Anti Climactic: Part one

Karl Jeffrey (CEO) and Simon Gardner (head of development) explain…

Develop: You’ve been very quiet for the past couple of years. What have you been up to?

Karl Jeffrey: After we sold Climax Racing in 2006, we had a whole new philosophy. You know the old multi-studio British developer model where everything is heavy iron and scale is all-important? Well, we’d come to the realisation that scale is meaningless without quality, and we’d prefer to make a smaller number of very good games.

Simon Gardner: We had a situation where we were taking in work because we needed keep the machine fed –

Jeffrey: – there were 450 people in-house, on salary – a huge fixed cost base that drives the decision making. We now want to subjugate every decision to the games. The decisions we were making were driven by business and financial needs, not by the singular objective of trying to make the best games.

Develop: How close are you to achieving that?

Jeffrey: I think we’ve come a long way. The company is a lot smaller, with just under 100 people. Over 50% of our man months are done outside of the company, whether by freelancers or classic outsourcing.

The singular objective is now to make better games. Every decision in the company, whether it’s hiring, location, tools, technology, management structure – everything is now subjugated to that single goal. Before, the tail was wagging the dog. You end up being a machine to maintain the employment of development staff, not a machine to make great games.

Gardner: The problem was the publishing world was consolidating, so the options were every decreasing. The client always has to be satisfied with the job you’ve done –

Jeffrey: – but it wasn’t a system that allowed for satisfaction. A non-specific example: we’d have a large team of 40 people about to finish a project, and the publisher would be broadly happy so we’d ask if they were ready to bring on the next one? They’d say, no, they’d want to see how that one sold first. Well, do we carry 40 people for six months while we wait? Just do the maths and it will cost more than the likely profit you’ll make on the game. You have to put those people onto something else.

So you’re not burning costs, you’ll tend to take on things which you know are perhaps poorly set-up: they’ve got insufficient time, the project itself is ill-conceived, or the budget is insufficient, or the publisher… you know there are certain publishers that are good to work with – constructive and helpful – and others that just aren’t so good. And you’re forced because of pure economics to make decisions that you know, with your eyes open, are not going to lead to a better game. So you can end up with a downward spiral – a lower quality game, an unsatisfied client, and so on.

Develop: Why did development studios grow so big in the first place?

Jeffrey: We were one of the first but Warthog, Argonaut, Kuju, you could list them all – the British way of doing things was to get ever bigger. Multiple locations, multiple studios, all the bragging rights were about scale – we’ve got 300 people, we’ve got 500 people, we’ve got 10 projects. You can call it empire building.

But if do an analysis of where the best games are coming from, they’re not coming from those big corporate groups of studios. They’re coming from the very focussed, boutique type studios. And so that’s what we’ve sought to become. Now two years isn’t really enough to deliver on our dream, but we’re a very long way from where we were – to a degree where it’s almost a different company.

Develop: Clearly part of the move to scale came with the rise of highly art intensive games like your own Sudeki project, though. If you take on those 30, 40 or 60 artists again, you’re still going to take on that business risk – unless you outsource?

Jeffrey: Yes, so we outsource hyper-aggressively, and we do it very well. It’s partly about the management structure and partly management tools. We have a Web-based management system that we push out to the outsourcers called Management in Games, and that’s unique, pretty much. It was one of the attractions when Disney took over the studio in Brighton – they took rights to use MiG in all their studios.

It’s also about internal structures, attention to detail, and pre-planning. Let’s be honest, nobody likes outsourcing because it makes the job more difficult, because in some ways you lose some flexibility. But as a business decision we would never go back, having done it.

Develop: What are the main advantages?

Jeffrey: One is financial. You move all these costs out of a fixed overhead to a variable cost, so you can expand and reduce your costs. That means there’s less money wasted on just burn and sustaining people for the hell of it, which goes directly to the game, which feeds quality.

Second is the selection of best in breed. We can go out and shop now. We know what we’re looking for – character models, voiceovers, sound effects, music, concept artwork. In the old model, we would have used a musician or a concept artist on staff. If stylistically they’re not appropriate to the game, what do you do? Do you fire them and hire someone else? Now we check out the best talent in the world.

So it’s not just about going to cheaper places – it’s not just a cost saving exercise, in fact it’s not even primarily a cost-saving exercise, in the sense of your man month rate being lower. It’s about accessing the best of type for a particular requirement. The financial saving really comes from not carrying the fixed overhead when you don’t need it.

Develop: That’s a very different model…

Jeffrey: Once you get past the aspect of organising the company in a certain way and disciplining yourself, it’s incredibly flexible. It allows you to field a large amount of people at short notice, field the right people, and then there are all these cost benefits. I can’t see any other model that’s going to work going forward, even for the big guys.

Develop: How developed is the outsourcing ecosystem to support all this?

Jeffrey: There is now a huge global outsourcing industry. Our database runs to hundreds of companies who we’ve done due diligence on all over Eastern Europe and South East Asia – China specifically. We try to continuously keep in touch. Some get bought and some go out of business.

Gardner: Our outsourcing manager at the moment is over in Lyon at the outsource tradeshow. It’s a networking event for all the outsourcing companies to meet the people responsible for placing the contracts, all in one place. We are pulling together a set of favoured companies.

Jeffrey: For concept artwork, we’ll tend to pick a specific artist, and pay them on a contract basis. If we have a very large amount of art or environment work to be done, we’ll go to an outsourcing company. And then there’s everything in between.

Gardner: For concept art, we might use an individual artist like Craig Mullins. Craig has done work for film studios, Star Wars, he did the Halo stuff. We can now access that talent, for individual set pieces of work. We have a fixed cost – we know we can pay for example $10,000, that’s the cost, we get what we want, and then we move on.

On the other side of the scale we work with another concepting company called Massive Black, a group of artists again based in America. We pay a fixed fee for them to deliver a fixed number of images. It’s all done online, and there’s a direct feed back to them – it’s kind of like having the artists here, but we don’t carry the fixed cost.

Jeffrey: It’s also about technology and facilities. What would the logic be of us having a mocap studio? None at all.

Develop: How do you replace the 30 artists doing the bulk of the in-game artwork?

Gardner: There are companies in Eastern Europe, and Shanghai in China. We set the style with our artists here. We say, “This is the quality of the characters or environments that you must reach. The concept work says this is what it must look like.” And then they have to execute on that contract. We take that artwork back, and we either approve it or we don’t. Then we do a final layer of polishing and lighting to make sure it sits perfectly within the game environment.

Develop: Intellectually this set-up makes sense, it reminds me of the old discussions about game moving towards the Hollywood model. Did those discussions come too soon – was the infrastructure just not in place?

Jeffrey: From memory, we were discussing this 5-10 years ago. We were always talking about it, then thinking, “No, that’s not our way”, and hiring lots of people and lots of offices, and HR departments, and everything that went with it.

And yet, pure economics eventually drove us to it. I’ve used the analogy of what happened in Hollywood. It was the big studios that went to flexible production last. It was the small guys that found it first and pioneered it, and I think it was always going to be the same in our industry.

The big guys can afford it: they have a slate that’s large enough to afford a back lot, stars on payroll, and all the rest of it. As Hollywood fractured, the innovation from flexible working models came from the small guys upwards. It was always going to be EA or the other big publishers who would find religion last. If you’ve got big budgets and big internal teams, it is very convenient to have it all on staff.

Gardner: We were forced to take this route as well because we had no option. A few years ago it was really hard to be an independent developer outside of the publishers. They were all investing internally. The new platforms had come along, and they didn’t want to give independent developers a leg up in terms of technology – they wanted to hold it all for themselves. We’re seeing the consequences now, where they overgrew and now they’re laying those people off, and divesting themselves of that. But that’s just cyclical, it’s always happened.

Jeffrey: I’d say five years ago we started doing it. We started our Flexible Workforce Programme – boring business speak – it was multi-tiered, like the layers of an onion. Right in the centre we have the directors, the key minds behind the business. Then the next layer out, the key individuals – lead artists, designers et cetera. And then another layer of people who are employed on payroll but are on fixed terms or fixed task contracts, so there’s a certain flexibility there. Then a layer of freelance individuals, not companies. And then another layer of relationships with other companies – the classic Shanghai based art-outsourcing company. But you can also say that using other people’s technology, middleware and suchlike, is an outsourcing of the technology requirement.

It’s the past two years that have really proven it – while we’ve been quiet we’ve exclusively followed that model. But we pioneered it under the old heavy iron studio prior to that, four years ago. That was really when we started.

Gardner: I’ve been here five years now and I had to do a presentation to the board after four weeks. I remember saying to Karl, “We have to reduce the size of the studios internally”. I could see like an express train the market was changing.

Sudeki was a large project – it peaked at 75 people. Nowadays that’s nothing, but back then it was a big team. I could see we’d be faced with a problem when we finished Sudeki of what to do with all those people. And so we said this [outsourcing approach] will be the model for the future. We took baby steps along that route, but three years ago we really went for it, because we had to – the market was difficult for us.

Only last week one of our clients was telling us about another developer they’re working with, and the problems they’re having trying to get outsourcing working. And I said, “Well that’s just what we do”, and he said, “You’re three years ahead of them”. We really are. We’re on the fourth or fifth iteration of projects now.

Jeffrey: At this point in time there are more people working for Climax in other companies than there are here.

Gardner: We also knew that repeat business with clients was vital. We’re on our fourth product with Konami in a row, for instance.

Develop: I can see the benefits from your company’s point of view, but how does it work if I’m an individual programmer or an artist?

Jeffrey: The Hollywood analogy doesn’t stretch completely. For that model to work everyone had to be placed in LA, or Bollywood, or wherever. A game – and software generally – can be produced in modules and put back together.

The guys in the office have to be together to put the thing together to design, create it and manage it. We’re in Portsmouth but it’s not hard to attract talent. It’s not Guildford and it’s not Brighton, but we’re getting a good core company here.

Gardner: We’ve made the model work and we’re still here – we’re becoming more successful and the quality of our products is rising. Because we’ve made that leap, we’ll provide more employment for those people, and the more successful we are the more that employment will be ongoing and the opportunities will grow.

Jeffrey: When you divide a project up into its elements, there are certain places you can divide it up into discrete component parts and some where you can’t. The old teleworking idea, where everyone was going to work from home and build software while looking after the kids, I don’t think that quite works, because there are some things that don’t naturally divide, like creating a series of environments. You wouldn’t give an environment to one artist here, and one artist there. You really want to have a production line, saying that takes 30 people to produce those environments and they need to be in one place – they don’t need to be in this place, but they need to be in a place. Concept art, that’s a unitary thing, and it’s low volume as well. A single individual does a single voiceover, that won’t change halfway through.

So innate in the planning there’s a natural way that things are split out. A model where there’s a central hub and loads of freelancers – I don’t think that works. But a model where there’s a central hub and some freelancers, and some people on short-term contracts, and relationships with other companies, that does work.

Gardner: Games development is a creative business, which means there are iterations as well. At some point we have to bring everything together, when it has to be worked on as a single component – as a game – that is balanced, tweaked and changed. Sometimes we’ll get work back from the outsourcers and we’ll change it for gameplay reasons. Those changes will be made with the artists here.

Develop: Karl, you’ve used the term ‘heavy iron’ a couple of times when referring to the old model. Presumably you mean UK development was following sort of old-fashioned metal-bashing factory model before?

Jeffrey: Yeah. Another analogy is the car industry. General Motors would own and produce every piece of technology in their car and every component. Then the Japanese developed these very flexible models based around close relationships. They don’t own the component suppliers and such-like. In that model, you try and keep in the centre the things that make you unique and add value.

Develop: The Coke model.

Jeffrey: Totally like Coke. With its bottlers it’s a totally flexible production model. If they’re doing their job right, they shouldn’t really ever see an actual product. It’s just the marketing campaign and the ‘secret’ ingredient, and the design of the packaging.

Develop: What’s the secret ingredient for a game developer?

Jeffrey: What are our core things? We believe design is a core competency of ours, and something we want to develop further. We’re really investing in design, because that’s a sort of ethereal thing where if you get it right it adds incredible value at very little cost. We think a small company can be better placed to be creative than a large company.

Number two is management competency, including competency in outsourcing. They go hand in hand – outsourcing is not a bolt-on, it’s integral to this model.

Also delivering projects on time and to budget, and not having these scare stories, which is quite unique in development, right? Most companies have scare stories. We’ve had ours but not now.

Gardner: The problem with projects that go into crisis is there are two outcomes: You get lucky and everything’s fine, or the studio goes under. Climax has been around in one form or another for 20 years now, so it must be doing something right.

Check back tomorrow for part two of Develop’s in-depth discussion with Climax.