How to get 40M installs? Marketing for Mobile Games | Brian Sapp VP UA at Jam City

This video interview was first posted on my personal blog site www.gamemakers.com. Also, I work as SVP of Games & Digital Platforms at NBCUniversal in games publishing. None of the opinions, comments, or written materials in this post or on my blog represents the views of my employer.

Content Overview (Full Transcript Follows):

1:40 What is the role of the User Acquisition/Marketing team at Jam City? What is Performance Marketing vs. Brand Marketing?

2:55 At a free to play studio like Jam City how much brand marketing do they do relative to performance marketing?

4:03 What are example activities on the Brand Marketing side?

4:40 What does your day to day look like?

5:50 Describe UA activities across the full life cycle of a game from concept to launch

8:28 Could you tell me more about what you look for in soft launch? How do you do UA during soft launch?

9:38 How is your team structured? How do you think about structure?

12:55 How does the UA team work with other groups at Jam City like PM team or others?

14:10 What does it take for a UA team to be as effective and competitive as possible?

15:50 How do you get good at optimizing Creative ad units?

17:57 What qualities do you look for to hire on to your team?

19:10 How do you think about doing UA for branded IP games vs. original IP games?

21:13 Both Facebook and Google rolled out machine learning based bidding which has been super successful. What are some of the newer applications of machine learning or other new trends in UA?

22:50 What happens to UA teams in the environment where machine learning becomes more pervasive and performant?

23:28 How did you get into games? What was your career trajectory?

26:05 High level thoughts on transitioning from SF to LA?

27:00 What do you like most about working in UA and working at Jam City?

Transcript:

Jam City is one of the largest and most successful mobile free-to-play game publishers in the world with a lot of hit titles such as Cookie Jam, Panda Pop the recently launched Harry Potter Hogwarts Mystery. More recently they have made a couple of acquisitions such as Disney's Emoji Blitz from Disney and Bingo Pop from Uken Games based in Canada.

How has the company been able to be so successful? 

Well, one of the key pillars of success in addition to the games themselves is the marketing side

of the equation which is more about user acquisition and how do you get players to come in to your game and play your game. Today we will be speaking to Brian Sapp who heads user acquisition/performance marketing at Jam City and we will be talking about a few key topics with respect to user acquisition. 

For example, from a marketing perspective what is the difference between brand marketing and performance or user acquisition based marketing? What does it take to be successful in the new world of user acquisition? The type of teams and how you would structure a team if you're building a user acquisition team for your mobile game and a lot of other great topics that we're going to talk about in today's interview with again Brian Sapp, VP of user acquisition at Jam City. 

Let's go!

JK: I want to start by just assuming a general audience; could you talk to me about the role of the user acquisition and marketing department and your role specifically here at Jam City?

Brian Sapp: Sure, so my name is Brian Sapp I oversee the UA marketing team. It's a team of about

24 people right now and UA marketing or performance marketing is really marketing that is tied very much to measurement.

And really the core of it is data. For most mobile gaming companies that are free-to-play, performance marketing is a big driver of both the growth and the revenue for those companies. For us, especially everything we do around what channels we spend with, what we're focused on, is really costs per user and then the revenue coming in from those users and the quality of those users. 

And so when we say performance marketing it's really around these KPIs, these data measurements that we really care about which is slightly different than brand marketing. Where that's more encompassing around awareness; different channels that you can't necessarily measure how effective it is. But performance marketing is all about measurement, all about data.

JK: In a free-to-play game studio like Jam City you mentioned both performance marketing and the brand marketing. How much brand marketing exists within Jam City relative to your side: the performance or user acquisition side?

Brian Sapp: So Jam City was a mobile first company and was built on the back of performance marketing. Performance marketing has been one of the biggest drivers allowing us to grow the

business the size it has. In this year alone when we released Harry Potter, the game has done over 40 million installs. And a lot of that was driven by one, it being a very successful IP, the other part is scaling user acquisition, and so we are definitely a performance marketing heavy company. We do do brand marketing it's just not at the core of what drives our growth and that's slightly different than some other gaming companies especially ones that are more entertainment focused such as Warner Brothers and NBCUniversal.

JK: On the brand marketing side are there examples of some of the brand marketing activities your group does do?

Brian Sapp: Yeah, so just to give you an example we have an integrated marketing team and they're working on activations that might be things like working with influencers around certain

video game events. It could be working with celebrities and having them come in both into the game and then marketing collateral.

They're always working on ideas that kind of fall outside of traditional performance marketing but they can actually help enable performance marketing to perform better.

JK: Can you talk to me about what your day-to-day looks like? What are you doing on a general basis?

Brian Sapp: Yeah, a lot! So we have eight titles that we are running UA at scale on. We also have three titles in soft launch. For me, at the highest level, it's setting the budgets for the games based on our goals and objectives which is usually around payback and how we want that money to come back to us. The other thing is kind of setting strategy and helping the UA team set strategy for each of the games and different types of tests that we want to run. And then working on creative analysis and kind of driving the creative team to really push ourselves around performance of UA creatives and then planning for the soft launch titles. Helping set kind of checkpoints into when we think these games are ready for us to give a green light to.

Between all of that, all these games are in different life cycle phases. It's kind of different strategies for all of them but really helping the team hone in on the right strategy.

JK: Can we dig into that a little deeper? Assuming a typical game launch from conception to pre-production to production to soft launch and then hard launch, can you talk about how does UA get involved? How do you work with whether it's the production team or other teams during each phase in the game development process?

Brian Sapp: Sure. So soft launch or I kind of call it the R&D phase really has kind of generally three phases. The first phase is the tech phase where we kind of get the game out in the wild and see if there are any bugs. In this case UA is basically trying to get enough users to make sure we can get a good read on the technology.

Second phase is generally early retention and so from a UA standpoint we are trying to drive cheaper installs between tier 2 tier 3 countries with a focus on getting statistical significance on retention. We're not necessarily focused on what's happening later in the funnel, we're not focused on getting our best performing creatives in just yet nor are we really ready. 

It’s really about stat sig and what's happening in games so the game team understands the metrics they're seeing and where they need to improve.

And in phase 3, we call it getting ready for the big leagues when we start going tier 1 territories we really start honing in on creative and seeing what works trying to understand what our predictive LTVs are and what our cost per user is. And then building out the business

plan for the game. Is this game viable or is it not viable? And then we go into a worldwide launch if it is viable and from there coming up with a worldwide launch plan which is going to be different for every game.

For Harry Potter we were very aggressive because we knew that the IP was so strong, there hadn’t been a free-to-play game for Potter in the market and so it was a pretty large global launch plan.

For a non IP title it might be a different strategy. It might be slow growth. It also depends on the genre. For some of our match-three puzzle games we're looking at multi-year paybacks therefore the risk is higher in the beginning of spending too much against it. And then once we're through the worldwide launch period, it's really a question of what's the strategy to grow the game and should we be aggressive now versus taking our time and then kind of later stage we start moving into retargeting and how effective is it?

Depending on the games lifecycle its definitely always a different strategy.

JK: Just going back and digging a little bit deeper… let's say in soft launch you kind of test a lot

of sort of technical stuff at that point and then you focus on let's say d1 retention or some early retention metric. And let's say the metric, it doesn't quite hit what you guys are looking for, what

happens then? Does it go back to the product group and then when they say they're ready to test again or how does that work?

Brian Sapp: Yeah generally we're spending very little amounts during this time so there's not a lot of risk for us to just keep spending on the UA side. And the product team will basically have these benchmarks that they want to hit and if they're coming below them they will, you know, every update they're putting in features to address that right. Do we see it improving or do we see it decreasing? 

Of course you you don't have to exactly hit your benchmarks to go worldwide, I think the more important thing before going worldwide is that the overall business plan makes sense. In other words, your cost per user is less than or equal to your LTV and that's what's most important however you get there.

JK: Can you talk a little bit more about your team’s structure? The user acquisition landscape is kind of complicated and there are various ways you can structure your team; whether it's based on channels or based on games. From a structural perspective how have you thought about this? How I'm going to set up my team to be the most effective

Brian Sapp: That’s certainly a challenge and I think there's no right or wrong way to do it. What I've done here at Jam City is basically the UA team is made up of four teams: there's the media buyers, analysts that help us build the dashboards and analyze all of the data, there's the creative team that's basically all of the people making the creatives and plus we outsource some creative, and then there's creative strategy team which is a small team but they're hyper focused on what competitors are doing what's working best for us overall kind of best practices in creative and applying it across all games.

Okay and so those are the four teams that make up the UA marketing team and then within the media buying team this is where it can get extremely challenging. Just to give you an example you can allocate a team around channels and you tend to see that more often when you have one or two games at scale or you see it with non-gaming, such as Netflix or Hotel Tonight,

where it's really just one product and those teams are pretty large and they're focused around channels or geos or sometimes both. We were set up that way when Jam City launched and went to scale with really only two games, Cookie Jam and Panda Pop, and at that time I think that made sense. 

But as we released more games, communication and strategy across the team became very challenging  when a buyer was only focused on their channel right. What was

being missed was kind of the holistic view and health of the game. So we actually ended up reorganizing the team around supercells is what we call them but they're basically small teams of four to five that are focused on a group of games. And they split up the channels amongst themselves but more importantly they're thinking holistically about the strategy of each of those games.

JK: Why do you think that strategy worked? Do you think it's because the supercells felt more ownership against those games? What do you think led to the increased performance?

Brian Sapp: I think with the smaller team you've actually had improved communication. You got three to four people working very closely together as opposed to twelve people working across everything together. That's one. It was improved ownership like hey we're responsible for these games no matter what and then it was putting the right personnel on those teams. The right experience of Facebook buyer, network buyer, and then strategist or someone who's a little bit more senior who understands like the overall game. How are paid installs doing? How are organics doing? Where are we seeing opportunity? To kind of going out and like coordinating amongst the team. 

So it's a combination of more ownership, better communication with a smaller team, and the right personnel. It's almost like a Special Forces team. They are way more effective when each person has their role and they work well together as a group

JK: How does your user acquisition team work with other departments within Jam City whether

it's the product management team or other teams? Like what kinds of communication happens and who do they have to work with so that user acquisition can be more effective?

Brian Sapp: So probably the closest working relationship is the PM team and the PM teams are responsible  for the business of the game. Every game is a living breathing entity. It has live events, has certain updates, and the PM's are also responsible for the P&L so the UA team is constantly communicating or ramping up or slowing down spend around events when it makes sense around tent pole events as well as like what the budget is. How we are pacing and how performance is doing? 

And then beyond that we interact with integrated marketing is doing some of the brand marketing activations, publishing which for us is kind of the App Store optimization team and UA actually handles a lot of that as well and then to a lesser extent PR and other teams.

But PM and integrated marketing and then publishing.

JK: There's a view that user acquisition plays a very strong component with the success of mobile games today and increasingly more important for the user acquisition team to be doing a really good job. So from your perspective in today's market what does it take to be as effective as possible?

Brian Sapp: So clearly it's become very important in the last couple of years is understanding how to buy on social platforms. When we say social we really mean the self attributing networks such as Google and Facebook.

Those two kind of control the market at this point and it's an absolute necessity to have UA buyers understand how to buy on those platforms. Outside of that, the networks are important for scaling but social is most important.

And then I would say creative.  Creative has become very important as the market has gotten more competitive and so really trying to understand the psychology of what drives players into

the game. How to differentiate amongst the competitors but also what drives the payers and what that motivation is. 

For example, at WB I worked on RPGs and we found that showing character progression rank up really brought in payers for whatever reason. That's what really motivated them and so for a puzzle category it's something else and so for each genre and for each game you're gonna find creatives that really resonate with the users you care about and then just getting smarter and smarter about that.

JK: How do you get good at that side, on the creative side? So let's say someone's part of one of your four teams. On the creative strategy team, what does it take for someone to be 

really really good at coming up with these ideas - you know for creative ideas so that you have high-performing creatives

Brian Sapp: That’s an interesting question. The reason I say that is that I've found that there's no faster way to get your ego beaten down than to come up with creative ideas that you are just absolutely positive are gonna work and they don't. And every UA team has had that experience and I think people who haven't done UA or come from brand marketing which is less measured tend to come in with like a lot of determination that they're right and they haven’t been beaten down yet.

JK: So do you think it’s important to have an experimental framework?

Brian Sapp: Yes, yes, and it's important to try a lot of different things. In fact, throw your preconceived notions out the window. 

To give you an example, I've run very expensive CG trailers and then run I would say lower quality five second ads and the five second ads of lower quality outperformed the CG Trailers. So that's not uncommon. And what it takes is just you know our creative strategy team is focused on IPM which is basically impression to install and then other later funnel metrics and we're hyper focused on improving it. So we'll look at what's working what's not

working and we try to like basically use the learning across all our games. So if it applies to one match three let's apply to all match threes so that helps accelerate learnings and then looking at the competitors and trying to understand what they're doing. 

We take feedback from the studios as well but sometimes the studios they're not performance markers and they think they that high quality art always wins and it doesn't and that's when I kind of go back to the fact that like we need to be looking at the data right?

You’ve got to understand the best looking creative doesn't always win. It's just a fact

JK: Going back to your team let's say there's somebody out there interested in getting into performance marketing and user acquisition what qualities do you look for when you're hiring people onto your team?

Brian Sapp: Excel skills are a must. We generally will send out Excel tests, I also will put together, often depends on the seniority of the role, but just put together an example question around if you had a million dollars how would you launch this game? There's no right or wrong answer it's more that I like to see three or four slides on how a person thinks.

So that kind of comes down to how they communicate their ideas and what thought process goes into it. Because it's important that when you do UA you communicate what your strategy is to the game teams. 

So basically it's data chops which is kind of Excel and focus on numbers and then it's communication and kind of strategy and being able to communicate well with slides and giving presentations. Then I think lastly is like you know creative thinker who can  think outside the box.

JK: So when I look at Jam City in terms of the product portfolio and the product mix, there seems to be a shift towards branded, IP games. You mentioned Harry Potter and there is the acquisition of the Disney Emoji Blitz team. As you think about doing user acquisition for branded IP vs original IP what are some of the considerations? What's the thought process behind when you think about doing marketing for a branded game versus original IP?

Brian Sapp: Jam City has always kind of been in both worlds. Some of our older titles like Book of Life are IP titles and Snoopy Pop for example. And they both come with advantages and disadvantages.

Some of the disadvantages of working with IP is obviously you have to get approvals from the IP holders which can be restrictive and sometimes slow down the UA creative process but the disadvantages of working with non IP is that it’s not well known.

Generally speaking IP titles have a much different launch than non-IP titles and what I mean by that is since IP titles generally speaking as IP titles are well known and are usually gonna see a very low CPI at world wide launch that is an opportunity cost that I for one don't think you should pass up on.

JK: So you want to be more aggressive at launch?

Brian Sapp: I look at the launch as almost  like supernovas and every IP is different and I've definitely seen IPs where there is no CPI reduction and you know you throw that out the window but for Harry Potter we saw amazing results. So we did push harder.

Versus non IP you're not gonna see that same momentum at launch so you're just taking a more calculated growth perspective on non-IP.

JK: You mentioned the most important channels for UA were Facebook and Google and certainly there's been in the last two to three years a lot of innovation with respect to like you know machine learning based bidding. Can you talk about some of the recent changes or some of the new things happening because it seems like there's a lot of application of machine learning whether it's to bidding or whether to stuff like Bidalgo applying machine learning to creatives optimization. Can you talk about some of the new things coming up? 

Brian Sapp: Two years ago Google rolled out UAC which basically I think at the time there was a lot of backlash because you could target YouTube words, search independently and then when UAC

rolled out they're like nope you can't do that anymore. The machines are smarter let them do it and I think it was a little bumpy for a while but now we're seeing great results and clearly it makes sense.

With a long enough timeline with enough data and the right algorithm all these big platforms are going to be able to basically optimize better than you. It can and save you a ton of time and now Facebook is kind of going in that direction. Some of our best performing ad sets for Facebook are just broad targeting male 18 to 55 and then you give Facebook your goal on

AEO, app event optimization, and it just does it’s thing and so I think everything is going in that direction. It's just a question of when it gets there.

JK: What happens in that environment? So does creatives optimization become the most important thing?

Brian Sapp: Creative optimization and I will say that targeting is still important in that using your own data to enhance your targeting. For example creating look-alike audiences off of an in-game event and testing a bunch of different in game events to localized audiences. And one out of five may be great. Separate from purchase event optimization which is still the holy grail.

But it's really getting smarter about your own data and how you can use that to your advantage.

JK: Shifting gears a little bit I wanted to talk a little bit more about you. So how did you get into the industry? What was your career trajectory leading up to coming here to Jam City?

Brian Sapp: We’re in LA now. I went to college here and like played in bands actually and was in the music industry. I was basically negotiating the rights for music to be in the movies, TV and film for years and didn't love it but I was doing it because I played in bands  and it was kind of like I could be close to it. 

However, this was like 2010, 2009, both those industries were dying and mobile gaming started right around that time and especially free-to-play and it just excited me so I actually quit my job and moved up to the Bay Area with six months savings and was running out of money crashing on a friend's couch.

And I interviewed at tap joy when it was only 30 people and I had gone on like 24 job interviews that no one hired me except for Tap Joy and so I got into Tap Joy when it was only  30 as like one of their first direct developer partnerships people and my job was to bring on game developers as advertisers and publishers and the rest is kind of history.

Tap joy just blew up, we were one of the biggest platforms at the time and I got to work with Kabam, Supercell, Zynga and all the early guys. It was a great crash course in all things both UA data, monetization,  game mechanics, we had our own publishing fund and we were like funding games. It was awesome for me personally in my career and then I decided I wanted to go compete because it was very consultative but I couldn't actually make decisions. So I had a great opportunity at WB games to oversee UA analytics and monetization and that was awesome and had a lot of responsibilities and got to learn a lot and was a part of a lot of big launches including Game of Yhrones Conquest and Mortal Kombat and Injustice 2 and so that really kind of honed me on how to work with IP holders especially and it was small. For being a big company we were a small scrappy team, didn't have a lot of resources and so was there for three and a half years. 

And then I had known people at Jam City for a long time knew about the great culture and really wanted to join a mobile-first company that had data at its heart and had resources So decided to join Jam city about a year ago overseeing the UA marketing team.

JK: What are your high level thoughts in terms of transition from the San Francisco Bay Area to Los Angeles?

Brian Sapp: Well it’s becoming more San Francisco like by the day. When I moved here I thought it would be difficult to recruit people from San Francisco and actually it’s been the opposite. It's been really much easier than I thought to get people to move down because the cost of living is a little bit lower. It's not like insanely cheaper, but the weather is a hell of a lot better and then it seems like LA is just really exciting.

Like you know we have two football teams come here now. They're building a train line and I take the train to work. I couldn’t do that before. The city is changing, they are putting in a lot of money ahead of the 2008 Olympics and then a lot of tech is moving down here so it's just it seems to be coming up right now.

JK: Last question. What do you like about working in user acquisition the most and then what do you like about working here at Jam City the most?

Brian Sapp: Working in UA the most is just the live operations of it. I love that every day it's dynamic. The data changes you know it's just it's living, breathing all the time and I like that. I like being always on call if you will. I mean that sounds like most people wouldn’t like that but I just like that it's always going and always having optimizations and things always change and all of our

games are living entities. And I love launching games, there’s nothing more fun then world wide launches. What other products can you get millions of people within like a weekend. I can't think of one so that's super exciting to me.

And Jam City it's just the culture: the willingness to take risks the willingness to be aggressive and you know it comes from Chris and Josh and the founders down. And then the autonomy. 

JK: Great! Well, thank you very much for your time. Everybody I'm here with Brian Sapp here at Jam City and that's it.

If you enjoyed this interview please subscribe to the GameMakers YouTube channel here or you can go directly to the www.gamemakers.com website.