What Successful Product Managers Do Differently in Free-to-Play Games

[Music] what successful product managers do differently in free-to-play games well this was actually the question that i asked myself some six to seven months ago when i was offered a role as a head of product management at rovio and the thing is even though i had the experience so i worked as a product manager first at digital chocolate then at supercell in rovio and then more of a game leadership roles as an executive producer at zynga and then studio lead at fun plus so even though i had the experience of working as a product manager hiring product managers and managing product managers i never really you know sat down and conceptualized what makes a good successful product manager and most importantly what are those successful product managers are doing differently so um i approach this in the uh

the typical core loop what i mean by that is in order to to really find out what the successful product managers are doing differently i first talked to my friends my colleagues and people i know from various amazing companies like facebook zynga wargaming natural motion fun plus ea you name it and i ask them two questions firstly is what makes a great product manager and the second question was more detailed and that was can you can you tell me about a situation where product manager was succeeding when they did more than you expected based on these questions and based on the multiple answers that i got from my friends and colleagues i sort of deconstructed the uh the information and conceptualized it into different themes and with those themes i got back to my colleagues and friends and got more feedback again deconstructed our feedback and so forth and so forth until this played so

as a result i got five elements that successful product managers do differently in free-to-play games but before we jump in into those five elements let's talk about what do product managers actually do and the first thing that you might think of is they animate a lot their presentations and you're absolutely right but in addition to that what product managers do is is actually a tasks of both highly creative elements like defining designing and managing key features discovering creating growth levers but also very many analytical tasks such as building quantitative models owning and analyzing game metrics and overall just managing the business performance but the role is actually even more complicated because it changes not only based on these tasks but it also changes based on the

life cycle of the product whether in pre-production production soft launch or live operations as well as the culture and the organization of the company so the way i see it and and the way i like to present it is i would say product managers have two key goals the first goal is owning the business side of the game and what i mean by that is that in in the internal battle of cpi versus ltv the product managers definitely represent the ltv side and they focus on the metrics of of growth of retention and the monetization and their design thinking and their their overall goal thinking starts with those metrics first and they make their decision based on that the second key goal that i feel product managers have is helping the leaderships the helping the leadership make decisions and why i have um what i have an image of a hitman is okay product managers don't really have the um

we don't have the ability always to pull the trigger but because we do the analysis because we do the uh the due diligence and because we do a lot of the designs we essentially have the the power to adjust the aim and that's why product manager's goal is to support the decision making all right let's go to the uh the details the first thing that i feel successful product managers do differently in free-to-play games is they use datum to find the low-hanging fruits and what i mean by low-hanging fruit is is something that requires minimum amount of engineering work a minimum amount of work actually from from art perhaps but has a significant impact let's start by talking about what isn't a lot low hanging fruit so the normal process of product management of future updates it goes through basically this cycle you start off with the analysis you end up in a hypothesis or design then you prioritize that design based on the cost of the of

the work and based on the impact and the next step is execution of of that design after which it's further analyzed and then more hypothesis arise pretty pretty um standard approach and to give a more of a clear example of this type of approach would be let's say um league update in clash of clans so as you probably are well well aware of this game and have played it before the way the leagues function and the way the update would have come is the analysis shows that players were abusing the matchmaking system by dropping their trophies and that sort of created a double whammy in the sense that those players abusing the matchmaking systems were able to be matched against weaker players they were able to rate them quite easily and thus get resources and progress faster through the content on the other hand the players that were raided by much stronger opponents were were you know feeling a little bit discouraged and and that would probably affect their retention quite significantly based on this analysis that that derives

both from quantitative metrics as well as qualitative and the hypothesis or the design is made to reward players to stay at certain trophy level and most importantly push them to acquire more trophies and thus be matched against even harder opponents because this design has a big effect and it solves a big question in the game it gets prioritized high it gets executed and then with further analysis the numbers are being tuned and the uh the leak features getting improved and improved with every upgrade so this is not a low hanging fruit this is what you're expected to do so let's talk about a low-hanging fruit a game called dragons of atlantis i don't know how many of you are familiar with this but just to remind you this was a kabam game and it could be described as game of war um minus key social features plus dragons okay at the time was fairly successful and the the product team on kabam were

you know were having a good time but the game had some some problems and the key problem was that they were experiencing a churn of players between citadel levels so the way it worked is you upgrade your citadel or your city hall and that unlocks several more upgrades both for your troops your research your buildings and then you grind through those and then you can upgrade yourself level again and so forth and so forth the problem was that they didn't have as much levels for those citadels and because of that the citadel levels were quite far away from each other and they're getting further and further away and and players were honestly just turning out the metrics were going down as they were progressing through the games and then players just felt it was too big of a grind to go to the next level based on the normal approach of analyzing hypothesizing prioritizing and executing the product team would have come up with a suggestion of doubling the content of the game implementing that and so forth and so forth but that's a very slow approach so instead the

the product team on the game did something much more simple and more effective they simply doubled the amount of citadel levels without it really touching the amount of level ups there are so instead of unlocking ten new things when leveling up to level six of citadel you just unlocked five and then on the seventh level you unlock the rest of the five as a result with this small update they were able to increase all the metrics from retention to engagement to monetization let's go through another example of an easy low-hanging fruit and it's a game called angry birds match and if you haven't played this game i would describe it as candy crush saga with angry birds and events and events is actually the important part of this game so uh let's talk about yeah sorry so events are important part of this game and there's two type of events there's the seasonal events that last for two weeks or three weeks they're pretty

large and visually they're very different and then there's these small events that are happening between the bigger seasonal events during their latest not latest but a couple a couple events ago when they were running the valentine's day event what they did differently is at the end of the event they were added a leaderboard so as the players went all through 70 60 80 different levels they instead of finishing the event they ended up in a leader board and that leaderboard was active for as long as the game was active the product manager in this game mateo spiri analyzed the uh the data and what he noticed that adding the leaderboard actually significantly increased the engagement because it kept those highly engaged players progressing through all-out event and instead of them finishing the event and waiting for the next one to start they actually stayed in that event and that brought a significant increase not only through engagement but also through through retention and monetization elements so based on this finding as the event ended we made a hypothesis and that was that

when running our smaller events those who last between 24 to 72 hours what if we made a little change and that change was that after you finish the event you can finish it again so you can get double the price and that's it it's just a small data tweak as a result the kpis or the baseline revenue pretty much increased 20 percent with a small data change so successful product managers are able to look at the data and make quite significant changes with without the without big use from uh from engineering team but even though that the small hanging fruits even though we're always looking for those that that's always not the case because sometimes you just can't find them or sometimes the target is just too far off and that's when when breaking down those those bigger goals that's when working on bigger features you're supposed to and you and the success ones are able to break it down to smaller targets so let me give an example of how not to

do it and this is my personal example so about this is an old game it's a facebook game called army attack from maybe eight years ago and uh i was running this as a lead product manager and i had the goal of increasing the baseline revenue by 20 and instead of you know taking the uh the smart approach of really breaking down where the revenue comes from uh maybe thinking about adding some some some core features increasingly gain increasing the retention i went with a much straightforward approach uh which was this [Music] as a result yes the the revenue increased and then the leadership was very happy but not for a very long time because what i ended up doing is i just changed player behavior player purchasing behavior and it just created more work and created cycles in in in our sales and overall just wasn't very smart so let's talk about a smart example of breaking down big goals a game called dawn of titans i'm sure

all of you played it and those of you who have played it probably know that it's a highly event driven game a pm on this game her name is kim kellerbaum um not anymore of him on this game but when she was she was given a similar goal than i was and that is increase the revenue increase the revenue of events by 20 and her approach was significantly smarter than mine so first she looked at the low-hanging fruit can we increase the uh the funnel for the event the dau of the event by 20 because that would i mean that would basically do it and the answer was no so what kim started doing is she started looking at the data and more importantly started looking at how the best players in the events are behaving and she found out that the best of the players were the most engaged that made the most sessions they were the most active they spent the most time in the game and most importantly they were very active in chats based on that she tweaked the event structure in a way that the next event was was really designed

to push players to act as those engaged players and those highly engaged players were pushed to act just in that way as the um or even even becoming more engaged next she um she built levers in order to push players to act in the way that that was designed and after that she created a custom dashboards together with data scientist team that would allow her to follow the event structure the event progress in real life time in order for her to pull the triggers at the right time as a result kim was able to hit her target to a t at the right time and just overall broke down a big goal into smaller actionable targets and did her due diligence so third thing turning numbers into a story and as i try to show in the slide before with the uh with the gun it's it's basically saying that we as a pm you have you have a power of of affecting the aim affecting the

decision making and as a game lead i mean this is often what you experience there's constantly new ideas coming in left and right and they're just dropping on you the quest raid bosses everything everything that you should be doing and this comes both from external and internal and you know you need you need to prioritize your roadmap and you need to make the right decisions and and sometimes it feels like the right decisions aren't made in the right time so the successful pms are able to turn those numbers turns those analysis more of a story and sell it through that as an example let's get back to dawn of titans before there were there were events for for kim to optimize and grow there were actually no events in the roadmap at all the two pms working on this on this team mr anil descoupta who's probably in this room as well as lawrence clark they created a hypothesis that by adding events to the game they would able to

do they would be able to increase the baseline they showed the examples from apani like games like puzzles and dragons in the east as well as greece games in the west such as crime city and world of war of nations i believe they were all increasing their baseline by running very good events constantly but that wasn't enough to convince the leadership that that events should be done so instead this happened it's a pop-up of a sale but it's actually not just a sale so what the guys did is they created um they created an event by themselves they tweaked the gotcha a little bit so that like the likelihood of getting a very rare titan was increased by tenfold and that event would run for about three days they added a pop-up to the game and they added also a message to the facebook community just informing players that this kind of event was going on as a result they were able to get some numbers for the story and the numbers were good because the game hit its highest

arp doubts during that time and it ended up in the game in the in the in the roadmap being prioritized with the events and and the game running through the events event cadence example number two uh clash of clans and japan so sometimes so why telling a story through the numbers is important because sometimes we just don't have enough data we think that something is really important we just don't have enough data to show it and and it's great to have a b tested grace to have sequential tests it's great to have the analysis but sometimes you see something else being done by others and you're trying to pull that data but it's just not as concrete and and that's when when telling a story is important and that's that's that was the kind of the case of the launch of clash of clans in japan so before the game was launched or relaunched in in japan the kpis were very good for for the audience but it wasn't localized or culturalized

and the normal approach when you're seeing good kpis is basically you contact your localization partner you set up the order for for that localization it comes in there's a qa and then goes to the next update and hopefully it increases your number a little bit but the story of japan was a little bit different at that time puzzles and dragons alone was making more money than clash of clans in japan than collateral clans it sorry at that time the puzzle and dragons were making more money in japan alone than clash of clans throughout the world so that meant that there was a big market the second part of the story was that successful like highly successful western games such as farmville and candy crush saga were not that big of a success in japan despite going through the normal localization route just by doing the marketing despite kind of doing their due diligence so the team looked for for uh for a way to be big in japan a plan was created and that plan would take actually three to four times longer to implement than to simply pushing out um

a new localization update yet the plan the story behind it was that that it really put the game into a position where it could be successful in the japanese market and thus the leadership decided to to go through this slow more riskier route and as a result the game was actually i believe first western mobile game that hit top 10 grossing in in the japanese market okay for leveraging the knowledge of the peers um this is a this is an interesting topic and it's actually yeah okay so leveraging knowledge of the peers uh there's there's two elements in in in the knowledge leveraging the one is that there's so much information coming both internally and externally that product managers need to handle at all time by external information i mean the information that is coming just through the market change you're looking at the market you're looking at what's going on you're looking at at the at your own category

you're looking at the games what kind of updates they're doing and then there's the internal information as a product manager pretty much expected to know not expect it you have to know what's going on with your game you have to know how your cohorts are behaving you have to know how your how your updates are influencing them and that knowledge is constantly increasing as well with all the new a b test with all the new game launches it's almost impossible to keep the track of that yourself and that's why leveraging the knowledge of of of your peers of others is important a good example of a culture of knowledge sharing is zynga azinga the knowledge sharing is being run by the team called central product management and the knowledge sharing is actually divided into three different categories first is the uh the numbers reviews and and or sort of a general business updates that are done every three to four weeks during those updates basically pms together with gm's go through their game can update what's going on on the business side what have been done how that affected and what it what is being done in the future

then you have a more intimate setting of knowledge sharing and that is products uh product launches so during product launches the product managers actually show uh what what kind of different a b test what kind of different analysis of the new features have been launched and share those though that detailed information with other with their peers and thirdly there's the knowledge sharing of of competitive insights where there's a specific team just looking at different type of games that have been launched and deconstructing those games and providing product managers with that information now you might ask what this has to do with the successful product management i mean this is just description of a process true but it sets up this next example um angry birds evolution are more specifically the the lead product manager in this game dylan treadrey um what he did was was really interesting so as the game launched as as he went through soft launch and launched there was constantly more and more need of information uh and from people were asking him both internally as well as outside the team

is what's going on what has been learning what are the a b test and so forth and so forth so instead of being a sort of a reactive knowledge chair he set up a meeting every friday at about 2 pm that lasts between 5 minutes to 35 minutes this meeting was called numbers review and everybody was pretty much invited to this meeting and it turned out in from from kind of like a reactive knowledge sharing to proactive knowledge sharing and instead of him trying to to um to answer those rapid questions he would have people come in on that specific day but in you know but it didn't just do that it did something more so the meeting was was really successful and it's successful in terms of you went there and you actually learned more about about the game and what we decided to do is he actually pushed it forward and and now it's something that we do company wide where we have a one numbers review meeting where product managers and product leads are are invited and we go quickly and in about one hour through the key updates as well as key a b tests and then sequential tests and

analysis of the different features and that kind of grew up into a knowledge sharing meeting so what dylan essentially did is he created a knowledge sharing group where that wasn't really before another example of of um of knowledge sharing is is um is is this blog the instructor fun i don't know if um actually do you mind like raising your head for a second if you've ever heard about it okay that's crazy um so uh okay um yeah so the constructor fund i i started writing to this blog about six years ago and and it was honestly just something that that i did in order to improve my own learnings and it's still something i do just to just to learn and throughout the time there's just more and more games that were launched more interesting games and the games got deeper and it's almost impossible for one person to to really cover those at the same time there were great people started writing their own blogs just as anilus gupta with both guns blazing and adam telfer with mobile free to play and instead of

yeah and and it's kind of turned into this weird um i don't know for them but it turned for me into this weird kind of competition was trying to write better the construction instead of learning it's just it's it was weird so what uh what i tried to to do with them is i asked them like would you like to to learn together would you like to um to work together and happily they they said yes and i learned that i didn't know anything so by surrounding myself with with uh with really smart people that that i truly respected is is was was you know it really leveraged the knowledge and in short what i'm trying to say is you might think you know something but until you surround yourself with people you respect who are able to criticize who are smart you actually you know don't know and final one is understanding why different people like different games

quite often surprisingly often i've you know talked to product managers and they weren't that interested about games per se they were doing this for something that um they felt would progress their careers and progress them to become somewhat service managers e-commerce people or you know app people but the but the most successful ones are the ones who are actually who actually like games and it's not just about liking games because everybody in this room loves games and you know there's fortnite players league of legends players candy crush saga players you name it but what really separates a successful product manager from others is they're able to get deeper into the games and actually understand why do different people like them not just playing always candy crush saga or always fortnite always league of legends and not just outsourcing this question to the quality of research team but actually attempting to answer it yourself that ladies and gentlemen concludes my

presentation and please open for questions [Applause] hello hey you mentioned a clash of clans had the strategy to go to japan and that they were really successful i'm really curious to know what was that strategy so the strategy like in details of what the strategy was yeah as detailed as you can be i'm a bit brief i'll be very brief and i'll just say that the strategy involved creating a lot of partnerships with with local partners in japan as well as focusing on culturalization of the game and getting the stamp of approval so that the japanese players felt that this game was made by japanese okay i expect somebody to ask the question of what's the difference between designer and product manager

uh hi great talk um i am a data scientist who's working with the pms oftentimes about how to improve the monetization and the metrics that we really care about i was wondering like uh if you can give us any tips about how to communicate with the pms from a data scientist perspective and uh especially when talking about the statistics and numbers um yeah for sure uh there's uh there's i'll tell you what the data scientist told me and that was that was basically the question of why like why are you looking at this specific number and then asking instead of answering always to the what the product managers are asking you ask them why are they asking that and then you get to the root of the problem hey great talk thank you thank you quick question where do you see the lines uh being

drawn between design product and production okay um so if you it all depends on the ownerships so let's put it this way so if you have to break it down into very clear ownerships i would say that the the product manager would would have the ownership of the business side of the game so making sure that the um yeah that it's successful product per se it doesn't mean that the product manager is the only person who makes those decisions but he owns that part the design owns the part that the players like love and just enjoy the game and the producer owns the fact that the updates are coming on time that they're not that they're not filled with bugs and and that the production is going forward as planned so that would be a very high level divide of those three all right thank you

you