Interview Transcript

Can you start by just laying out exactly your role at Unity?

I was heading APAC strategy where Unity was running strategies as a task force group, so I engaged in a couple of significant transformations, in terms of business model change and going out of gaming. From that perspective, I have much to share.

So you were working mainly with the developers, but also branching out into new sectors and industries?

Exactly. My later experiences with Unity was working with developing businesses with automative film and animation and architecture and engineering companies, as well, for VR and AR companies.

I want to start by diving into the engine market. How would you compare Unity versus Unreal?

If you’re interested in game development, you definitely know but, very simply put, Epic Games, they develop games by themselves. They are the user of their own engine. Unity, on the other hand, had opportunities to develop their own games, but they decided not to, for over a decade. What I’m trying to say is, Epic has their own view of developing a game engine, probably for the sake of their own development. Whereas Unity is developing businesses and technologies, mainly for the longer tail of game development and their studios.

Why did Unity decide not to move into publishing?

I think it was just a strategic decision as to where we should focus. If you really focus on developing your own game, there will be a certain type of segment or genre of games that you want to develop. But I think you need to want to be more universal. Unity’s mission was, basically, democratize game development. That means anybody who is interested in developing their own games, to represent themselves, their taste, they can develop their own. From that perspective, the hurdle to adopt Unity, learn Unity, should be lowered. Unity is approaching it as game engine development platform, rather than wanting to build their own games.

How would you split the market share of Unity versus Unreal and the different platforms?

There’s no hard number that I can share but, from my perspective, I would say in mobile games globally, I assume that Unity’s share would be around 50% and, probably, Epic Games would be less than 20%. But it really depends. If you look the game share for the top 20, globally, the Epic ratio would increase quite significantly. Whereas if you look at the top 1,000 games, I think Epic’s share would go down to less than 5%. Unreal has been used mainly by the bigger studios.

That’s because Epic is more complex or is used for more complex games and graphics and technology?

Exactly. One interesting example is how the game market evolved in the South Korean market. The South Korean game market has grown rapidly, for the past 20 years, driven by PC games. The South Koreans are huge PC game users. 10 years ago, they shifted to mobile, with the legacy of Epic Games or the Unreal engineering team. Then it actually expanded into the China market and they have some Unreal legacies, as well. Going back the early 2000s, there was no Unity and there was no third-party engine; there was no other competitive third-party engine. A lot of Korean major game studios didn’t have competitive in-house engines and they adopted Epic’s Unreal engine. Then the legacy continued about 10 years. Junior and senior engineers and developers are very acquainted with Unreal; that’s where legacy started and it still continues in South Korea and in China and neighboring countries.

How would you compare the technology between the two engines?

I think one of the greatest assets that Unreal has is AAA quality, overall, and the branding. If you think about how developers perceived Unity versus Unreal, there was a very clear separation or distinction, between the images and the branding they have. Unreal is for AAA quality games and Unity was more for casual games. But then Unity really tried hard to catch up on performance and technologies and filling the branding gap so that they could produce AAA games, as well.

I think the gap is now minimal. Although there could be some gaps, I don’t think it’s really significant, as of now. What’s more interesting is the gap that needs to be filled in, not technology or performance wise, but more branding and marketing wise, or perception wise.

Is that just because Unreal has a perception that it is more complex or better quality, versus Unity?

Exactly. It is also interesting, if you look at what’s going on outside gaming. A lot of AR/VR content used to be created by real-time render and I think they have a different perception, compared to the perception that gaming developers have.

What’s the perception of the AR and VR developers?

Let me put it this way. If you look at the market share of game engines, in the gaming industry, let’s say that Unity’s market share will be approximately 50%, whereas if you look at AR/VR content, Unity’s market share goes up to 80%, maybe even 90%. That means that the legacy of the brand that was deep rooted, doesn’t exist in industrial applications, outside gaming. Developers in the gaming sector or the industrial sectors, they have different experiences; they have different brand images. Epic’s presence is, I think, much weaker, outside the gaming industries.

Does the engine suit AR/VR much better at Unity, or is it, again, just a marketing perception thing that increased their market share?

I think, relatively speaking, in terms of real-time rendered backed content creation, gaming definitely surpasses all other contents, in other industries. For developers who want to create real-time content in other spaces, I assume most of them lack experience or expertise, in terms of creating it. Having said that, as you mentioned, Epic has perceptions of being difficult to learn and hard to code. That actually applies similarly for the rest of the developers outside of gaming industries. In that sense, Unity is in a very good position to be approached more comfortably and make Unity easier to adopt and try out. I think that’s probably why Unity’s share in AR/VR content is so high.

Could you share some context as to how the engine business has evolved? What I mean by that is, how Unity have looked at the engine as part of their total business with the ad network and not going into publishing?

To me, I think there are two questions there. Let me separate them out, from my perspective. One question would be, what are the kind of directions or trends that we are seeing in the gaming market, in general? The second question would be, how has Unity evolved its business model and how did it successfully monetize.

Going back to the first question, I think, maybe three decades ago, there was no significant third-party gaming engine companies. They all had in-house engines and some were good and some were bad. But then, as the gaming industry developed, some started to emerge. The best ones survived. The interesting thing is, Epic tried out a model where they were developing their own games but they thought their in-house engine was so great that it was sellable, as a product or service, so they started to sell it.

Unity, on the other hand, I believe that they didn’t have to tackle this gaming engine market in the beginning. In the beginning, they concentrated more on how rendering could work more efficiently, in making the gaming application. Then it developed into a general gaming engine model. If you look at the trend over the past three to five years, I think it’s quite interesting that a lot of the traditional major gaming studios, they stopped developing or using their own in-house engine and they started to adopt third-party engines, such as Unreal, or even Unity, for their major titles.

I think there are a couple of reasons for this. One would be, as the market grows significantly, if you think about the different types and manufacturers of handset, the number is increasingly incredibly, every year. Each in-house engine, they hit an inflection point where they can’t really keep investing to embrace all of these changes, by themselves. At a certain point, economically, they sense that it is better to adopt a third-party engine. They also hit a point where a third-party engine, including Unity, improves their quality and performance and is easy to integrate into their own system, or team, or engineering capabilities. All these pieces come together in a harmonious way and, I think, over the last two to three years, we have started to see a lot of major titles are made with a third-party engine. That’s the general trend that I see.

For the developers and publishers that have shifted from in-house to Unreal or to Unity, is this the software CRM of enterprise businesses, for game developers? Do they look at the engine as a mission-critical part of the technology stack, where the retention and stickiness are very high, for developers?

It is actually related to question number two, about monetizing. Let me put it this way, the biggest revenue driver for the Unreal engine is, basically, revenue share; taking a certain percentage out of their customer’s success. That could be really big.

Like 5% of revenue?

Something like this. As far as I know, the rate gets lower, depending on the size of the studios. But Unity’s model, if my memory is correct, before 2015, Unity’s business model was, basically, perpetual license sales. It was more offline sales driven and, pretty much, one-time sales. You have to engage with them about a year and a half or two years later, when the new version comes out. We didn’t have multiple variations, add-on sales, and we didn’t have a lot of customer engagement; we didn’t know what they wanted. Back then, the subscription model or SaaS, was not generally widely used. Then Unity took the initiative, in the gaming engine market and decided to transform to a SaaS company.

I think it was late 2015 that we announced that we were now a SaaS company. Unity did a great job, in terms of designing the customer lifetime journey, for the first time. If you were well aware of Unity’s model, it would be easier to understand, but let me just explain, from acquisition to onboarding. For onboarding we have developed multiple content and multiple engagement channels, so that developers are really linked with Unity. We also have a higher level of learning materials; we developed a certificate for different courses, such as AR courses or graphic designer courses. With the subscription model, eventually, we had three different product lines. One was Personal, which was free; Plus, which was less than $20 per month. Then we had Professional. Professional was actually targeting mid to large-sized gaming studios. Plus was more targeting money-making amateurs and small gaming studios. The free version was targeted to hobbyists and students.

With all the learning courses and with the success of development, we intended that each user should try out different tiers or lines of the product. The journey didn’t really end there. Throughout this journey, we had Unity’s Asset Store, where you can actually images or any digital asset, such as graphics or backgrounds or sound that you tried out or made by yourself, for your game. This was an App Store type of model, where Unity takes 30% and then creator gets 70% if revenue was created. Eventually, Unity had a LinkedIn version of a job marketplace, for Unity. Unity wanted to set up a place where anybody who wanted to build their own games, could look out for talent that they needed. With a certificate that we developed, they can actually prove what kind of skillsets they have, with references that they personally had.

All of this created the whole cycle of adoption of Unity and learning Unity, developing with Unity and then making money with Unity. That was the full stack of what Unity do, with their own journey.

Plus the Ad network?

Exactly. So what was Unity’s business model? Going back prior to 2014, 2015, when we had perpetual license sales, the sales were okay. It was not that bad but it wasn’t great and it was hard to scale. With the SaaS model, from game engine license sales, we saw some growth but it didn’t increase exponentially. What leveraged Unity’s growth was the in-game video Ads network. Let me just give you a bit of context. Back in 2014, 2015, Unity had quite sizeable market share, in the mobile gaming market, as well; about 30% to 40%. But it was then more casual games, where a lot of Unity’s customers didn’t make good money and so that affected Unity. Since we had so many games that were made with Unity and so many gamers playing games made with Unity, we decided to acquire one of the decent in-game video ad networks. That really boosted Unity’s growth.

What’s more, Unity had a good strategy and it also acquired a game analytics company, in a similar timeframe, so that the game analytics could provide better insights, in terms of Unity’s customers. Game developers can target much better for any campaigns to their gamers; the customer’s customers. Eventually, just like a virtuous cycle, it incentivized game developers to adopt Unity and develop games with Unity. So there was more Unity Ads revenue coming up and the accuracy or performance of Ads plus the analytics got higher and higher.

How did you look at the lifetime value of the developers?

If you think about perpetual license sales, you have a certain license price. Let’s say, $1,500 to $2,000 per seat. Then if the studio is mid to large sized, then you probably need enterprise support, where if you have an issue, you have to ask Unity, because it’s hard to solve, internally. That was a serviceable product, back then. You could probably sell source code, if the studio was acquainted with Unity. All these companies represent an average lifetime value. It was quite simple.

Was that $1,500 to $2,000 per year, per seat, or is that over a three-year license?

It’s forever, because it was a perpetual license. Usually, in Unity, every 18 to 24 months, we had major updates. These major updates could be sold at 40% to 50% of the original price. When we calculated lifetime value, for big accounts, we had historical data and we had average update periods and then we had average recycle or renewing contract periods, globally. Then we had a combination of seats and services and source code, all together. We can have small, medium and large ideal prototypes or models where they have a certain number, with a certain cycle period. We can assume that if we acquire this sized studio, in this region, then we could expect this much lifetime value coming up.

It got trickier and more lively when we transformed to a SaaS company. As I mentioned, onboarding and everything, it was really complicated and you have a lot of layers, including certificate revenue and things like that. Then the Asset Store revenue really kicks in. Asset Store did exist prior to the SaaS model, but with this frequent engagement with the SaaS, the Asset Store can be really sellable. It’s quite interesting to understand that if we had an accurate calculation, in terms of lifetime value, in the perpetual license model, then it could indicate the pricing point for the SaaS model, as well. It’s a really good indication to understand the willingness to open their wallet, basically.

What was the retention, on average, for the larger accounts, versus the massive long tail of developers using Unity?

I don’t have the hard numbers but, obviously, for the major accounts, their retention rate was really high. Unity fans would use Unity 100%, but most major gaming studios, they probably use 30% or 40% of their entire gaming product with Unity. Having said that, let’s say a major gaming studio had 100 gaming developers and they are split and are developing five different titles; maybe 20 developers per title. Probably they are using Unity for three titles; that means 60 developers out of 100. These three titles have a developing period of 12 months and then usually, if this ends, maybe not simultaneously, they have another upcoming title, as well, so it goes on, one by one. But eventually, if you flatten that out, usually the retention rate is, I believe, over 70% or 80%, for a major account.

Why is it so high? I think that if that studio was developing games 100% with Unity, the retention rate can go up and down. But a lot of major studios, I think, are using Unreal for several titles and using Unity for several titles and maybe several with their in-house engines. The specific share is more sustainable, in the longer run.

You’re saying the retention is around 70% for paying users of Unity?

70% was going back to my example of 60 developers out of 100 are using Unity. 70% means that, after 12 months, maybe 70% of those 60 people will still be using Unity.

They would be paying, as part of the studio, for Unity and then they will be working on new titles, but then the others just drop off and go and work on different engines and different titles?

Exactly. But then the new title has a chance to use Unity or Unreal, but they are pretty consistent.

So the studios just buy different licenses, per seat, and then just have a flexible model where they can take developers off different engines and move them around?

Yes, exactly.

How does that compare to Unreal, which is more of a fixed revenue share?

I believe, recently, Unreal has adopted a Unity-like model. For the smaller studios they don’t really force the revenue share model. They just sell licenses in an affordable way. I think subscription model wise, the business model develops in a similar way. I think what matters is large studios. With hobbyists, the retention rate is really high. Hobbyists can just adopt and play for a week and then they can just leave. They probably come back six months later. If you calculate it that way, the retention rate for Personal will be really, really high. But that’s actually the intention of building Unity Personal.

How do the large studios look at insourcing the engine, potentially, versus using a third-party engine? If a game is a big hit, do I want to bring some of the game engine in-house or am I happy to keep paying Unity because it’s a fixed price, no matter what?

I don’t think there’s a clear answer. As I mentioned, devices are all connected, but the variations are going crazy. We are trying to do the same game with my cell phone, my tablet and my smart TV. You have to invest a lot of money into all of this, just to develop your own in-house engine. As long as the third-party engine performance is great, I don’t think that there is any reason to develop your own. At the same time, if your game becomes very successful and you become one of the major accounts, I think there is some room for you to request to Unity or Unreal, that they develop a certain feature that you might really want to highlight in your game.

What do you think is the biggest risk for retention on Unity?

I don’t think there is an obvious or immediate downside for Unity, in terms of retention perspective. Rather, I think there’s a risk for further upsides.

So you think that there are not that many risks to Unity falling below that 70% retention number, but more opportunities to actually increase it.

Exactly. If you think about it, you have to think about competitors or alternatives, if you think about the risks for retention rate. Except for Unreal, you don’t really see any competitors evident here, in gaming engines.

Is that because of the scale? It costs so much money to build these engines, no one is going to compete now.

Because of scale and because of the price point. I don’t think that new entrants to the market are able to provide an affordable price point, as Unreal or Unity.

What do you think about the price point and the opportunity to increase the price, for the larger accounts?

From my perspective, I think for the larger accounts, there are many more routes to price increases, per seat. Compared to the value that Unity’s engine brings to big accounts and if you think about the size and scale of the business they are creating with the Unity engine, I think it’s quite insignificant.

What’s the average revenue of a big studio, versus the amount that they pay for Unity?

We can quickly do the math. I’m not talking about a billion-dollar company. Let’s say a company with a revenue of about 100 million. They probably have about 100 Unity developers alone. Let’s say, everything combined, they are probably paying about $3,000 per year, per seat, including the seat and additional services. $3,000 per seat and 100 developers, only makes $300,000. If you think about it, it’s nothing compared to the 100 million that they are making. Maybe that is why they are adopting Unity so much.

What are the add-ons that can increase that price? There is obviously a fixed fee for subscription to the engine. What are the other add-ons that can increase the revenue, per user?

It’s like the chicken or egg question. Going back to the example that we mentioned, if a studio generates $100 million per year, if they used Unreal, just the revenue share alone, say 5%, that’s $5 million. So it’s incomparable. It’s way bigger; about 15 times bigger than Unity.

Is Unreal that much more expensive then? They had the 5% rev share, but they also had a discount if you distribute on EGS. It’s a bit confusing in that sense, but it does seem as if it is far more expensive, no matter what.

I think they can still get 5% to 10% more than Unity. The strength of Unity could also be perceived as a weakness of Unity, in terms of, if we set a per seat price so low, then it’s very hard to make a strong argument to increase the overall price. There are few areas that Unity can really increase their revenue from the value that they bring to customers. Services is pretty much tied to man hours, so it’s not very, very scalable. Another would be multi-porting. Unity has been favored by game developers for early ages, early stages, for Unity’s multi-porting feature. Multi-porting is, basically, if you build a game for Android’s Google Play market, Unity allows you to port the same game to Apple’s App Store games market. Usually they have different coding languages and you need to recode anything. But Unity helps really shorten that period of time.

If a certain game is made with Unity, it can actually reduce the time to move. Multi-porting is basically free, right now. But for big studios, if you think about how much money they can make, if they port to another marketplace, for them specifically, I think you need to charge a significant price for that.

Does Unity really focus on improving lifetime value by add-on services, but also the Ads network, as well?

Exactly. It’s like Amazon’s flywheel. If every component, every stage of this lifetime cycle strengthens, the entire platform gets much stronger. You mentioned Unity’s Ads network. Although the lower tier of Unity’s customers can leave very soon and the retention rate is quite low, but it is still highly likely that they will adopt and try Unity’s Ads network. It leaves a very long tail of the Ads network, eventually.

In terms of the economics then, back to that example, let’s say you have a studio with $100 million in revenue, 100 Unity developers and they are paying $3,000 each a year. Is that just the subscription cost for the platform and the engine and then the Ads network is separate from that and the Asset Store revenue is also separate?

Correct. I mainly assumed the game engine revenue and other relevant services. Ads is another animal. There’s no apple to apple because Epic then has to include Fortnite revenue and things like that, so it’s very hard to compare apple to apple.

In terms of when you look at lifetime value, do you include any Ads revenue from those studios in that account or is that separate revenue?

It’s a separate revenue because the dynamics are very different. Company-wide, if you look at it, I think there is a very high correlation between the subscription revenue and the Ads network revenue. It’s run separately, but there are a lot of linkages.

Briefly, on customer acquisition costs, how did you look at the cost of acquiring a developer on an on-premise or perpetual license model, versus the free-to-play and subscription world now?

I think a very accurate and detailed lifetime value calculation or projection, allows you to set a certain limit of expenses you can burn, for marketing. Without those references, it’s very hard to make a judgement, in terms of how much you could spend for these campaigns and things like that. Of course, there is no simple answer but I think, going back to recycling the previous example, if you have a seat that generates $3,000 per year, let’s say we can assume the average lifecycle is 2.5 years, per seat, then you know how much money you can actually get. Then you have a certain number of campaigns, events, marketing plans and R&D resources and you can allocate them. The numbers are quite internal, but I think this gives you how much you can spend, per quarter and annually, on acquisition costs.

What was the payback, roughly, on a developer, typically? Let’s take the $3,000 a year developer, at a big studio, how much would you want to pay for that seat, in terms of acquisition costs?

How much would you be willing to pay, to earn $3,000 a year from a seat? It really depends. I think that there is no straight answer. $3,000 is pretty much the willingness to open the wallet; I think it’s pretty much fixed. That’s my assumption. My answer would be, how can I lower the cost? If you calculate that it’s $700 to acquire such a seat, how much can we lower the cost. Going back to the lifetime, flywheel cycle, in itself it is a good marketing tool. It spreads out in the market. From my guesstimation, ball-park range, I think $700 could go down to $300 each.

Obviously, the Unity website has great learning material and videos and content. I guess you were just using traditional channels, in marketing ad spend to drive traffic?

Correct, but I think, as we know, the 80/20 rule is the same thing. Unity’s major revenue comes from the large studios. Maybe it’s 90/10; I don’t know.

What is the risk with these bigger studios, in the engine business?

That’s the context that I mentioned that there’s not a significant risk that I see, in terms of retention rate, because you don’t have other alternatives. They tested out Unity and Unreal for five to 10 years, already, so they know both engines very well and they don’t plan to develop in-house engines, in general.

How have you seen the shift from boxed products to free-to-play impact the development community?

It’s quite simple and obvious that free-to-play games are free, but they’re not free. Somebody has to pay money. Who pays it, usually the ads-engaged parties pay and they pay it back? Having said that, fundamentally, the balance between immersion and the distraction of the game – as the intrusion of ads grew – has become really significant, for the past few years. I think that’s the major shift that I see for free-to-play games. Obviously, the ads have focused to develop where they do not distract so much, in terms of the context of the game flow and the content and nature of the ads.

How have you seen the CPM change, on the Ads network, over the last five years?

That’s also a very interesting topic but I don’t think I’m the right person to share, because I’m not the Ads guy.

Just back to the point on developers; do you think we’re going to see more consolidation in the development community or do you think there are lower barriers to entry, more people can create free-to-play games, but less people can earn actual dollars in revenue from them, because it’s so difficult?

Overall, I’m seeing a huge consolidation. As I mentioned, there will be a more connected approach and less distinction between the devices. For the same game, it could be free-to-play on a smart TV but it could monetize in other ways, on your cell phone. I think, going forward, it will be very flexible, more connected, more integrated. If you think about non-gaming areas, like automative architecture, or any industrial applications, let’s say if those developers start to adopt Unity more and more, usually in industrial applications, we don’t see that free-to-play is kind of given, but nobody knows. It could include another type of Ads network that is non-game. The distinctions and segmentation will be blurred.

Which industry, or other sector outside of gaming, are you most interested in or excited about?

Definitely AEC; architecture, engineering and construction. I think AR can really give huge opportunities for our generation and one of the applications that was amazing in the past year was the growth of AR applications in construction or interior design. You know what would be here, even if it hadn’t been physically built. That gives a whole lot of difference, in terms of marketing but also, construction workers have a much better understanding, in terms of context. If you’re a plumber and you know where this should be built exactly, with AR, that gives context that only very experienced engineers or workers had, previously.

Who does Unity compete with, in that field?

Unreal.

But are there other companies that offer something similar to this, to those end markets? Or is it purely these engines that are actually competing for all these other sectors?

For all the industrial sectors, out of gaming, I think Unity’s main competitor is still Unreal. If you think about it, one good analogy will be gaming. In gaming, I think Unreal’s target customers are pretty much limited; big houses, major titles. Unity, on the other hand, targets a much longer tail. From that perspective I think, outside of gaming, we have pretty much the same situation. If you look at automotive, you very few limited OEMs. They have plenty of engineering capabilities; they have a lot of money. They could focus on developing using Unreal, rather than Unity, if they can guarantee better performance.

But if you think about construction, interior design, we are talking about very small companies through to large companies; a much longer tail. For this segment, there is a limitation in terms of, these people can learn real-time rendering machines. From that perspective, Unity is better positioned to tackle the longer tail. The same thing in film and animation.

How would you compare the Unreal engine and Unity, in those different sectors, like film? They are both used at the same time, so do developers typically use them together, on films or construction?

I think Unity is way better positioned in the industrial spaces. For automotive, honestly, I’m not really sure. I think Unreal has a good position, as well, in automotive. For film and animation, architecture and construction, I think Unity is better positioned for the reasons I mentioned. It’s easier to adopt and learn but also, Unity also developed an industrial template that is specifically designed to merge with existing, particular tools that they are using, in those industries.

For instance, in the architecture space, they are using beam software, so Unity are building templates that layers between the beam software and Unity, so that the beam users can naturally adopt Unity, with less learning. Those kind of layers, bridging those gaps, filling the gaps, I think Unity has done a great job. Very simple, better technology, less learning, less cost wins, eventually. That is really what I expect to see how Unity performs in the industrial spaces.

The last question is just around Epic. Epic have Fortnite, they publish, they have EGS; do you think that those different, and potentially large, revenue streams, can then lead Epic to continue to reduce the cost on Unreal and, potentially, compete more specifically on price, with Unity, which could change the dynamics and market share?

That’s a very good point. A lot of comment that I shared was actually focused on the game engine only; but we are talking about company-wide competition. For that, actually, I am quite in line with you. There are mainly two reasons for that. Firstly, I think Unreal is much better positioned, in terms of growing its business, hitting bigger opportunities, because they are making great games. Their games are very successful. It creates better branding, definitely gives more cash and then, eventually, if you have such a level of success, maybe you have a better idea of how to develop your game engines, going forward.

Secondly – and I think it’s more important – I have a deep respect towards Epic Games, because they have been the winner or the top tier third-party game engine for several decades now. I think they are still very eager to learn and evolve. Whenever Unity published something very nice, in terms of features or different business models, I think Epic watches it and kind of adopts it into their own play. I think that’s the real, fundamental competitive edge that Epic has. From that perspective, if Epic maintains those two characteristics, going forward, I think there is a risk for Unity.

But it’s obviously much more difficult to adopt and that could be the benefit for Unity, in serving the long tail?

Correct. But I think, if Epic decided they wanted to be adopted widely, I think Epic could have done it a decade ago. But they didn’t. So they are staying there. I think it is just a matter of economic sense. As I mentioned, when the industrial market for AR/VR gets so large that it’s bigger than the gaming market, I think there’s a chance that Epic can play just like Unity, to be adopted very easily.