Creative Operations
March 2026
The Building Block System
The ninety percent gain: how we architected marketing inside a production pipeline that was never designed to support it. Seven award seasons. Six films. One system.
90%
Reduction in content delivery cycle
48→6h
Turnaround per asset request
7
Award seasons served by the system
Core 20%
Assets that drive 80% of global marketing impact
If you wait until the end credits of The Boxtrolls, there is a scene that does something none of the film's marketing could do as efficiently as those two minutes of footage.
Travis Knight — President, CEO, producer, and lead animator on the film — is on stage, animating two of his characters. The characters are having a conversation about whether they might be puppets, controlled by external forces. As they talk, the camera pulls back. You see the stage. The rigging. The camera equipment. The animation capture system. The rapid-prototype replacement faces in their trays. And Travis himself, one hand on each puppet, moving them frame by frame.
The whole production — the months of planning, the hundreds of steps, the unreasonable commitment to physical craft at a scale most people's careers will never approach — is visible in that single pull-back shot. And audiences feel the weight of it without being able to explain why.
That scene took months to plan. It required Travis to animate himself into the film he had just spent years making. And it worked because it showed rather than told — it gave the audience the "Santa's workshop" feeling that had always been LAIKA's greatest asset, without once explaining it to them.
In the early years of the studio's social and marketing operation, capturing a moment like that would have been an enormous ask. The planning, the setup, the disruption to an animator mid-shot — none of it could be justified without a clear framework for how and why that footage might eventually be used. And so the crew member finished the work. The shot was completed. Production moved on. The moment was gone — existing only in whatever footage happened to have been taken while it was still happening.
That is the problem the Building Block System was built to solve.
The pull-back shot that showed more about stop-motion production than any behind-the-scenes package we could have made. Travis Knight animating his own characters in the film's post credit scene. ©LAIKA LLC
01
The Nature of an Irreversible Medium
Stop-motion animation is, at its core, a medium of irreversibility.
A puppet's performance, once captured, is rarely revisited. A set, once struck, is gone — the physical reality of stop-motion production means a set cannot stay up for the duration of a shoot. It is dismantled, the materials stored where possible, the moment existing only in whatever footage was taken while it was still standing. An animator spends three weeks moving a character one frame at a time to produce ten seconds of footage. That specific performance, in that specific light, by those specific hands, could theoretically be recreated — but the production schedule is planned with meticulous precision, and any unplanned request carries real cost. A surprise addition to the call sheet can mean a delayed shot, a delayed sequence, a delayed end date, and additional shoot days that nobody budgeted for. There always has to be a compelling reason to go back. LAIKA produces four to six seconds of usable footage per week. Everything that happens inside that process happens exactly once — not because it is physically impossible to repeat, but because repeating it is a cost the production cannot absorb without a very good reason.
This creates a structural tension with marketing. In most industries, marketing happens after production. You finish the thing. Then you sell it. The content needed to sell it can be generated after the fact — photoshoots, interviews, demonstrations. The product still exists. You can go back to it.
In stop-motion, going back is possible — but the cost is rarely justifiable without a compelling reason built into the plan from the start. In high-fidelity production, marketing has traditionally been treated as an extraction process. You finish the work, then you go back in to mine for assets. At LAIKA, in the early years, this extraction came with a 48-hour delivery window — if the request was achievable at all. Every request for a creative review or a global asset meant a two-day lag to find, prepare, export, and deliver. Global licensees, distribution partners, awards campaign managers — everyone was in the same queue.
"The question was never whether we could go back. It was whether we could afford to. The system changed that calculation."
02
The Wrong Question
For a long time, I was asking the wrong question.
It started on Corpse Bride at Warner Bros. — the first production where marketing requests began arriving while the film was still being made. The requests were reasonable. The answers were difficult. Where was the documentation of this fabrication process? Could we pull BTS footage of that specific sequence? Did we have reference material for the colour pipeline work? Some of it existed. More of it didn't. The decision about what to capture had been made by whoever was present at the time, with no framework guiding it.
The question I was asking then was: how do we get what we need faster? It was a logistics problem, and I was treating it as a logistics problem. Better file management. Cleaner folder structures. Faster processing pipelines. These things helped at the margins. The fundamental problem remained.
The fundamental problem wasn't retrieval. It was perspective.
The moment that crystallised this came during the Coraline production at LAIKA. A licensee asked for production reference material — photographs and documentation of how specific props and set pieces had been built, to inform a licensed product range. We gave them what we had. It wasn't enough. And sitting with that, I realised the real issue wasn't that we'd failed to store something correctly. It was that nobody had thought to capture it in the first place — because we'd been capturing what the studio found impressive, not what a licensee, a distributor, or a marketing partner would eventually need.
That's the same tension at the heart of how LAIKA approached its social content in the early years — showing what the studio found extraordinary rather than what an audience would respond to. The Building Block System was the operational answer to the same underlying problem: we needed to stop thinking like a production and start thinking like the people downstream of it. To anticipate the requests before they arrived. To understand what marketers, licensees, and distribution partners would need — and build the process around capturing that, not around what felt most significant from inside the studio.
That reframe is the entire foundation of the Building Block System.
03
Distil: Finding What Actually Matters
The first step wasn't technical. It was analytical.
We stopped trying to react to every bespoke request and started mapping the pattern behind them. Across multiple productions, the content that drove genuine commercial impact clustered around a surprisingly small set of asset types: EPKs, production assets, style guides, character materials, key behind-the-scenes footage. Everything else — the answers to every possible press question, the B-roll for every conceivable editorial context, the assets for every regional market's specific requirements — was genuinely endless.
The Core 20% wasn't a compromise. It was a discipline. By naming what was essential, we could stop trying to capture everything and start planning specifically for what mattered. Each asset type was assigned a semantic tag based on its likely downstream use at the point of capture: awards campaign, licensee style guide, distributor toolkit, social content, press materials.
This tagging happened before any request arrived — which meant when a request did come in, the answer wasn't a search, it was a look-up. The metadata was already there. The asset was already organised. The only question was whether the rights were cleared and the format was right. That shift — from reactive search to proactive tagging — changed the nature of every downstream conversation.
04
Plan: The Marketing Wish List
The hardest part of the Building Block System wasn't technical. It was cultural.
We stopped asking for assets after the sets were struck. Instead, we went through the script and watched animatics as they became available, identifying key elements of the story that would need coverage. We met with department heads and asked directly what areas they expected would be the hardest problems to solve — because those were usually the moments the audience would find most extraordinary, and most worth capturing.
We repeated this process at multiple points in production to determine two things: what the hero plot points were that might need additional marketing coverage, and what the behind-the-scenes challenges were that best illustrated the production's commitment to craft. That, paired with the standard marketing material requirements, formed the Marketing Wish List — and we worked it directly into the shooting schedule.
We captured timelapse of an animator spending two months on a complicated shot — Norman walking down the street talking to ghosts in ParaNorman. We documented the complete fabrication journey of a character on The Boxtrolls, from concept art through to finished puppet. These were not afterthoughts. They were planned modules, scheduled alongside production, captured while the moment was still alive.
This required deep diplomacy and a shared language with the filmmakers. The production process was primary — the films had to be made — and every request to capture something during production came with a real cost in time. The Wish List had to earn its place in the schedule. That required being honest about the value of what we were asking for, and being willing to accept "not this production" when the cost genuinely outweighed the benefit.
The Giant Skeleton from Kubo and the Two Strings — at the time of production, the largest stop-motion puppet ever built. Captured as a planned module on the Marketing Wish List. When the Oscar campaign needed it, the footage was already there. ©LAIKA LLC
The Kubo and the Two Strings Giant Skeleton sequence is the clearest example of the system working exactly as designed. The Giant Skeleton was, at the time, the largest stop-motion animated puppet ever built. How we animated it — the rigging, the scale, the technical problem-solving — was the kind of behind-the-scenes story that could support an awards campaign, a distributor toolkit, a press package, and a social content series simultaneously.
Under the Building Block framework, the capture of that process was pre-scheduled as a module. It was on the Wish List. It was in the production plan. When the BAFTA nominations were announced for Kubo, when the Oscar campaign was running, when a distributor needed contextual production footage — the Giant Skeleton material was already there. Tagged. Processed. Ready.
"We didn't search for content. We unlocked it."
The discipline was in building the shared language with production that made the Wish List possible. Not imposing a marketing framework on a creative process, but demonstrating — consistently, across multiple productions — that the "sell" being built alongside the "make" made the "make" more visible in the world, not less.
05
Construct: From Stage to Licensee in Hours
Data logic is the only way to scale craft. We were always going to be dealing with a large volume of assets and requests arriving simultaneously from multiple sources — editorial, marketing, awards, licensing, distribution partners across twenty territories. It was impossible for any one person to hold all of that knowledge, and I had no interest in being the bottleneck who had all the answers.
The system started on stage, at the beginning of each shot, with a digital slate. Before a single frame was animated, the team filled in the relevant metadata: which animator was working, the camera team assigned, the AD responsible, and any specific frame usage notes — "only use frames 1, 7 and 9," for example. At the end of the shot, notes were added. All of that data was stored in blob files attached to the shot record.
Once a shot was submitted, the render farm took over automatically. From a single submission it generated everything simultaneously: MXF files with all metadata embedded for the editorial team, projection files in the correct colourspace for the screening room, low-resolution videos logged into ShotGrid so any team member could review and add notes without needing access to the full pipeline. Multiple outputs, multiple destinations, one submission. On Coraline, that process was either impossible, manual, or incredibly time-draining. By ParaNorman it was standard.
The conform side of the system handled the downstream delivery problem. When a distributor sent in a trailer for review, editorial would match cut it and export a shot list from Avid. That file went into a conform tool we wrote in-house. By selection it could generate whatever was needed: a high-resolution clean QuickTime for sharing, a conformed image sequence for the screening room in 2D or 3D, a ShotGrid-logged version for notes and approvals. One file in. Multiple delivery formats out. The choice of outputs was a checkbox, not a conversation.
The directory architecture underpinning all of this was designed to be humanly navigable as well as machine-readable. Three to four levels of organisation: movie directory, category, shot name, version. The art department working on concept drawings would select environment, location, and artist — and their files were published and stored accordingly. The rapid prototype department could file by sequence and shot, or by character and action, depending on what they were working on. The flexibility matched how each department already thought about their work.
That last point was the key to adoption. I worked with every department to understand how they naturally named and organised their own files — what they called things, where they instinctively saved them. The publish tool formalised that natural behaviour rather than replacing it. Once teams realised how easily they could share, find, or present their work through the system, adoption followed without requiring it to be enforced.
By the time we reached Missing Link — the fifth LAIKA production — everyone knew how to use the tool, because it was essentially the same tool across all departments. A set dresser could shoot test frames on stage, route them through the pipeline, and have them available for review in editorial without needing a specialist team to manage the process. There were steps and protocols, but the beauty of the system was that it was simple enough for anyone to use. The bottleneck I had been in the early years had been replaced by a studio-wide shared language.
48→6h
The delivery cycle for creative review — before and after the system. From a distributor's request to assets in their hands. Fast enough to respond to a BAFTA nomination announcement the same morning it breaks.
Across multiple productions, the system grew from supporting a handful of core campaign needs into the bedrock of almost every initiative the studio ran. The modular, pre-tagged asset library served LAIKA's own awards campaigns and the studio distributor's campaigns simultaneously — but it also supported regional distribution partners around the world, licensees building consumer products, press and publicity teams, and internal presentations when the studio was pitching the next film to potential partners. The same assets, captured once and organised correctly, could serve all of it without rebuilding from scratch for each new request.
Each production made the next one faster. The tagging schemas improved. The Wish List became more precise. The relationships with production deepened. By the time the system was fully mature, what had required significant effort and negotiation on the first production was running almost invisibly — part of how the studio worked rather than something imposed on top of it.
That kind of compounding is what separates a system from a workaround.
06
The Conversation the System Made Possible
Stop-motion production is genuinely hard for people outside it to calibrate. Requests that would take another film's marketing team an afternoon to turn around can take a stop-motion studio three months to fulfil — not because of organisational inefficiency, but because the physical reality of the medium makes certain things impossible after the fact. A simple request for a character in a new pose, interacting with a new object, in a new environment — in any other medium, that's a few hours of work. In stop-motion, it may require a puppet that no longer exists, a set that has been dismantled, and an animator whose schedule has moved on to the next production.
For years, the honest answer to many marketing requests was either no or "maybe, but it will take three months." Neither answer serves the campaign. The first closes the conversation. The second undermines confidence in the operation.
The Building Block System changed that answer. By considering what was consistently needed across productions, applying experience about what would be requested before it was requested, and building a baseline of prepared content that covered the most common requirements — the answer became something different. Not exactly what you asked for, but I have this, this, and this. Which of these serves your campaign best?
That shift, from no to here are your options, changed the nature of every marketing conversation. What had previously been a difficult negotiation about constraints became, more often than not, an energised discussion about possibilities. Distribution partners who expected to be told what couldn't be done were instead presented with a set of prepared alternatives — filmmaker-reviewed, production-approved, immediately available.
That last detail matters more than it might appear. The content wasn't just available — it had been seen and signed off by the filmmakers themselves. In an awards context, where authenticity and creative integrity are everything, filmmaker-endorsed materials carry a weight that no amount of post-production marketing content can replicate. The campaign wasn't just fast. It was trusted.
"Not exactly what you asked for. But I have this, this, and this — reviewed and approved by the filmmakers. Which serves your campaign best?"
When It Didn't Work
07
I want to be honest about where the system struggled, because the failures are as instructive as the successes.
The Distil step — identifying the Core 20% — required discipline that was sometimes hard to maintain under production pressure. There were productions where the Wish List crept upward, where the instinct to capture more outran the honest assessment of what was actually needed. When the list became too long, the production relationship deteriorated. Requests that had been positioned as essential started to feel like a marketing tax on a creative process that was already under strain.
The remedy had two parts. The first was the evidence — going back to the data about what had actually been used in previous campaigns rather than the instinct about what might be useful. Which assets had driven real downstream value? Which had been captured, processed, and never once requested? The discipline of the Core 20% was not a one-time decision — it had to be re-earned on every production, against the actual record.
The second part was more important, and it came from having the filmmakers genuinely inside the process from the start. Because they had been part of building the Wish List — because they had agreed on which areas of the film they wanted to highlight, and understood why capturing certain moments mattered for the campaign — conversations about whether a capture was achievable became collaborative rather than adversarial. We weren't a marketing team pushing against production resistance. We were a group of people with shared ownership over the same problem.
When a planned capture was at risk — when a shoot day got redirected and a Wish List item might be lost — the conversation became: if we don't get it this way, how might we solve it later? The filmmakers were asking that question alongside us. That changes everything about the dynamic. A problem the marketing team owns alone is a problem that production can deprioritise. A problem the filmmakers are invested in solving is a different kind of problem entirely.
Not every miss was recoverable. There are moments from specific productions I wish we had captured and didn't. But the misses, when they happened, had a reason and a shared paper trail. That is fundamentally different from the earlier model, where misses were invisible until the moment someone asked for something that no longer existed.
What Transfers
08
The Building Block System was applied at Baobab Studios in 2020–21 — on the Namoo and Baba Yaga awards campaigns. The medium was entirely different: VR rather than stop-motion, digital environments rather than physical sets, rendered assets rather than captured footage.
The underlying constraint was identical. High-intensity production. Compressed marketing window. The need to have the right assets ready at the right moment for an awards ecosystem that wasn't built with the medium's specific requirements in mind. VR offered one advantage stop-motion doesn't — environments could be staged and rendered intentionally after the fact — but the Distil and Plan steps were just as essential, because the rendering time required to produce those assets intentionally was itself a resource that had to be planned and scheduled.
The framework also transferred to the consumer products programme I built at LAIKA. The same tagging structure that made marketing assets retrievable made product licensing assets manageable — style guides, character reference turnarounds, production photography, all pre-tagged and available for licensee use without a bespoke request-and-processing cycle for each new partner.
"The constraint changes. The principle doesn't."
The principle that underlies all of it is simple: the cost of capturing something at the right moment is almost always lower than the cost of not having it when you need it. Every production person already knows this informally. Most production environments don't have the infrastructure to act on it systematically. The Building Block System is that infrastructure — built not once, but refined across seven productions until the effort required to maintain it was a fraction of the effort it had replaced.
The Art of Namoo — the same Distil, Plan, Construct approach applied to a VR production. The semantic tagging framework built at LAIKA, adapted for a digital environment. ©Baobab Studios
Closing Principle
The puppets from Kubo have long since moved to LAIKA's archive. The Giant Skeleton has been carefully dismantled. The stage where Travis Knight animated himself into The Boxtrolls has been reset for the next production.
But the footage is still there. The characters holding green screen signs, animated to interact with whatever context needs them next. The Giant Skeleton rigging documented from every angle. The timelapse of Norman's walk cycle, two months of an animator's work compressed into four minutes. Tagged. Organised. Ready.
That's the Building Block System. Not a technology. Not a workflow. A way of treating irreversible moments as the assets they are — before they're gone.
References
ShotGrid (Autodesk) — production tracking and asset management platform, foundation of the Construct pipeline.
Autodesk customer story: LAIKA — independent third-party validation of the pipeline infrastructure.
Dragonframe — adopted by LAIKA from ParaNorman (2012), developed in close partnership into the global industry standard.
American Cinematographer, October 2005 — Pete Kozachik on Corpse Bride's digital migration. No stable direct URL exists for the article itself, but the ASC's in memoriam for Kozachik references it and is linkable
AWN / VFXWorld, September 2005 — Bill Desowitz on-set report, Corpse Bride: Stop Motion Goes Digital.
LAIKA — the studio whose productions the Building Block System was built for.
Pareto Principle — Juran, J.M. (1951). Quality Control Handbook. McGraw-Hill.