns out he's a prince it's amazing and with this it also this was like because we had all this data but then we could also look at what was happening in the news and you saw Prince William and Kate they just had their first two kids it was really on people's mind so we can take sort of again take that Hollywood approach take those normal things but add this data element to it to really feel like we're honing in and I'll talk more about it later but the world maybe was one of our best performing stories to date that we've created that was entirely original so you get this great concept then the next step to creating a story is really use those best practices of Hollywood screenwriting this you know if you've read save the cat if you've read story by Robert McKee these books really lay that out they go through kind of great three acts story structure they go through kind of pacing dialogue like

all the good things of storytelling because there's a true craft just doing storytelling well and I think that applies independent of the medium the other piece that we've taken a lot for episode because it's episodic is from the TV world lady of a writers room so the idea of you have a set of writers and creative people in a room working to break down the story together and come out of it with an outline that captures every story be seen within that act structure and we found that as a really great thing but you can't just take the Hollywood screenwriting process apply it exactly because Mobile's different so how is it different so the first is piecing of your story on mobile so in the first five minutes of movie they do set up in movie to get through the first act it's like 15 to 20 minutes in books it's 60 to 70 pages if we're talking about a novel an episode the first episode we tend to get through the entire first act and if you don't do that it will feel too slow so let me tell you two stories so one and one of the stories we did was

Mean Girls sorority rush it was a follow-up to our Mean Girls senior year which is a follow-up to the Tina Fey movie Mean Girls so the first version of that the whole first episode was you arriving at school and anyone that seen the movie pitch perfect that's basically the first five minutes a pitch perfect you're arriving to school it's fun it's enjoyable of the pacing feels right for movie we did that it felt way too slow it just it was like you were bored as we were playing through this thing we'd created so we had to speed it up and we jammed kind of that into the you know the first couple taps and then went on into the inciting incident and the rest of the first act another example is we had a writer that we brought in and brought onto our team who spent a lot of time in the soap world and he the first day in this writers room outlined the first four episodes of a story and the next day came back in and the team they looked at it and are like that's episode one that is not the first four episode that is episode one so I think really making sure that that pacing is mobile mobile length is important the second thing is

thinking about the mobile device and what is the native format of mobile if you think of iMessage you think of snapchat you think of Instagram you think of your email it is portrait that is the native format of mobile I think that is not the native format of Hollywood it is a landscape but I think for mobile designing and that native format feels better it feels more natural and it's the one we have very much stuck with I think assuming you take that step for kind of interactive fiction you put on mobile you need to really think about how screen real-estate works it's really hard to have a ton of characters on screen at once in a portrait mode it's better for close-up shots fewer characters so if I think of like if we're gonna tell Lord of the Rings on episode we would tell freddo's story more than we do the 10,000 person battle I think the last piece that's really different from Hollywood is you have to build interactivity in from day one so in the Hollywood world you can you create a linear story there's no way for the audience to interact with the movie screen and a theater there with TV shows maybe that will come soon but it's not

there yet and you know VR it will certainly come but again yet so I think this is a change to the process and it's thinking as you're breaking that stories you're in the writers room as you're putting together that story outline what are they going to be these moments of great interactivity these moments of choice these moments of branching so we found that we want to think about it then but the other thing we found we have to do is add a couple days onto our writers rooms to have specifically focused on choice and interactivity and to really flesh that out to make sure that we're building that into the core of what the story is instead of Dan dating interactivity on after the fact um it's an example when we work on Pretty Little Liars we have this idea there's a moment in the story where you're at a party there's a photobooth at the party a photo pops out and you have this really tough decision are you gonna protect the four Liars you're gonna protect Spencer and Hanna and Aria or you're gonna protect yourself and your family and we knew this would be one of those weighty wonderful decisions where it's a true choice that has agency one where 50% she's won 50% she's the other but we

designed it from the beginning and then could write the whole episode with that in mind so after you finish this process hopefully what you come out of is a great story and by that I mean you have an outline of a story with interactivity built in that's thought of for the pacing of the device now to iterate and turn to the first thing to do is write scripts but I think the key is go as fast as you can to get that script onto device so much changes I think and especially until you have lots and lots of cycles writing the script for device which we certainly don't yet it episode two we and we learned this the hard way by making some mistakes so we had one story we did internally called bad boys girl it was where we spent six weeks on the script we got it really right we kept tuning the language tuning the dialogue making the romance feel great making the plot move forward and then we put on a device and it just didn't work people were bored it was not fun to tap through and had we done that sooner so we had to go rewrite and blow up the whole story and work over it I think a second one that's having our first season of Mean Girls Mean Girls senior year so without we partnered with Kiwi

Smith's she's awesome she co-wrote Legally Blonde and 10 Things I Hate About You she's like great at writing Romans comedies and she knows how movie scripts translate to movie screens so we got that script it was hilarious we were laughing out loud we thought it was so funny it was like we thought maybe were funnier than Tina Fey then we put it on the app and everyone was so depressed because it was not funny anymore the comic timing just didn't translate and the way people expected so had we spent less time perfecting everything in the script form getting it to device sooner and making it a playable we would have learned that and I think key to being able to do that quickly is having tools to get to a playable fast so we built a tool set you know we released them anyone here can use them if you like um but some of the key things to that tool set is having a library of art that's there that you can just grab so you don't have to wait on all of the custom art for your story you can put art from your library in there I think the second thing is the tool set should really be usable by anyone or essentially anyone it shouldn't require animators Plus artists plus engineers to

get your story to be playable then once you've this playable story the question is how do you do story feedback so we think of it in four phases first of all there's a core group and this core group needs to be able to cover the key elements of the story while understanding the genre you're in and so things aren't that like for us story structure is a key thing and the fundamentals of kind of how the outline works in that romance for a lot of our stories is a key elements are really being honed to that and how to make a romance feel real and great and wonderful interactivity and choice people in narrative design backgrounds are great at that voice and dialogue and kind of getting that right and connecting with your audience and as well as visual storytelling and really understand not just as how are the words doing it telling the story but how are all the non word elements doing of telling that story so we get feedback from that group but then they've seen the thing a bunch so we want a few people to really understand story there outside of that group that can be fresh eyes then going to a broader group often including people that kind of understand

metrics better in tuning and those things that can give very specific feedback on the story but maybe aren't people with creative story backgrounds and then finally we'd love to show it to the whole team and we can't at this point for us we can't do that with all of our stories we like to pick a few and show and we get amazing feedback even from people who aren't from the demographic who don't love the genre but they still can be really thoughtful and the questions to ask as you're going through this is most importantly on mobile it's does it keep you engaged at every tap does it keep you engaged is the story moving fast enough does it have the right amount of agency that it gives to the readers is the dialogue working and fundamentally is it fun and is it connecting with you emotionally so if in there we do the third phase we want to then go and iterate with our viewers and for this this is very much what we've always done is a mobile games company a pocket gems is really think about how can we soft launch something how can we collect data and user feedback how can we then go iterate on the story and continue repeating this

cycle until we believe in the thing that we're launching and then we can get it to worldwide launch so how do we iterate so the first thing we look at is the overall rating so for anyone that's put games out on mobile this is equivalent Google Play or iTunes rating that you get from your users it's a pretty blunt thing it tells you like do people like it or not how much do they like it and eventually you can start having relative comparison points this is really good to say you know in my on track to have a hit is my concept to hit is the overall story a hit is my execution in general going in the right direction and it can help you decide which things to invest in overall but it can't really tell you what to do to iterate so that's the next set of things so a first is story starts and so this is combined with things you can control and things you can so a set of things you can control you can control the title you can control the description you can control any visual images we call them story cards inside episode and and so for that we've learned some things so we've learned that being closer on faces is really

good like you see on the troublemaker image on the right because it can show more emotion and capture that and we've learned for our audience like you see in bad boys girl kind of love love triangles or really exciting and ones with two guys tend to do better than ones with two girls turns out um but you can tune all these you can tune the description you can tune the title and make it really exciting there's an you also have to understand where are you showing it in other words like an episode if we put it is the very top card it'll get far more initial starts and far more people coming in if it's a licensed property that people already know really well it'll get more starts so I think that there are things there you have to understand to calibrate but then the final thing you can't control is is the thing solutely a hit or not and I think that that in the end will determine how many stories ties you get but these are all things you can tune to help give you a better chance then it's what do we iterate on inside the story so we look at what is the retention duty episode and we can also instrument Singh's story beats so you can go all the way down and

see where people kind of playing where they not playing where they falling off what we found is this is really useful interesting data but it doesn't tell you what to do it only tells you where there's a problem so in that first season of Mean Girls Freddie rush when we were working with Kiwi we put it out and we thought because in episode 5 you're competing for any of you who remember Mean Girls you're competing in this version with Regina George to win student body president and the election the results come in in Episode five so we thought that'd pull people along turned out no not enough not enough happened but we never said one in five to get people there we had to have something really fun happen in Episode three and key we then had this great idea she's like what if there were a party and what if Glenn Cocco threw that party and what if you had all these like highs and lows of a high school party and all the fun things and it was great it uh Turley fixed it but unless we had this data we wouldn't have known to go figure out a way to solve that retention issue in episode 3 in this connect this is just like if you think of kind of

retention in games but for us we broke it down to how do we do it within the story itself the next thing is choices and so for the choices themselves to make sure the player fields agency to make sure the choice is great the key thing about them is you want people choosing both sides equally if 90% of people pick one side of the choice and 10% the other that wasn't a meaningful hard choice it felt like there was a right answer versus this was a choice where you got to shape this story and where you got to define yourself as a character and that's another thing that you can track and look at we also look at dollars earned per story start is a way to think about monetization and in large part because we want to make sure that we can kind of know the stories that people love and decide where to invest more over time and there's a pretty big range and what we find here and we find that we can actually move this quite a bit and the primary way that we move this is by adding premium branches and choices so when I talk about those and we found that we can do strong work to

optimize them so here I don't know if you guys can read this that well and but what we started with was one where kind of you're saying hey I want to put on a new outfit you're gonna go meet a guy that you're into but you're like I want to put on a new outfit we give you a single choice for the outfit then you put it you try it on or not and you're like okay now I'm gonna go meet the guy it was a good choice it did alright we found with working with the the set up the choice and the result we were able to increase at almost a hundred percent and conversion and so with the set up we found that we really wanted to create good like stronger self encouragement to say hey if I put on this outfit he's going to love it I think then for the choice itself we found if we gave people two options they liked that better I mean we simplified the language around what those options felt like and then in the result we found that if we really emphasize the payoff of making that choice and help people feel really good about it and encourage them to make future choices like this I think it

separately in Pretty Little Liars with this moment where we've thought at the beginning it was gonna have great choices if you could help the four liars in their lives and you would pay to help Aria you'd pay to help Spencer you'd pay to help Hanna and what we found is that didn't work it turned out that the choices that work better were things that about yourself in your own life and kind of whether it was the romance you were interested or other things that tended to work better within that world and we had were able to really move up the engagement with those choices so this is kind of we stay in this process until we get to something great then we push it out to worldwide launch and that moment of launch is the moment where you learn did you you actually have a hit on your hands or not you don't really know for sure upfront but there's so much you can do to give yourself a better chance along the way but for us that's not the end so the step what we've been taking now and kind of we're just in the process of is thinking about how can we support agency and identity across stories not just within the individual story so because in a lot of our stories

you get to pick an avatar create a character but it's locked inside the world of that story so we're launching profiles where you can create a profile you can have an avatar for that and then the ideas over time for that avatar to be able to be the pro the avatar you use three different indistinct stories so that it better sets you up to feel feel this identity that persists across episode and across the world of that and really takes that to a next level so kind of summarize I think everything starts with understanding your audience and picking a concept of that audience is going to love I think then you need to embrace your mediums pacing its narrative design formats its interactivity formats well at the same time taking those best practices from Hollywood screenwriting and keeping those in mind you then want to go from a script to a build on your device as soon as possible because it's really the translation can is can be quite tricky

and unexpected from there you want to establish the right internal review loops because you want to feel good about what you put out but then you want to pile it to gather data and feedback and honestly we've never felt like we put something out too soon to pilot we've often felt like we waited too long spend too much time internally but I think the sooner you put it out the better and then you want to iterate and in this iteration don't be afraid to blow up your story I think that's the biggest thing nothing about that story should feel sacred it should feel like all you're trying to do is make something really great and wonderful that people love and if it's not working change it try again you it will get there eventually so thanks very much I'd love to take any questions if anyone has any [Applause]

hi thanks for the talk I just had a question more about how you come up just by making games and episodes games how how the studio was born episode sure so in terms of how episode was born we when we started pocket gems was always with the thought of we're mobile entertainment company not purely a games company and in entertainment story tellings the half trillion dollar a year market games is you know the tiny market of only 150 billion so we've always thought of what is need of storytelling on mobile but by storytelling we met this long-form Hollywood storytelling so we prototype for a couple years we launched a first version in 2014 with kind of a pretty small team and it just it worked and it fit so we've been investing more and more in it ever since okay thanks sure I have fantastic talk like a quick question in terms of monetization your stories are five minutes how many

of those episodes would would have player be able to do for a dollar and how did you guys go about that process of figuring out the sweet spot for how many passes or the stories a person can do for when they when they spend sure so thinking about what's the kind of sweet spot for story so for us the and our suits are five minutes our stories might last multiple episodes the same way TV seasons do so they're usually at minimum 12 episodes and that can go on we have some of our user stories that are now over a hundred episodes and in terms of that we don't actually charge for the episodes and we tart have kind of in-app purchases within them and then we also have a pass system that kind of the longer you've been in the app the more free passes you get every four hours but they continue to refill in terms of figuring out the sweet spot a lot of a/b testing I guess is the answer for the past system and but then also for the with the story length itself it depends partially on the story it I don't think

we know exactly what the optimal length is we haven't sort of so clearly fit it the way TV has it's like twenty two and forty four minute lengths great thank you sure hello very nice talk um I wanted to ask you how do you choose which choices should be the ones that are monetise able to with choices to monetize and a second question what once you made that decision did you a B test those choices so could you say the second part again once you make the decision okay I want to monetize this choice did you do an a/b test on those choices sure um it's not there's not a science to it exactly but we look at choices or branches that feel really meaningful and that will give some viewers something that feels like really meaningful value because if they pay for something we want them to feel great that they've got something of real value so either the branch has to be big enough or it might give them an outfit they wouldn't have otherwise got so that's kind of the the main criterion is

our viewer gonna feel good about having made that choice and done it and then yeah we test we test each one and we can kind of look at the performance of the choice okay thank you sir how much ongoing development are you investing in like the core of an episode right you have dialogue choice mechanics like items and outfits etc like all the animation stuff like are you still actively developing and adding to that toolset for the authors or is it does it feel kind of like done right now definitely not done so our tool we're investing very much in the toolset and want to continue putting on you in great ways for people to tell stories and surprise and delight their viewers is there any like sort of preview of what feel like that you could I don't I'm just interested in what feels missing or like what's the next biggest bang for your development buck sure I think a couple of things that are coming right now so we launched Pretty Little Liars in December with an entirely new art and animation style I think is more realistic more subtle we're working to put that out to our whole community

where we can put weather effects out right now we have a kind of first version of that so there's are a couple of things on the roadmap cool thanks I really enjoyed the presentation I was wondering if you could comment on your Studios experience mixing minigames with interactive fiction sure so experience mixing minigames interactive fiction um we haven't really done that that much as the answer so we have choices in there but we very much tried to say how can episode be an alternative to TV for a generation whose mobile only and used to be an active participant in their entertainment more than we've thought how can we mix interactive storytelling with kind of games in that sense so philosophically we've just taken a slightly different path Thanks a great content and super impressive concept just overall well done how did

you guys determine your target audience to begin with sure no I think it's great question how we determine the target audience and it's just to some degree it was thoughtful in some degree its chance and I think the chance piece came of we in our casual studio and the early days i pocket hams had created a game campus life and it had a great art set and had our paper doll engine that allowed kind of animated interactive characters so you could change outfits on so we could just take that art set and it was a in in the pilot episode and that focused on this audience I think some of them were the more thoughtful parts of it this is an audience that feels pretty underserved and and it's an audience that seems like storytelling quite a bit um and so forth and it's an audience where the kind of stories they like are tell about in a mobile format quite well versus say like a superhero story they'd be much harder to tell in a platform based way do you see any opportunities for extending the platform to boys I think we do um over time for sure we'd love to be a platform that hits all four

quadrants and but kind of right now we're focused on just putting out great stories for the current audience and growing organically all right thanks hi there I'm one person aside I'm curious about incidences of long-term choice and how your audience feels about them it seems that since they like things in a bite-sized form do they get frustrated when there's a longer-term choice where maybe they decided okay that's a choice I made a long time ago maybe I shouldn't have made I think that's an interesting question what do the professor with long-term choice I think it very much depend it means they wanted every moment be emotionally connected to that story so if the long-term choice is helping them stay connected to the story and feeling good I think that's what matters more than whether the choice is short-term a long term I think the other piece is all the best choices are the ones that feel really really meaningful and again it matters less that they're

sort of long-term it's just are you connecting deeply and emotionally with your audience at each moment excellent thank you that was fantastic well thanks everybody really great to talk to you you