What I Learned Building CGI Studios - The Bit Nobody Talks About.
- Richard Wright
- Apr 29
- 6 min read
There are two kinds of knowledge in this industry. There's the knowledge you can learn - the software, the pipelines, the techniques, the workflows. And there's the knowledge you can only earn - through years of pitching work you weren't sure you could deliver, hiring people you had to trust completely, and sitting in rooms with clients whose expectations were enormous and whose patience was finite.
I've spent over two decades earning the second kind. And I want to share some of it here - not as a retrospective, but because I think the lessons are more relevant now than they've ever been.
Building Something from Nothing
Running a CGI studio is not glamorous in the way people imagine. From the outside, it looks like a world of beautiful imagery, cutting-edge technology, and creative freedom. And those things are real. But the inside of a studio - the actual daily reality of building and leading one- is a relentless negotiation between creative ambition and commercial survival.
You are always managing tension. The tension between what a client wants and what they've budgeted for. Between the time a project deserves and the deadline it has been given. Between the technology you wish you had and the technology you can afford. Between the talent you want to hire and the talent you can retain. Between the work you want to make and the work that pays the bills.
The leaders who thrive in this environment are not the ones who eliminate that tension. They're the ones who learn to work within it productively - who understand that constraint, handled well, is often the engine of the best creative work.But there is one tension I've learned, above all others, is non-negotiable. And it sits at the heart of every technology transition I've ever navigated.
The Floor You Cannot Drop Below
When CGI established itself as the primary tool for product visualisation, something fundamental shifted in the relationship between studios and clients. Photorealism stopped being a premium offering and became the baseline expectation. Clients weren't paying for images that looked almost real. They were paying for images that were indistinguishable from photography - that had the weight, the texture, the light, the material truth of the physical product they were selling.
Once that standard was set, it could not be unset.
This matters more than most studio conversations acknowledge, because every significant technology transition brings with it a period of compromise. Real-time rendering, when it first arrived at a quality level suitable for client work, was not as good as the best offline renders. AI-generated imagery, today, is not consistently at the level of the best studio output. Early adopters of any new tool accept a quality trade-off in exchange for speed, cost, or capability gains elsewhere.
That trade-off is reasonable in some contexts. It is fatal in others.
The discipline I've carried through every transition is this: never let the adoption of a new process or technology become a reason to deliver work below the visual standard the client has been promised. The client does not care that you're pioneering a new pipeline. They care that the image on the page, the animation on the screen, the experience in the headset, looks extraordinary. If it doesn't, you don't get a second chance to explain why.
This sounds obvious. It is surprisingly easy to get wrong.
How quality gets eroded
Quality doesn't usually collapse suddenly. It erodes gradually, through a series of small, individually defensible decisions.
A new tool gets introduced mid-project because it promises to save time. The team is still learning it, so certain things aren't quite right, but the deadline is close and it's close enough. A client brief is interpreted slightly more loosely than it should be, because the technically correct interpretation would be expensive. A render gets approved at 95% of what it should be, because 100% would require another day and the schedule won't bend.
Each of these decisions has a rationale. Collectively, they establish a new normal - one where 95% is acceptable, then 90%, then whatever the current pressure demands.
I've seen this happen to studios that were genuinely excellent. The work that built their reputation was produced under conditions that no longer exist - when they were smaller, when every project was personal, when the founder was across every frame. Scale, speed and commercial pressure introduced distance, and distance introduced compromise.
The antidote is not to stay small.
It's to be deliberate and relentless about the standards that cannot move, regardless of scale or pressure.
The visual quality floor is one of them. It is, in fact, the most important one - because it is the thing clients can see.
Evolving without compromise
During the real-time rendering transition, I learned this the hard way - and then, I hope, got it right.
The promise of Unreal Engine and real-time workflows was transformative: interactive experiences, configurators, immersive environments that no offline pipeline could produce. The capability gain was genuine and significant. But early real-time output had limitations. Materials behaved differently. Lighting was harder to control. Certain levels of detail that offline rendering handled elegantly required significant extra work in real-time environments.
The temptation, and I watched some studios give in to it, was to let the technology set the quality standard. To present real-time output to clients and allow its limitations to become the new definition of acceptable. It was often those studios that had come from the games industry, or were more about code than creative, that made this mistake.
We didn't do that. The standard we held was simple: if the client cannot see the difference between this and what they'd expect from our best offline work, we move forward. If they can see the difference, we don't - not until we've solved it. That meant some projects took longer. It meant some pitches we didn't win, because we wouldn't promise what we couldn't deliver at the level we required of ourselves. It also meant that when real-time output did reach the quality bar, our clients trusted it completely, because we hadn't spent the preceding two years training them to accept less.
The same discipline applies now, as AI tools enter the workflow. There are applications where AI assistance genuinely enhances quality in ideation, in texture generation, in certain kinds of environment building. There are applications where, right now, it introduces artefacts, inconsistencies, and a particular kind of uncanny smoothness that a trained eye - and often an untrained one - immediately recognises as wrong. The job of studio leadership is to know the difference, and to ensure that the adoption of new tools never outpaces the quality they can reliably deliver.
The commercial logic of uncompromising standards
This isn't just aesthetics. It's business strategy.
The CGI industry is not a commodity market, though some clients, and some studios treat it as one. The studios that compete on price alone find themselves in a race they cannot win, there will always be someone cheaper, somewhere, willing to produce something adequate for less. The studios that compete on quality, on the reliability of their output, on the trust that a client places in them to make something extraordinary - those studios build relationships that are genuinely hard to displace.
Every significant client relationship I've built over twenty-five years has been built on one thing above all others: the certainty, on the client's part, that the work will be outstanding. Not good. Not competitive. Outstanding. That certainty is not built by individual brilliant projects; it's built by the consistent absence of failure. By the knowledge that when a brief lands, it will come back as something better than expected, not something that had to be negotiated down to acceptable.
That reputation takes years to build and can be damaged very quickly. A single project that falls below standard -for whatever reason, however understandable, creates doubt. And doubt, in a client relationship, is extraordinarily difficult to repair.
What great studio leadership actually is
I've thought a lot about what separates the studio leaders who build something lasting from those who don't. It isn't creative genius, though that helps. It isn't technical mastery, though that matters. It isn't even commercial acumen, though without it nothing survives.
It's the ability to hold two things simultaneously: a genuine, uncompromising commitment to the quality of the work, and the practical intelligence to build the processes, the teams, and the culture that make that quality achievable at scale, under pressure, and through the disruptions that will always come.
Those disruptions are not the enemy. They are the test. Every time a new technology arrives and promises to change everything, the question it's really asking is: do you know what you stand for? Do you know which standards are absolute and which are adjustable? Do you have the discipline to evolve your methods without surrendering your quality?
The studios - and the leaders - who can answer yes to those questions are the ones still standing when the dust settles.
I've tried to be one of them. I intend to keep trying.
By Richard Wright