Gaming AMA with YC Visiting Partner Holly Liu

by Y Combinator11/9/2018

YC Visiting Partner Holly Liu recently did a Gaming AMA with Startup School founders. It was so good that we wanted to share it here.


If you were starting a new game company today, assuming all other conditions and game mechanics, retention, were equal, and UA costs were the biggest decision factor, would you: A) Develop your own IP B) Use public domain IP like Sherlock, similar to your Camelot game C) spend money at market rate to license a hot/the hottest IP you can like an upcoming movie D) spend money to license a less hot IP that has been sitting around

If UA costs are big factor, I would work to reduce the CAC as much as possible. If you don’t have as much money, I would use a public domain IP like B – in fact this is the exact reason why our first game we chose the Camelot theme – it was public domain and market accessible.

Are there any great mobile UA resources or channels that are cheaper/undervalued, but overlooked because they perhaps don’t scale for the big players?

Unscalable UA channels overlooked are any of the ones that tend to be more grassroots. I would consider looking at forums, reddit , Discord etc… I don’t know which one will work the best for you, but go to where the potential gamers are at.

Mobile games are notoriously hard to raise money for, understandably, and also require a fair amount of capital to experiment with. Where should mobile game founders go to find investors?

Yes they are hard to raise money for! If you have the means, bootstrap the game! If not, there are a couple of ways to raise capital especially if you have a team or pedigree to raise money. You can go on Angellist to find founders who were in gaming still investing. There is also a list of VCs here for Series A:

And a plug for my friends at – if you want to crowd fund your game.

Is product market fit thought of differently in mobile games? Is it more retention based than “people are beating a path to your door/demand is exploding”? Seems many breakout mobile games are more the result of game with good retention and great UA strategy than they are about releasing it and it “taking off”

It’s much closer to player / market fit. And it’s closer to consumer – if you have your servers melting, you have product market fit. But when you first release your game, definitely drive a small amount of traffic to it (few thousand) in your target and work on your retention numbers, then I would start driving more traffic once retention is good.

Can you tell us more about how you see the future of mobile gaming (Hyper casual to High-core)? Do you see gaming ecosystems and how big companies (Facebook, Twitch, Google, Apple etc) are thinking about it?

So I may be jaded here, but because mobile has become more mature, I think the ones that make it a big hit will tend towards more core and niche than casual. This does not mean that the theme cannot be casual, but mechanics will likely be core. This is because content is so readily available that it is easy for someone to leave quickly, which is what casual tends to do more of. Even all the popular casual games definitely loop in more core mechanics – progression, player level ups, etc…

In terms of gaming ecosystems – I’m a big believer in growing platforms provide gaming ecosystems. I think the larger ones (except twitch) focus on platform for all types of applications and not just gaming. It’s just that people love games and the highest revenue generating app tends to be games on any given platform. I think this is how PC games and mobile games are able to get their start. I do think there is still room for console or some type of hardware (VR?) gaming, but getting mass adoption will be difficult until you can reduce barriers of entry.

A little long-winded question: Sam Altman and others have said in the past that ‘starting a hard startup is easier in the long run than starting an easy startup’. Which means, if you’re solving a hard problem for people, initially it can be painful, but gets better over time because more people will be motivated to join you, invest in you, and be part of your mission. I worked for one such company for many years, and it was well worth it. Now, I was into game development growing up, and I got back into it in recent past with some potentially good ideas. But I started struggling to motivate myself because of questions like ‘it’s not real value to people’, ‘it’s not impactful to society’ popping up in my head all the time. It felt like making games wasn’t akin to solving a real problem for anyone. Do you ever struggle with such questions as a game designer? Do you think gaming companies have a disadvantage in terms of, 1) how much lasting value they actually bring to people, and 2) how much talent they can attract as a result? How do you get around this? And what’s your opinion on communities like that are trying to mix in game development with societal impact?

I get questions all the time about what I’m doing is frivolous. I absolutely disagree.

  1. This is how human beings learn and behave. Most people who tell me how they don’t like games or play games – I am absolutely astonished and say “well I think there hasn’t been a game made for you.” <= AND this coming from someone who did not grow up in the gaming industry OR as a gamer. Games are how we learn and connect. The first game a human being plays is peek-a-boo he/she is learning “object permanence”.

  2. Ever notice that building a great gaming company can come from anywhere in the world? You have amazing game makers in every continent. To make a Hollywood blockbuster movie, you should go to Hollywood. To build a big tech company you should go to Silicon Valley. But not games – you don’t need to leave. This to me drives home point #1.

  3. Games obviously solve a real problem for people or else they would not be spending time on there. This is the hardest problems of all, because usually what games are solving are not tangible – they are problems of loneliness, loss of purpose, sense of low worth, boredom, lack of motivation, and yes even physical pain.Countless people use games to solve all those problems.

  4. Games is entertainment. It goes a step further in that it is interactive entertainment. In entertainment you get the privilege to influence culture. People cannot be what they cannot see, as game makers we enable them to not just see (and be) what they cannot see. Help them see the greater good in themselves as a game maker.

I can go on and on… but those are some of my initial thoughts.

Games drive forward adoption of technology, they are the forefront of change, and I think its’ up to game makers to be responsible with that power. That is how you make lasting change.

My general sense of games + social impact: I certainly think games for social impact make sense, but you have make that into a game in it of itself and not really tell players it’s for social impact and then surprise them! what if Pokemon go was about exercise, that would be no fun!!

1. When you first started, how you dealt with spam/bad content in the community? What policies, mechanisms you and your team implemented? (anything besides manually scanning the content).

2. Did you used social media to grow the community? If so, what did you do? Do you consider social media an effective way to grow a community? When the value of the product lays in the community, it is hard to start.

I felt like I started 3 companies in the last 11 years, but I will try to answer the most relevant experience to you question.

  1. We had an admin tool for our fan communities and appointed officers to help us clean content. We did do it manually but we outsourced it to officers. For our games, we created some tools to help within the chat rooms to enable others to mute others and then eventually automated the prevention of spam.

  2. When we were early on Facebook for our fan communities we went into existing groups to tell people to install our application. When we did our mobile games we spent most of our time with UA – mainly because it was cheap. UA is really good when it’s an untapped channel but hard to grow. Now with UA being so expensive, I would consider doing grass roots social media.

How to model games where “Player exp” will potentially decrease?

Many “games” start from level 0, then the player exp/levels grow monotonically, so players can feel the rewards for their effort.

What about the other type of “games” – think Fitness tracker, flashcard memory apps, that players actually start somewhere, and if they stop they actually go backwards. Giving them reward points for progress itself might be misleading – they actually gain weight / forgot, say. While punishing them & deducting their exp could be frustrating.

Within games there are other things that start with a number and decrease (instead of the other way)
1. Energy – often times I’m limited by the number of lives like in Candy Crush
2. I usually start with a # of coins and I need to use it over time
3. In many pet and sim games, if I do not take care of them, they die
4. Limited number of actions (spins) done in a day

My startup is a gamified Q&A app where a global community of language learners practice by checking each other’s writings. Our gamification consists of leveling up, unlocking features and earning virtual currency that can be used to buy power ups. There’s some social interaction that consists mostly of users following and helping each other.

1. Given that we have to prioritize features, in general, how do you know when it’s a good time to build gamification features over building social features, or vice versa?

2. Some of our users are a bit upset about having to unlock features, but we use this as a motivation for them to help others (everything is free). Do you have any tip on how to make users see unlockable features in a positive light?

3. We have avoided adding paid features because we fear this could create an stratified community where things are not earned by merit alone. Is there a good way to charge for extra features without affecting the dynamics of the community?

  1. I’m not sure the difference between the gamification features and social features? How does someone level up and how can they use their power ups?

  2. Sometimes you can give them a sample of what the feature is unlocked and lock it more based off of the second time they use it. Almost give a free trial so they know what they are trying to unlock.

  3. You can charge for things that do not impact whatever they need to earn points or level up. You can limit the number of things someone can get checked and then charge them money, split that money with the translator. In a game like League of Legends they only charge for skins or cosmetic things so as not to impact the core of the game.

What are your thoughts on virtual reality gaming? Do you think it will go mainstream and if so what is your prediction to when?

Ah, opining on VR! A favorite past time of a gamers and techies.

I think right now the VR gear is to cumbersome for it to be mainstream. You have to set up a whole rig and room. Compare that to the phone, where you just have to pull it out of your pocket and start playing. On top of that, most of the content I see for VR is just not that great. So I’m going to gear myself up, look more than an idiot than I already look for less of an experience? No thank you.

I liken it to a bit of snorkeling vs. scuba diving. Snorkeling is much more accessible to the masses, less gear and less training. I can see some really great things. But if I want to see even better things, I need to learn how to scuba dive, put up the gear and prep a lot more! The content just better be worth it.

I do eventually think it will go mainstream if it can solve some of the usability issues and phenomenal content on the other side.

1. Do you see e-sports or e-sports betting as markets with exponential growth?

2. Which e-sports platform or game has the gamers that are most obsessed with statistics or live statistics?

e-Sports has a larger audience than NFL. That’s pretty crazy!

  1. Yes e-sports is definitely a growing industry.

  2. I do think betting can be big – my issue with betting is it’s been so regulation heavy in the US and have seen it kill regular sports betting companies (a la Fan Duel)

  3. Not sure.

What would you do if your product was engaging and fun, but didn’t have enough k-factor or new users to scale to meaningful size organically? How long do you iterate versus try something new?

I am assuming that the monetization strategy is ads or something with a lot of audience and that usage is important. If that is the case, I would give it 3-6 months iterating to crack the k-factor and then consider moving on.

We are building facebook messenger chatbots for restaurants for getting new customers and re-engaging existing customers. We have gamification as a core part of our user acquisition and retention strategy. We are using contests like Spin the wheel (variable rewards) and steal of the day (fixed rewards) to acquire restaurant customers on the chatbot.

We will also be using variable reward based games as a way to increase customer loyalty for restaurants. For instance, we will send out a message saying your spin has recharged, spin and see what you get. Along with this we are also building a gamified referral system with a social aspect. My questions are:

1. What are some best practices of building product retention using gamification?

2. In your experience, what are the key elements of a gamified reward system which make it addictive? For instance, I really admire Google Tez scratch cards as a product feature.

3. Are there any resources you would recommend to gain deep insights into gamification?

  1. I don’t think gamification builds retention. I think the core of your product needs to be retentive and sticky. Facebook does a great job at creating a core loop first and then giving you signals and gamifying it.

  2. I’m not familiar with Google Tez scratch cards, but in a non-game, I think creating a core loop of a behavior that can become a habit. Games work differently in that they set a core loop and the game is the purpose, whereas your purpose maybe different entirely.

  3. I would consider a lot of work from BJ Fogg to help. as well as Dan Ariely:

P.S. Bonus article:

We cofounded, which is a marketplace that lets gamers hire coaches for 1:1 online coaching and replay feedback. We’re pretty excited that someone from the gaming industry is doing an AMA here!

1. How can we combat cyberbullying in online gaming? Gamers are notorious for talking crap to each other and sometimes creates a toxic enough environment to the point a shooting happens (the recent Madden shooting) or swatting.

2. Can you tell us what your goals were when you oversaw the worldwide People and Culture function?

3. Coaching in e-sports is still uncommon and not a norm in society. How can we change that?

Congrats on launching!

  1. I think there are a couple of reasons why gamers can get out of hand: 1. usually highly achievement zero-sum oriented games, you attract competitive people 2. Layer on anonymity and it can become a place that is really difficult. If you are a game developer or a platform you can enact decency rules or better yet penalize players that go too far. I can’t find the article but I think LoL implemented a user tribune to help combat poor user behavior. This was several years ago I remember reading about that. In terms of the Madden shooting, I’m very very sad about that and I personally think it’s related to a larger problem that is not applied to gamers.

  2. The people and culture function was related company building. Once you get product market fit, company building is your second job as founder. Our goals were very aligned with the business and would change as the business would change. An example of this is we needed to focus on building a long term relationship with the customer, from the culture side, we saw that our only metric of the customer that was shared widely was revenue. This was not a great way to build a relationship with our customers. So we began looking at other metrics such as retention , alliances and also added in qualitative ways to build a better relationship with our customers.

  3. I think e-Sports needs to become more mature. Once there is a clearer paths of longevity like the NBA or the NFL then all the support and feeder leagues can happen. Right now it’s a growing but not fully mature field. This is what makes it exciting!

I’m keen to know more about user behaviour insights and learnings from Kabam, corporate social network.

Specifically, what motivated people to engage and initiate new discussions among peers? And you were to do this again, how differently you will do it?

At my startup, StackRaft, we are re-thinking professional social graphs to connect people with global and remote opportunities.

With the corporate social network we had very little traction. In our opinion it was because people did NOT want to keep talking about work when they were done with with work. They were at work to work, not socialize.

If I were to do it again, I would find something that they wanted to do and tie it with what they needed. As we are moving towards a gig and freelancer economy, I do think a social network that is wider than the company may make sense. These people are always looking for the next job and their company is themselves.

I’m actually building a mobile game. Targeted at kids and using voice as a component. My main question is what is the best way to test, validate and market your game on a shoe-string startup budget. I’d love your thoughts on my game if you had time as well.

Here are some initial ideas:

  1. Recruit friends with kids and see if you can do a “playdate” with them – offer free child care or offer it as a class! Parents of toddlers are always looking for ways to entertain.
  2. Go to the library and offer it as a class for toddlers
  3. Create a signup list for the beta
  4. Infiltrate some facebook groups and recruit users there.

Hope that helps!

P.S. I’ve assumed that this is a game for toddlers because of the name.

We’re on a mission to connect the human brain to the computer using an AI-powered headset device. We’re currently exploring potential applications for our product in the gaming industry and I’m more than happy to have the opportunity to ask you a few questions.

1) What do you think is the next big thing in the gaming industry when it comes to character control?

2) Do you think that adapting a game content based on physiological signals might be a good idea at all?

3) Do you envision any disruptive innovations in improving the video game console (components, electronics and/or any other hardware-related improvements to be made), controllers/joystick upgrades, etc. happening soon?

4) Are you familiar with any gaming studio company that is experimenting with hard tech innovations or products at the moment?

  1. Funny enough, I’m bullish on voice as the next big gaming platform. Mainly because it’s immersive and so easy to get into.

  2. Could be really interesting!

  3. I certainly think the more you can remove barriers of entry the greater mass adoption will be. I honestly think consoles and controllers and terribly unusable. This is how I think mobile gaming was one of the reasons it could grow so fast

  4. A YC company called Plexus is creating a smart glove, but they re not a gaming studio. There are a ton more YC companies that are hard tech innovations, less with gaming.

1. Why didn’t the social network for corporates work? and how did you pivot from fan communities to mobile gaming? Especially what made you or helped you take these pivotal decisions.

2. Assumably, first thing a B2C based founder does is that – sharing the amateur(beta) version of the product to friends, and family and ask for their suggestions, Is this a good way because this set of users don’t give us the straightforward reviews? how did you manage to get product market fit for “Kabam”?

(Funny thing is my mom[traditional South Indian] understood the business, she is playing with the app and literally is doing the marketing for me which I did not realize until a few days back – she calls up her friends and ask them to try out the Android version).

  1. I think the way we executed it, we sent an unsolicited email so many employees were confused and hesitant to use it. Many companies thought we were not trustworthy. Basically no one was using us, so that is why we pivoted. When we pivoted we learned it was really hard to build a platform and for us that meant we could not get users to come to us, so we learned – go where the users are. We were incredibly lucky as Facebook had just launched their platform and they were growing like gangbusters.

    Product-wise we learned a couple of things:
    a) We did not hit people’s passion points. People were not searching for stuff about work outside of work. Therefore we couldn’t ride off of any free traffic from Google by people searching for our stuff
    b) We needed to provide value within an ecosystem we could play in. This meant instead of going after enterprises (we didn’t know inside the company their rules), we could create fan communities around TV and help TV show creators get to know their audiences more.

  2. Yes find the customers and watch them use it. IT’s more important to watch them use it than to have them give you feedback. You certainly should ask for feedback, but in B2C, what and why they say it is worthless compared to what they do. I feel like in user testing which we did, it is there to look for patterns. But most people cannot tell you if they will use it or buy it = they may say it and not do it.

When asking users questions, ask them about the last time they used a product like yours, ask them to show you how they found it it, what they did, etc… Easier for users to be telling a recount than what they WOULD do.

1) At a high-level, how did you approach game design when leveraging others IP vs. creating your own?

2) Did you typically use multiple structures for agreements? Were the majority of rev-share based with some level of a guarantee? What were the core variables that you found really important to discuss on both sides to have deals that ended up with both parties happy?

  1. In leveraging someone else’s IP you have to be true to that story and things in that world. In many ways for us it de-risked execution because we did not need to invent new characters or story line. Therefore game design for us focused heavier on systems design, progression and level ups. In creating our own we needed to think about story line and environment more in addition to systems design.

  2. By the time we worked with IP at Kabam we were a larger size. There were definitely rev-share with a minimum of guarantees. Anytime there is attribution of traffic to figure out rev=share that should be clearly spelled out how it’s done. Also you should talk about marketing plans of supporting one another pre-through launch. It is good to stick a person to help manage this integration and ask for support on their end. Also make sure you know what the rights are for. A lot of IP licensing is sliced and diced on platforms and geographies. Make sure you know what you are getting.

I’ve a few questions about the early days of Kabam.

1.) How did Kabam, in general, choose what games to make?

2.) What was the first game?

3.) How did you decide on the idea to build Kingdoms of Camelot?

  1. Our CEO really liked games, almost all kinds – MMO, RPG, FPS, and strategy/ city builder games. At the time Facebook only had the clicking games from Zynga and the one that made the most sense to build was a strategy/city building game. It was within our capabilities and really put a graphical interface to some of those text-based Facebook Games (remember Mafia Wars?)

  2. The first game was Kingdoms of Camelot on Facebook

  3. We chose Camelot as a theme because it was in the public domain, accessible and fun rich deep theme.

I am a cofounder on an educational game company targeting adult learners (you may read Anthony’s question, too).

My question is evolving around how you think it is possible for game based startups to get an MVP for investors out the door, in a decent amount of time and with an acceptable quality. For me this seems way harder to pitch to investors as one has to fight the aspect of doing a game (some investors already see this as less serious than web services) at the same time as fighting the effect that people focus way more on graphics in a game than on its functionality (compared to websites). Using game engines and assets already helps a lot, but it still seems that games make it hard to match the growth rates and development times expected by startup investors. Can you provide insides on how to solve this problem? 🙂

This is a hard problem. In general investors shy away from investing in games not because it’s “less serious” but because it is “hits driven” and it’s content. So just because you make one game, it is unclear if you can make another game that is as big of a hit – in other words it is not as repeatable. Where this is different is when you have a fundamental shift in the platform, and technology or even business model to show that the games are not just one hit. A great example is the free to play model or really the IAP (in-app purchase) model. That business model was very new to the gaming world as before it was just console games (you pay up front, no more and no less) .

Other ways to de-risk it for investors is to run a crowdfunding campaign and show that there are players or capital lining up for this game. You can do this at

Or you can go to a gaming platform like roblox try to start it. If you are in a country that gives gaming grants like Australia. That is another option.

If you have a good gaming pedigree, you can look on Angellist and find some founders of gaming companies that angel invest.

1. Futuristic – where gaming and gamification will be 10-15 years time and how it will impact our societies and industries – please use your boldest predictions?

2. Any thoughts how legal industry may use, learn from or benefit from gaming industry (or gamification)?

3. What are the key learnings/advices that you can share from your role of Head of Worldwide People and Culture function? Any surprising ones?

  1. My thoughts are not as grand – I think voice is going to be a large immersive seamless platform. Platform is still incredibly nascent.

  2. Move off of Word Perfect! My feeling is that the legal industry is one of the slowest to adopt new technologies. This is because they bill by the hour. Anything that reduces their time means less profits. I think there are other things that need to be fixed first in order to start changing. The one thing the industry can learn from gaming is to align incentives with its end goal. It doesn’t seem to be the case right now.

  3. Key Learnings – people will surprise you! Good and bad. I feel like HR is one of the most under appreciated professions. They are like the complaint desk for the company. And, therefore can not attract as high of a caliber it can. It is changing a lot as more founders are moving into the company building phase. Companies that do it right have a great free system and people infrastructure to move the company in the direction it is needed. And give your HR person a hug today 🙂

How can you gamify things like elearning platforms i.e. leaderboards, quizzes etc.

I feel like for non-games you have to be true to your core loop of behavior. If you don’t have a core behavior, then you should ask yourself what is that behavior and build around it. Gamifying can help reinforce it.

I know that one of the things that I really need for our team is a strong game designer. How would you go about finding one in the Bay area (I’m in Mountain View) and what qualities and questions should I be asking them?

  1. What kind of game are you making? Game designers come in all forms: map design, level design, systems design. You will likely need to be more specific. It is okay if you don’t get the terminology write, just plainly write “what you will do” in bullet points on the job req.

  2. Assuming you do not have a lot of capital, If you have a solid team, then you’ll be able to attract a solid game designer. If not, likely you can attract an inexperienced one. If you have traction, you can also attract a solid game designer. Without these it is cash or equity. This is the beginning of what you will need to do for the rest of the life of the company – recruit.

How much weight do you give to a top-down analysis (e.g., platform, category, etc.) versus a bottom-up analysis (e.g., core loop, innovative mechanics) when selecting a new game development project?

I weigh top down analysis much more – platform and category. Sam talks about how you cannot create technological waves, but you can certainly ride the wave. This has been our experience at Kabam from a user point of view as well as a fundraising perspective. It was so much easier for us to fundraise when the gaming industry was doing well, irrespective of our performance. When facebook or mobile were growing so were our games.

Parts of our product has some gamification factor in it, particularly a reputation system that incentivizes people to engage and contribute quality answers to a question (symbolizing credibility).

1. I was wondering if you had any tips/resources for going about designing a reputation system?

2. We would like to test our MVP soon and was wondering what the process of getting and growing initial testers was like in your experience?

  1. I’m not sure off the top of head about designing a reputation system. However, here are a couple of tips when we were designing our communities:
    a. Find elements in the badge that show trust. Usually the basic thing is tenure. Oftentimes you will see “Member since” , so you can see how veteran the person is and vet their answers accordingly
    b. Two sided reviews. The underclassmen should be able to rate the upperclassman and vice versa. I think uber and Lyft do this well. This creates trust on both sides and helps keep people honest.
    c. Rating . A rating will flow from quantity and quality. There may be more elements, but those are off the top of my list. If you are creating a community, you will need to think about roles and moderation and what they can do.

  2. If you can get a couple of folks within a school some underclassmen and some upperclassmen that would be the best. If you are close to a school you can advertise there, or you can post in a Facebook group. I would recruit a couple from each group and have them use your product. Ask them questions and watch what they do.

    Once you have your product ready, you should consider the platforms of where students hang out and go advertise there. I’m not sure if you are targeting high school or college or some other level of schooling or even country, so a little general on some of the sources.

1) At a high-level, how did you approach game design when leveraging others IP vs. creating your own?

2) Did you typically use multiple structures for agreements? Were the majority of rev-share based with some level of a guarantee? What were the core variables that you found really important to discuss on both sides to have deals that ended up with both parties happy?

  1. In leveraging someone else’s IP you have to be true to that story and things in that world. In many ways for us it de-risked execution because we did not need to invent new characters or story line. Therefore game design for us focused heavier on systems design, progression and level ups. In creating our own we needed to think about story line and environment more in addition to systems design.

  2. By the time we worked with IP at Kabam we were a larger size. There were definitely rev-share with a minimum of guarantees. Anytime there is attribution of traffic to figure out rev=share that should be clearly spelled out how it’s done. Also you should talk about marketing plans of supporting one another pre-through launch. It is good to stick a person to help manage this integration and ask for support on their end. Also make sure you know what the rights are for. A lot of IP licensing is sliced and diced on platforms and geographies. Make sure you know what you are getting.


  • Y Combinator

    Y Combinator created a new model for funding early stage startups. Twice a year we invest a small amount of money ($150k) in a large number of startups (recently 200). The startups move to Silicon