Is traditional game outsourcing on its death bed?

Streamline Studios has worked on big games such as Street Fighter V, BioShock Infinite, Killer Instinct, Final Fantasy XV, and Pillars of Eternity.

But chances are you’ve never heard of Streamline, which is based in Kuala Lumpur, Malaysia, because it is an outsourcing company, developing technology or parts of games for other developers and publishers. I met with Stefan Baier, the chief product officer at Streamline, at the MIGS 2016 event in Montreal. He gave a talk on “Why outsourcing doesn’t work.” That seemed contradictory for a guy who works at an outsourcing company, but Baier believes there’s a shift happening in the industry that means traditional game outsourcing is on its way out.

Baier believes that games will go the way of other big industries like autos, where game companies will rely on major partners for a lot of their expertise in specialized services as well as flexible development work, while “transactional” outsourcing is likely to suffer. Transactional outsourcing is where the main objective is to do a simpler task without expertise, where the main benefit of outsourcing is lower costs.

Here’s an edited transcript of our conversation.

GamesBeat: Your topic was interesting. Outsourcing isn’t an easy thing to do.

Stefan Baier: No. Part of it is because the majority of the industry—it’s still a place with a lot of room for improvement. The talk had kind of a provocative title in “Why Outsourcing Doesn’t Work.” What I mean is that a particular kind of outsourcing doesn’t work, and it’s going away at some point in the near future. It’s this transactional outsourcing, throw it over to someone in a short-term relationship. Games are becoming so complex that you want to retain that knowledge in your company and in the relationships you have.

This like what you see in other industries. If you look at manufacturing a car, a car has maybe 30,000 parts, but you don’t have 30,000 outsourcers. Large companies find three, four, five trusted companies to work with for 10, 20, 30 years. That’s where the industry is going. It’ll be more about expert partners, relationships that go deep and are deeply integrated. You’ll have trusted relationships with people who have engine access. You’ll strategize together. You’ll talk about where to take the technology together. You’ll talk about where the partnership goes.

With our company, we have 200 people. We’ve expanded very rapidly, especially selecting our own partners, our own clients based on this idea. If people are interested in building these kinds of long-term company partnerships, that’s our model in the future.

GamesBeat: What’s your own company’s history and experience with outsourcing?

Baier: We’ve been around for 15 years. We started out in Amsterdam, then established ourselves in Malaysia in 2010. Our headquarters are still in Malaysia, but we also have people in Las Vegas, Tokyo, Frankfurt.

We’re at the model we wanted to have from the beginning. When we started in 2001, it was more of a transactional relationship. We were modeling, texturing, doing services for other companies. Then, over time, we began to retain more expertise from the partners we worked with.

That’s one reason we were able to build long-term relationships with Epic Games. We worked with them on Gears of War, Unreal Tournament, those kinds of games. We worked on Terminator Salvation, on James Cameron’s Avatar, on those productions. We’re doing a lot of work with Capcom now. We had a joint press conference with Square Enix in Malaysia together with the Malaysian government a couple of weeks ago, discussing some of the partnerships we’re having.

A lot companies out there — big companies, ambitious companies – recognize for their own future that they need to identify other companies that think like them, that are philosophically aligned with them, that think about value-based relationships and not so much transactional relationships. That’s where we’re coming in.

We’re based in Malaysia, but it’s a very international crowd. We have about 25 nationalities from all around the world, attracting the kind of senior talent we can bring together. We’re focusing specifically on triple-A game productions. That’s always been our specialty.

GamesBeat: Some of the companies I’ve written about in the space do seem like they have very transactional businesses. Dhruva Interactive out of India, they made a lot of the cars for Forza, a lot of Microsoft games. That seems to make sense as a transactional thing. But a lot of games aren’t like that.

Baier: When you’re looking at why certain things are outsourced, something like cars makes sense, because with a few simple instructions you can say, “Make me 50 of these.” The plans are very straightforward. But when you look at the games it leads to—the games that embrace outsourcing the most are sequels, like the Call of Duty games, games that come out every year. For that the transactional model is very successful, because you don’t need to reinvent the wheel every time.

For products that have more of a creative angle, that are more ambitious or try to do things differently, things get more complicated. As any developer knows, that’s why you use a scrum methodology internally. You try something, it doesn’t quite work, and you iterate on it to try it again. Development is difficult.

The scrum method is very different from the transactional model. It’s almost like a waterfall method, following a linear plan. It’s not very compatible with scrum methodology. What that means for us is that the projects we see are more integrated across all the different departments between the two companies. We’ll have the preproduction team on our side talk to the art directors and concept team on their side. At the same time, we’ll be working on prototypes, level design, things like that.

As you do that, you build knowledge in both companies, both partners, about what it’s going to be like once production begins. You’re part of preproduction, the conceptual aspects of the game. You’re part of full production. You’re part of finishing the game, and even DLC, entire aspects of the game afterward. It’s possible to do that transactionally, but honestly, games are getting more and more complex. As a company you need to decide where you want to stand in the value chain. It’s not easy to do this work, but that’s been the trend in the industry, the demand.

When you talk to developers, nobody wants to deal with micromanaging other companies or dealing with 10 or 20 different vendors. They’d prefer to have trusted partners they can always work with. That’s one of the key reasons we’ve seen so much response to our model. A few companies working together for the long run.

It takes time, because you get to know each other on a different level. There are more conversations between different departments. Generally, we want to have the right people talking to each other – art directors, technology directors, all these guys have to talk to each other between the two companies. It can’t be funneled just through one outsource manager. One person can’t handle the interactions between three to five departments.

GamesBeat: A lot of people believe that outsourcing is inevitable, regardless of the form it takes. Do you agree with that? Or do you think a certain kind of outsourcing shouldn’t exist?

Baier: Well, what does outsourcing really mean for most people? It’s this idea of sending something to some unknown overseas location, taking away jobs, this otherness. It’s strange for us to think of it that way, though, because our company—we’re based in Kuala Lumpur, but we have many nationalities working together, and we’re just making games for the global market.

The whole is more than the sum of the parts. If you have companies that are experts in certain fields collaborating together, they’ll achieve more than a single company trying to do everything by itself. I don’t know if you want to call it outsourcing, or if you just want to call it focusing on core strengths and leveraging other companies that think like you and complement your strengths. I don’t know if that’s inevitably, but I do think it’s good business to invest in that. But it requires a business strategy to say, “I’m going to invest as a company in long-term relationships with other companies.”

GamesBeat: For some companies, it makes sense that some jobs are temporary, right? Does a main developer really need a motion-capture stage, say, for the lifetime of their studio?

Baier: Exactly. Looking around here, for example, there are some other companies that no one knows about, but they’ve been in the business for five or 10 years and become experts at something. A group over there is very good at backend architecture for games. They’ve built for very large companies. It’s another example where someone has become experts in their field, and every time they do this for a company, they get better at it. That’s who you want to work with, who you want to have a relationship with extending over years. You want them to be successful, because you’ll benefit as well.

To me that’s different from ordering a couple of props. This is working with experts in their field. It’s similar for us. We often educate or collaborate with our partners now and they’ll ask us, “How would you do this? Have you seen this done? How have you solved this problem before?” That comes about because you can’t be an expert at anything. You don’t need the hardware for everything. You need to focus in on where you feel your strengths are as a company.

GamesBeat: You’re also arguing for more expertise, as opposed to simply low-cost alternatives.

Baier: That’s exactly where it’s going. Of course you can always get it cheaper, but this is about driving value. This is not about going out and trying to save a penny. You need to understand that working with another company, whichever way you turn it, is a significant investment in time and internal staff. All the hidden costs that come with that, you need to account for them. You need to know what you want as a return. Any large company, there’s only so much bandwidth for building partnerships. You need to wisely choose the ones that will bring your overall value up the most.

If I were a company and I had to choose, I would choose partners that might not be the cheapest, but they’ll be that perfect balance between the expertise I get access to, the value and balance they bring, and the stability they add, instead of going for the cheapest option of the day, or the hyped name of the day. That doesn’t work. There’s too much rotation.

Games are difficult. You need to keep everything as stable as you can. We see this with the people we work with. They can afford to pay for the right company. What they can’t afford is to have a $100 million launch go wrong. That happens when you go with the company that’s underprepared. If the man-month rate is $5,000 or $10,000, that still doesn’t sound like a lot of money compared to a $100 million mistake.

GamesBeat: If you put a timeline on this, how recent would you say the thinking is?

Baier: We’ve been operating in this model in some sense or another for 10 years already. 2006 or so, this started to come up. Educating people on how normal maps work, taking that whole generation forward, working with Epic and then broadening that horizon. But it’s really the last few years where it’s become dominant for us.

It’s also being able to choose our partners more purposefully. We do care about having the Perforce engine integration, having the builds attached, having full access to it. The last five years have seen that model take off for us in a strong way. At this point we’ve become essentially exclusive.

It’s still an education thing sometimes. People come to us and say, “Oh, are you making props?” Well, we can, but generally we do that as part of a bigger goal-oriented project. We usually have milestones that are less about making 10 cars and more about, “What are you trying to achieve in this chapter of the game? What are your pain points internally at the main company? How can we help with that?” That’s why we have an engineering team, content team, animation team, design team, all working together in one location. That’s another choice we made. We didn’t split the team up into different locations.

GamesBeat: Is this keeping you competitive with other large outsourcing companies?

Baier: We don’t really consider ourselves part of that space. We don’t compete on price or volume. Someone else can make you 50 cars in a month. We don’t compete with that. We ask, “How can we partner with you and help you be successful?” Whatever that means, it tends to be a long-term answer. Something that takes investment on both sides in getting to know each other. It means traveling. It means talking to each other and finding weaknesses on both sides. It’s people.

We try to help build those communities. In Malaysia there’s a conference called Level Up KL. The Malaysian government is very supportive of the industry. We’ve helped put on that event as well. Our CEO gave a keynote mentioning the other companies in our space. “These are the local heroes in Malaysia.” We want them to be successful as well. We’re all facing the same challenges.

GamesBeat: What are some things about games that you think drive production? How do you deal with repetitive tasks? Procedural generation has made some tasks easier.

Baier: Certain types of games will lend themselves more easily to mass production. Replicating a city, in some sense, is easy. You just take pictures of the real city and make this many trash cans, this many brownstones, and you’re off to the races.

When you’re dealing with a more complex overlap of technology and creativity, it gets harder. Let’s say you’re making a fantasy city. You need to spend more time in preproduction. What does this world actually look like? How does everything function? How do you drive gameplay ideas into that? Then you have to constantly navigate between the technology that’s being built, which has particular needs, and what you build with it. You also have gameplay requirements which are often shifting in the first years of production, and the creative and content production aspects.

All of these factors rotate around. When you look at mass production, they usually wait until they’ve passed through most of that, and then there’s one year or less of insane production. That’s when they go to China and say, “Give me 300 people to pump through this stuff.” But that’s a painful process. It all happens in a short amount of time, and it burdens the main developers with parsing through 300 man-years of work before launch.

We avoid this with our partners because we work with them from the very beginning. We can help solve these problems, so it never piles up in the same way. It depends on the product you’re making, the ambitions involved. We tend to work on complicated projects – either creatively complicated, technologically complicated, or both. That’s where our strengths come out. We tend not to work on sequels where all the problems are solved. If it’s Unreal engine, sequel number four, you’ve built the city before but it’s in HD now, you don’t need me. But for pushing the creativity angle forward and keeping it stable, that’s where we tend to come in.

GamesBeat: I still see production problems with different games. Mafia III was a good example. They spent five or six years on it, but the last couple were the most focused for them. They raced to the finish line and had a very good story and all that, but players complained about bugs.

Baier: One thing that happens there is that they underestimate the importance of involving partners early and giving access to the technology. I don’t know anything about the Mafia engine, but—open-world engines, they’re tough to deal with sometimes. They’re usually built internally. The teams building them never think about what happens if they move that technology to a partner. We’re experienced with that now, but there’s a whole process at our IT team, how they work with other IT teams and take this technology somewhere else, make sure the plug-ins and everything work.

The earlier we begin to do that, the more we avoid those issues. Because we have the technology on site – most of them don’t do that – we can solve bugs as they appear on our side, instead of having to wait six months until they have time to implement on their side. These production problems are a key reason why we tend to get involved early.

The way we put it is, “people, process, technology.” It’s about people. You need to travel to each other. You need to have the right people talking to each other. Process, there needs to be a regular understanding of how we deal with each other. Technology, both getting their technology on site and our technology to them as well. We have our own management tools, Stream Frame, a web-based tool for managing our internal teams and our interactions with other companies. That allows us to keep everyone on the same page, even if it’s a very large production.

Again, you need to invest in your own technology. If you challenge your partner to invest in making their tools more compatible for off-site use, it’s only fair that we do the same thing with our technology too. That costs money, but you have to spend that money. We don’t compete on price, but we reinvest our money internally to build a broader and more balanced experience for our partners. That’s where we see the future going.

GamesBeat: How much are some of the other tool vendors helping? Unity’s whole pitch is that you can run your game on any platform. They take over some of the work of porting, which simplifies a task outsourcers used to do.

Baier: It definitely does. It’s allowed us to move the conversation a bit from the technical side to the creative side. You see more companies venturing into making games. The Unreal engine, Lumberyard, Unity, that’s all very good for the industry as a whole. It establishes a more common baseline for everyone.

About half of what we do is still custom engines, especially the bigger and more complicated stuff. We’re attracted to complex projects. But we’re seeing the conversation shift when you talk about, say, an Unreal game. Porting comes up less. It’s more about, “Can you take over a part of this game?” Because everyone is so familiar with Unreal, it’s easy to do that. We also develop our own internal games with Unreal.

The porting conversation goes away, but other places where help is needed come up very quickly. Especially now, with games as a service, it’s less about porting and more about season two, season three, thinking in the long term. That plays into a model where it’s all about understanding how you’re thinking today as a partner. We’ll be able to anticipate and work with you two or three years down the road. A lot of these companies really do plan their content that far out, even after launch, if they see there’s traction.

GamesBeat: What are you working on now that’s filling out your ambitions?

Baier: We can’t talk too much about that. What I can say, this press release we had with Square Enix a couple weeks back—obviously we’re a long-term partner with Capcom, as well as others. We see a model where, in simple terms, a lot of developers are just tired of having to micromanage all their partners.

A lot of conversations we have with new companies, we hear the same thing. “We’ve used these outsourcers for a long time, but it’s exhausting. We’re looking for new options to try.” Then we tell these guys how we do stuff, and it sounds great to them. They can hand us this whole chapter, this whole responsibility, and we can run with it. Many teams are not as big as they used to be. They’re so preoccupied with a certain aspect of the game that they’re happy to say, “You guys can handle multiplayer. You guys can handle chapter two.”

It helps with attracting and retaining senior talent, too. There’s a greater sense of ownership in the work. If you’re there and assisting from preproduction to launch, that’s a huge sense of being part of a project. It doesn’t just feel like a black box to us.

A lot of the stuff we’re doing right now is in development for the next few years. But the relationship with Square Enix is a good example. They’re quite forward-thinking. At the press conference, they sent one of their lead designers to Malaysia to speak on our behalf. Those kinds of partnerships, you’ll see more of them. People understand that games are complicated and you need to find partners to work together.

More companies are emerging. Some of the larger game developers and publishers are trying to consolidate. Some people are driving this idea that you can only outsource lower-value work and only the experts inside the publishers could manage outside companies. I beg to differ. There are some companies that might be comfortable with that, specializing in that kind of work. But many companies, in North America and South America and eastern Europe and elsewhere, are saying, “We’re outsourcing, but we’re also experts. We can do much more.”

The trend is going to be more and more obvious. There are companies like this right outside there. They might not be very branded. But I’d call them the industry’s best-kept secret, the experts who actually get the work done. They’re just not on the box. It’s a crucial change that I hope will come in the next three years. Other companies will do what we do and say, “We have an identity. We can be part of the conversation.”

GamesBeat: Is virtual reality much different? I see tool companies coming along everywhere to supply different pieces of VR, like eye-tracking. It’s like your analogy with car companies. There are 30,000 parts in VR, and you’re not going to be able to master making all of them.

Baier: You see the seasonality. A lot of companies come up. There’s the “app for everything” syndrome. That’s not what I mean. Rather, I see this phase passing. There are lots of suppliers now, and sometimes it becomes overwhelming. Eventually one of these tool-makers will establish domain expertise all across the vertical. There’s a wave of excitement in VR with 10,000 companies or whatever, but by the end of it there will be just a few expert suppliers who are very good at certain things and they’ll stick around in the long run.

Some people at publishers look at all these companies and think they need to put a lot of effort into managing it, but it’ll manage itself. Experts will emerge, and they should be recognized and given a place at the table as the experts in their space. VR is just the latest example.

Disclosure: The organizers of MIGS 2016 paid my way to Montreal. Our coverage remains objective.