GREG HARTRELL: Thanks toeverybody for attending. Today I want to spendsome time reflecting on what we’re seeing in thegames ecosystem in general and observations around whatplayers are actually doing, talk about thearchetypes that we see and how your gamedesign can cater to those archetypesand a lot of the tools that we have that allowyou to connect with them. So before I preppedfor this talk, I spent some time actuallywatching people play games. Turns out great opportunitiesto work with many of you, but also to get to seepeople play all the fun stuff that you make. And some observationsare this– mobile games, by and large are still, onaverage, a very single player experience. This is a portrait of achild sitting on a couch staring into a screen, verysingle player lonely type of experience. And whatever the little girlor the little boy is playing, I promise you they won’tremember a year from now what they were playing. And that’s a weird kindof phenomenon for me because when I reflect on thegames of say the ’80s and ’90s, I remember playing with friends. I remember playingwith families. I remember running intorandom people at arcades and enjoying a verysocial experience. In the console biz, we createdthis kind of gamer archetype, right? It’s the guy sittingin his basement staring at a screen witha controller in his hand. And we put a headseton this gamer persona and we call that social. But if you’ve played onthose types of experiences, that experience usually involvessome 13-year-old screaming something about your mother. And that’s not necessarilythe best experience. We did a little better. And then there was brieflythis phase in gaming that we called social gaming. We were all very excitedabout the possibilities of bringing people together. I don’t know how youfeel, but I never had the day where sendinga putty knife to my friends was a fun experience. And it just turns that we seeless of those types of games today as a result. So the reality is that gamesare better than this, right? Ignore video games for a moment. We know that gameshave the capacity to bring peopletogether in a way that other types of mediums andother types of entertainment are not capable of. And there’s evidence aroundus when you just take a look. I quote Greek philosophersto get immediate street cred, you should too. Plato uses this quotewhere, you can discover more about a person in an hourof playing a game with them then you can in ayear of conversation.
then you can in ayear of conversation. It’s a profound statementthere, that perhaps unlike any otherkind of medium, you can connect withpeople in a deeper way with games then you can with anyother action you can think of. A more academic version ofthis is by a Dutch historian, a guy named Johan Huizinga, ifI can pronounce that correctly. And in this quote– his thesisof his seminal book called, “Homo Ludens,” which if you’rea Latin speaker, that loosely translates into “man at play.” And the thesis of the book isthat playing in its own right helps develop usin ways that are more fundamental tothe person that we are. It shapes our personalities,it shapes the relationships that we create. It even shapes the way thatcommunities come together. And even at a meta level theway nation states kind of work with each other. So all play, becausewe’re attracted to playing with each other,has meaning in that sense. So it’s a very profound kindof way of looking at it. And what those quote arereally supposed to demonstrate is that playing gamestogether are fundamentally a human trait. We build friendshipsaround them. We express ourselvesthrough them. We build a sense ofbelonging to a community when we engage in thosetypes of experiences. And if I again use the physicalworld as a barometer for this, there’s some obviousexamples here. I investigated retiring. It’s actually fantastic. I think I’m going toskip the whole work thing and skip to retiring. They’re constantlyplaying games, right? Card games and boardgames and anything that they can gettheir hands on. They don’t just merelydo it to pass away time. The underlying tone here is it’sreally to help groups of people meet each other. And you would otherwise notknow who this other person is in this retirement community,but you build a relationship through games. Games also have the capacityto get large groups of people to do really inane things. This is a picture of– so inthe Northeast United States, there’s Wall Street inNew York and it turns out there’s a neighborhoodof people there. And every year atChristmas, they get together in do thisannual freeze tag event. And so if you know what theNortheast is like in the United States– there it getsactually pretty cold. But somehow they gethundreds of people out in the community to comeand play freeze tag events because it’s compelling to comeand play with a lot of people. This is also a good exampleof a physical world. At MIT, they have thisannual mystery hunt they’ve been runningsince the 1980s. And I think in thelast version, they
And I think in thelast version, they had about 2,000participants with teams ranging from 50 to 150 people. And the goal ofthe mystery hunt is you solve all these veryarchaic or arcane puzzles, I should say. And you eventuallysolve all these riddles to learn where some hiddencoin is around the campus. And the reward for doing allof these crazy puzzles that involve mental and maybeeven real gymnastics is that you get to write thepuzzles for next year’s event. And you see this graphhere where it just shows the intensity of the teamsthat are participating in this solving puzzlesover a time series. And it captures the essence oflarge groups of people coming together to experience areally fantastic entertainment experience. And so the thesis isbasically, at their best, games bring us together. And when I look backin video game history there’s really great examples ofwhere we did this really well. There’s the arcades ofold that brought together random groups of people to play. The more recent incarnationis the bar-cade. Some genius decided that they’lltake their love of arcades and their love ofalcohol and combine them into one establishment. And so I’m glad to see thatkind of take a resurgence. But living roommultiplayer games– we remember games like”Goldeneye” and a lot of the Nintendo products of old. And of course, MMORPGs,which in my mind are really justreally elaborate chat apps that have a game on topof it while you’re playing it, right? And of them have thesesimilar types of properties. They bring peopletogether and it keeps people engagedin very deep way. So if we accept that mobilegames are largely still single player experiencesand playing games is human and we’ve been better atthis before, the question is, can we change? Can we be more effective? And what my thesisis yes, we can. And here’s why. You heard this stat fromMichael this morning where we have a billionactive Android users. This is 30-dayactives, not totals. That’s a massive activecommunity of people. But use this stat asyour next measure here. Three in four of those users areplaying games from Google Play. That might be the largestgroup of people playing games ever put together. And it of course makesGoogle Play the best place to grow and distribute your gamein the mobile ecosystem today. But what these twostats really tell you is that really,everybody’s a gamer now, right? It’s not a demographic in a box. It’s not just 18 to35-year-old males.
play pc games on android It’s not just 18 to35-year-old males.
It’s not just 18 to35-year-old males. Anybody can play a game becausethe device in your pocket is capable of entertainment–such rich entertainment. And we have these cloudservices that can actually connect people togetherin a very deep way. So the call to actionto me is really clear. Let’s make ourgames social again. Now here’s the challenge. If you have abillion active users, how do you succeed inan ecosystem this large? When you only had afew million users, it became a lot easier to thinkabout, well, what kind of game am I making? How am I going toconnect with them? When you get intothe billions, it becomes a much harder challenge. And so this is the reason whywe made Google Play Games. This is Google’s game networkfor Android iOS on the web. And we announced recentlythat this game network added 100 million users injust the last six months alone. And that velocity, we’revery confident in saying that it’s the fastestgrowing mobile game network, perhaps ever. And what’s fascinatingabout Play games is that it fills thisreally unique need when you have anecosystem this large. And that is– you’re creating aconcentrated network of people who love playing games. And as a result you want tobe able to connect into that and get your contentexposed to them. So since we launched itat Google I/O in 2013, developers have been integratingservices like achievements, leaderboards, multiplayergifts, and newer services like saved gamesand quests to be able to enhance the retentionand engagement of their games. And we know that games aredoing better with those services today. More tangibly though,Play games gets you access to very highquality gamers, people who love playing games. And that’s the performance edgeyou need it in an ecosystem like we all see today. We know when we look atdata for the developers who have integrated gameservices– and I’ll take this moment to say thatif you have integrated them, thank you, because youare the reason why we’ve been successful andwe love everything that you’ve donewith game services and we want to see you do more. But tangibly, wewant to make sure that the services areactually providing value. It’s not just somethingyou throw into your game because Google told you to. We see these players, they playlonger, they play more often, and they monetizemore frequently. And that is a reasonalone to make sure that you get your gameconnected to those users. And being able to see thedouble digit percentage increases we’ve seen in sessionlengths and duration played.
increases we’ve seen in sessionlengths and duration played. And in some cases,revenue bumps depending on the types of designsthat you’ve chosen. So this is where the question–it’s not merely a question of did you integrate Play Games? We all know that the gamethat your designed matters and the thematics and howyou approach your users still works. But how do you engagea billion people? So we’ve done numberof studies here and one way to think aboutthis is that in this ecosystem there’s an opennessfactor for how people want to engage withsocial interactions. So there’s people who wantto do very high levels of social interactionand very low. And so we have thesethree personas, right. So I’ll call them thecompetitor, the completionist, and the stealth gamer. And you can plotthese personas, kind of on that graph ofsocial openness, right? With your competitorbeing the most open to being directlycompeting with other people, to your completionistwho’s actually not very interested incompeting with others, but is more interested insingle player type interactions. So I’ll talk about thesethree personas in more depth. So the competitor is aninteresting kind of persona because they’ll engage in almosteverything in your game, right. It’s achievements, it’sleaderboards, it’s multiplayer. This is the person–they own every console. They set up theirsocial media page to brag about theachievements that they earn from any platformthey earn it on. And they’re really engaged andthrive in direct competition. On the oppositeside of the spectrum we have thesecompletionists, right. These are peoplewho don’t really like engaging in anykind of social activity. However, they thrive inprogressing through your game. They want to beimmersed in it, they want to accomplish all thegoals that you give them. They want to see the storyline. They want to finish everything. This is like themobile equivalent of the gamer that wanted toearn the “Geometry Wars” million point achievementwithout ever dying. If you remember gameslike “Final Fantasy VIII,” there was that one summonthat you could do at the end and it was like atwo minute video that you had to watchjust to watch it cast. And you had to completeevery last corner and inch of the game tobe able to earn the right, or the braggingrights to do that. That’s this person, andin mobile, they exist. The last categoryhere, and they’re kind of in the middleof this social openness spectrum is what wecall the stealth gamer. This one’s a trickyone because they’re the person you ask, hey are youa gamer, and they’re like, nah, I’m not a gamer. But then when you lookat them on a normal day they’re playing games every day. This is like my mom, right. She’s like, no Idon’t play games. But she’s sitting thereon her tablet every day playing with random people. And so what we knowabout this persona is that they will engagein social interactions but they will do sopassively, right. The social interactionshave to be quick, they have to berelatively frictionless. Or you have to kindof involve them in kind of like a communitytype of atmosphere as opposed to makingit more head-on competition or somethingthat they would otherwise be frightened of. So with these three personas,we can think of the tool kits that we have withthe games network and how we can connectyou to each one of them. So I talked about thecompetitor persona. The toolkit you have here–an obvious one is leaderboards and we can talk aboutturn-based multiplayer and real-time multiplayer. So the nice thingabout our leaderboards is that we have socialand public leaderboards. So public leaderboards are kindof for your hardcore audience because there’s a set ofpeople who will always compete to be number one. For the average person, theymight never be number one. But I could be numberone amongst my friends. And so what’s greatabout your competitors, they’re going to intuitivelygo for the top scores but they’re going to see theirfriends leaderboard and say, hey, I want to competewith my friends. They’re more inclinedto invite them, it’s more likely to turninto an acquisition for you. And that’s a fantasticreason for leaderboards to be implementedinto your games. Encouraging the friends tobe invited into the game is just another way to notonly engage the persona, but to grow yourgame incrementally. Real-timemultiplayer– if you’re building a synchronous gamethis is a fantastic technology that gets you into the ecosystemin a very tangible way. “NBA Jam” here– ifyou remember that game from the ’90s, now on mobile. They implemented ourreal-time multiplayer. And what they usereally effectively is what we call ourauto-matching feature. And auto-matching– theway to think about it is that you’re gettingaccess to the hidden social graph of the game. It’s basicallyconnecting people who are playing actively right now. And so you see incredibleboosts in engagement because you’re able to simulatethe arcade interaction of old, but on a mobile medium. And so you’ll playa game like NBA Jam, you’re playing one or twosessions, you bail out and then seconds lateryou’re connecting with somebody else who’savailable to play right now. For a more, let’s sayintimate experience, you can still invitefriends directly in the game and this is kind of ascreen shot of doing that. Google will sort and rank usersdepending on your interaction with them recently. So recent playersand active players will show up at the top. And friends who haven’tyet come to play the game are deprioritizedin an effort to make the match experience relevant. For turn-basedmultiplayer, it largely follows a similar type of model. Turn-based multiplayeris much better for games where the sessions canbe played in increments. So maybe it’s overhours, or over days, or it’s literallya turn-style game. Same type of thing, except whena turn is pending in the games experiences that we havein the Play Games app and inside the game themselves,this is a call to action for me to come back. So if I went on vacationor I just got busy and I didn’t decide tofinish playing the game, having that card sittingin my inbox that says, hey Dan’s turn started remindsme that I was playing this game and is a call to actionfor me to come back and that’s an importantretention point for you. An important part ofturn-based and real-time, though, is the socialdiscovery part of it. And that is whensomebody invites me, I do get a notification acrossthe entire Android ecosystem. So the way that this worksis that we prioritize how noisy thenotification is, depending on whether I have a connectionwith them and their circles, or whether I don’tknow them at all. So here you can see Dan’sinvited me on the left here. And I see his face and I seehis name and it buzzes my phone and it vibrates, andgreat, Dan’s there, I’m going to go and play a game. But somebody can invite mewho’s not in my circles. And I’ll get a silentnotification in the shade and when I pull thatdown and I look at it, it’s like, OK, great, I’mgoing to go see who invited me and understand that Iwant to play with them. So when I dive into oneof those notifications, you’ll get an experience notdissimilar to this, which basically allows me tostart a game with somebody I know– Dan on the left. Or somebody I don’t know–this stranger on the right. Not really a stranger. And I can decide whether Iwant to play with them or not or mute them or declineor whatever have you. But here’s the growth hack. If I don’t have the game,this will redirect you to the Play Store. And then someone’s goingto install your game, start getting playing thatthey otherwise wouldn’t. And because it wasintroduced to me by a friend, they feel a lot more inclinedto start playing the game because there’s kindof a social angle to, well if Dan likes the game,I should like the game too. OK. So we talked about thecompetitor persona. There’s the otherside of the spectrum which is the completionist. There’s two really importantways to engage this user and we just see it time andtime again through data. Great achievement designand supports saved games. Great achievementdesign– the way I always describe it is if Ifinish a game like “Hit Men Go” is the game really done? And I could just turn itaway and forget about it. But it turns out that when Iadd achievements to my game it allows me to explain to usersthat there’s more depth to it. That there’s more to do. And so you can easilycome back and you can see, here’s some achievements thatI haven’t quite finished. They’re still very incremental. It gives me othertypes of objectives, it calls me back into the game. And we’ve seen games withgreat achievement design boost monetization. They’ve boosted engagement,they’ve retained their users in a deeper wayand consistently, compared to games that don’timplement achievements, they outperform themalmost every time. If you want tips ongood achievement design, we have a YouTubevideo called, Game On– Achievement Point Pointers. This is a good opportunity totake a picture of the screen. If you don’t take a picture ofthe screen, you can Google it. It’s a fantastic video byour resident, Todd Kerpelman, who is a former Pogo designer. I mentioned saved games. So this is kind of intuitive,but you have these solo players who increasingly, whenthey’re very engaged, are switching between devices. Our saved game systemworks cross-platform. And really, thepoint of doing this is that you never want somebodyto play level one again. If they’re going to switch adevice or buy a new device, the worst situation is you’llcompletely churn out your user if they can’t pick upwhere they left off. Or my other favorite example–I think we’ve all experienced this– is you get the supportcall if somebody says, hey I thought that I would fixthe game if I just uninstalled and reinstalled it, likethat works normally right? And then you realize,oh wait a second, you’ve lost all your data. You’ve basically resetthe game, there’s not very much I can do for you. That’s a bad experience. Saved games completelyremoves this. And we consistently see thisas one of the top requests from any one of ourplayers in Play Games. Easy to implement, braindead from that perspective. And here’s what we addedat I/O– the ability to attach cover images. So you’ll see here this isme playing “Leo’s Fortune.” What they do is theytake a screenshot of where I last leftoff in the game. So when I come back fromvacation and I scroll through and say, hey, what gamedo I want to go back to? I see the screenshot thatreminds me, yeah, that’s right, I did left leaveoff on level three. I’m going to comeback and I’m going to pick that gameover other ones because I feel like I couldgo back and probably win that level. So it’s kind of adigital bookmark so that users can getcalled back into action and help you get a little bitmore of a re-engagement flavor to your title. While we’re on thetopic of growth hacks, I want to insert this one. And this is the concept ofgames your friends are playing. This applies reallyto all the personas, but I’ll call out– becausethe completionist is still a play alone typeof player, they’re still attracted to this. There’s somethingthat’s about, hey, let me see what myfriends are playing. It’s a way for meto discover games. It’s a denotion of maybe thegames have a higher quality, or whatever have you. So you see a screen shot here. We have the Play Games appwhere you can see games that I’ve played recently, like”Brave Frontier” or “Frozen Front.” Or in the store whereyou can see games like “Heroes of Camelot”and “Sky Force.” This is a very strongcall to action. It’s zero effort to you, otherthan implementing Play Games. We take the signals,we pour them into our discoverychannels and help your name get better discovered. And so if you’reinto growth hacks, finding those extra percentagepoints of additional user acquisition, this is a veryeasy thing to do for your game. OK. And so that brings meto the stealth persona. So we talked abouthow the stealth persona is kindof tricky, right. They don’t like thedirect social engagement, so you can’t use like the oldmemes of direct competition. They don’t think ofthemselves as a gamer, but they play all the time. But we have a fewtools that we think will constantlyresonate with them. So the first one is that wecreated a game gift system. We know from our research thatif the social interactions are extremely discreet,very easy to get into, and low frictionGame Gifts is a way to kind of trade a quickobject between two players and keep them engaged andinvite them into the game. And this serviceis pretty simple. I can just kind ofvisualize it this way is that when youimplement it, players select a gift that theysend to another player. We store if for sevendays on our server. And we send a notificationin a similar type of way that we do for multiplayer. When the other user receives thenotification, they tap through, they select the giftthat they want to play. And similar to whatwe did in multiplayer. It will ask them to installthe game if they don’t already have it. So it’s a great opportunityfor another social acquisition. But it’s also something thatresonates with this persona because I’m notgetting challenged. I’m not being put on the spot tobe able to do this right away. It’s a very simple thing. Oh, I’m receivingsomething from somebody. That sounds kind of cool. Don’t send them puttyknives, by the way. I don’t want any putty knives. And then ultimately,they receive the gift, they’re elated, and theyget engaged into the game. And you can recreatethe experience inside of your own game like we dowith any one of our other APIs. Another service that welaunched is called Quests. And the way to thinkabout Quests first of all, it’s a way to create time-basedobjectives in your game without having toupdate the game. And what it does is it createsthis community feeling, right. There’s a lot of people doingthis big kind of community event on a weekend. Sometimes this isreferred to as live-off. So this might be thoughtof as live-offs in a box. But the way we’veimplemented this is that it’s analytics driven. So the way Quest works isthat you send the Quest system events of what’sgoing on in a game. When somebody levels up, whenthey modify their sword, when they find the rare blacksheep in your game. And through learning whatplayers are doing inside of the game, you cancraft quests then you run in a time span,like on a weekend, or for an entire week basedentirely off of that data. I’ll show you how this works. So you start bydefining a set of events that happened in your game. Let’s say I’mmaking a zombie game and I have red, green,and blue zombies. And so I would fire anevent every time somebody killed one of thosethree colors of zombies. And I integrated that in mygame and the Quest interfaces exists to show a user whena new quest is available. Then I monitor those events. So let’s say I’m prettysatisfied with the number of red zombies and green zombiesbeing killed in the game, but I’m not really happy withthe number of blue zombies being killed, because there’smore monetization events tied to them, they’reharder to kill, people have a tendency of buyingmore things in my game because they’re harderto kill in general. So I can go into thedeveloper console and create a Questthat’s like, hey, we’re going to have a 5,000 zombieblue zombie killing fest this weekend. And the Quest system will trackevery time somebody’s completed killing an existing blue zombie. And when youaccomplish that goal, it can push down a customblob of data to your game that you caninterpret and decide how you want toreward your users. So common examples wouldbe a blob that says, grant in-game currencyto the user who succeeds. Or give them somekind of in-game reward or progress them through thegame in a new type of way. The best partabout this, though, is I didn’t have to update mygame at any moment in time. It’s entirely based onjust sending the events and the questsautomatically populate in the Quest UIthat’s integrated. And so you could runthese quests every day, every week, every weekend andconstantly reengage your users. But use the data to decide whichquests are the most effective, and are you helping shape thebehavior of your users in a way that you’re retainingand engaging them more? We had one developer who showedtheir Quest data recently and they saw users that engagein quests went on to do, I think it was 160% moreother types of sessions in their game, just by virtueof engaging them in this way. So I highly advise–it’s a dynamite feature if you’re in to live-ops. Very easy to implementand everything is documented ondevelopers.google.com. The last thing thatI want to talk about is, of course– OK, we’vegot these three personas, we have all these tools. How do you knowyou’re successful? So you have to have away to measure success. So Google– we’re a verydata driven company. Game developers whowere wildly successful are also very data driven. One way that we helpyou here is– you’re familiar, probably with thestatistics we have in the Play console today. But there’s aseparate game focused set of statistics that aregiven to you just by virtue of integrating Play games. And so when youimplement Play Games, there’s basically a zero effortdashboard that’s given to you. I’ll dive into ita little deeper. So for example,your daily dashboard gives you a summaryof your active users, your new users, retention. And more recently we’veadded demographic data. The demographic data– whenwe hear from developers is essential because as a gamedesigner you want to know, hey, is this game actuallyresonating with the audience I thought that wasgoing to be playing it? If I made a game thatI expected females between the ages of18 and 30 to play, I would hope my datawould show that. Another angle here is that ifyou’re into user acquisition and you’re trying to comeup with the right type of creative assets, knowingwhat kind of demographic your game has, inreality, can help you decide what kindof creatives you use and how you engage and goafter the right audience. Another essential here isour retention dashboard. This is for your abilityto retain new users. So a common examplehere is that you want to take a look at how manynew users are coming back– day one, day two,day seven, day 30. And so you cankind of understand the attrition rate of your game. And let’s say that you hada tutorial in your game and you wanted to try andimprove that in an effort to help people get back on dayone and day two more often. Here you would see, youknow, on March 8 here, I did an update to mygame the day before and I saw a boost in my metricsbecause apparently my tutorial was better at convincingpeople to stick around. And that’s a usecase for this data. Again it comes from justintegrating Play Games and having people sign in. AUDIENCE: Do you havea formula for that? GREG HARTRELL: The formula? AUDIENCE: Yeah. [INAUDIBLE] GREG HARTRELL: Yeah,this one’s simple, right. It’s basically that youhave a cohort of new users that came in on a specific day. Here’s the percentagethat came back day one, day two, day three. It’s a simple multiplier. The last one is inthe engagement stats. Today we support achievementsand leaderboards. So when you think aboutengaging these personas you also want to use the datato know that you’re effectively doing that. So you can see whether theachievements that you added to your game recently areactually getting earned. Whether they’re toodifficult, whether it takes too long to earn them. Or if you’re into thecompetitor type of persona, maybe you have acustom leaderboard for some very specialpart of the game and this is a way for you to getan indicator of whether they’re actually engagedin that content. And whether the changes thatyou’re making to your game design are actually resonatingwith that type of user. All right. So I’ll wrap up. The summary today isbasically this– Google Play games connects your game toa highly concentrated network of very high quality gamers. We have the services,achievements, leaderboards, multiplayer, and gifts. And the new stuff likesaved games and quests. And they’re all showing usdata that games are increasing engagement and retentionin very dramatic ways. These players play longer,they play more often, they monetize more frequently. And the service can caterto those three personas if you’re very deliberateabout in your design. But going back to thebeginning of this– remember, everybody’s a gamer now. We have a billionof them out there. And that makes this a muchmore challenging, but also a very exciting time for us. Android and Google Playhave created an ecosystem that large that gives youthat kind of opportunity. And we know that games at theirbest have brought us together. And so if I leave youwith one message today it’s, go forth andmake your game social. And with that– if youhave any questions for me, I’ll be wandering outside, happyto talk with every one of you.