5 Things Your Creative Brief Needs Before We Start a CGI Project

Here’s something people don’t expect a studio to admit: almost every project that goes badly, goes badly because of the brief. Not the render. Not the software. Not the deadline. The brief.

We’ve delivered CGI and food commercials for FMCG and F&B brands across Egypt and the GCC for years now, and the pattern is boringly consistent. The projects that run smoothly are the ones where we knew what we were building on day one. The projects that turn into five rounds of revisions are the ones where everyone assumed everyone else understood the goal.

The frustrating part is that the fix is cheap. A good brief takes an extra hour of thinking on the client side and saves a week of production on ours. So let me be specific about what we actually need.

1. What this asset is for — and where it will live

This is the single most important line in any brief, and it’s the one most often missing.

“We need a video of the product” tells us nothing. A 6-second Instagram Story, a 30-second TV commercial, and a static key visual for a supermarket shelf-talker are three completely different products. Different aspect ratios, different pacing, different amount of detail you can even perceive, different hierarchy between the product and the packaging text.

If it’s for Instagram Reels, the product needs to read in the first half-second and survive being viewed on a phone at 60% brightness. If it’s for an in-store standee printed at A1, we need to build geometry and textures at a resolution nobody would waste time on for social.

So tell us: platform, format, duration, and where it sits in the campaign. Is this the hero film or the cutdown? Is it running alongside live-action footage we need to visually match? One sentence here changes every technical decision we make afterwards.

2. The actual packaging artwork — final, not “almost final”

I’ll be direct about this one because it costs more time than anything else on the list.

CGI lives and dies on the packaging artwork. We don’t paint a label from a photograph; we take your print-ready files and wrap them onto 3D geometry. That means we need the real thing: the AI or PDF print files, with the dielines, at full resolution, with the correct spot colors and finishes marked.

What we get instead, more often than we should, is a JPEG pulled from a presentation deck or a photo of the pack taken on a phone. We can work from that for a rough pre-visualization, but it will never hold up in the final render. Text goes soft, gradients band, logos develop halos.

And the other half of this: “final” has to mean final. If the artwork changes after we’ve built and lit the scene, that’s not a small update. Re-wrapping a label means re-checking every shot the pack appears in. If a new SKU version is still in legal review, tell us up front and we’ll structure the build so that layer stays swappable.

3. The finish and material reference — not just the color

Brands talk about packaging in terms of design. CGI artists have to think about it in terms of physics.

A matte-laminated carton, a glossy UV-spot varnish, a metallized foil, a frosted PET bottle, and a clear glass jar all behave completely differently under the same light. That behavior is what makes a render look real or fake. Nine times out of ten, when a client says a render “looks a bit CG,” what they’re reacting to is a material that’s too clean or too reflective for what the real product is.

So we need to know:

• Is the finish matte, gloss, soft-touch, or a mix?
• Any foil, embossing, or spot varnish? Where exactly?
• For bottles and jars: glass or plastic? Clear, frosted, or tinted?
• For food: how does the product actually look when it’s fresh versus after sitting?

The best thing you can send us is the physical product. A real sample on our desk answers a hundred questions that no PDF will. If the product doesn’t exist yet, send us the closest thing that does and tell us how the new one differs.

4. Honest references — including the ones you hate

Reference boards are standard practice now, which is good. But most of them only show what the client likes, and that’s only half the useful information.

Send us three or four references for the look you’re chasing, and be specific about why each one is there. “I like this one” is much less useful than “I like how hard the shadow is here” or “the color in this one feels closer to our brand than what we’ve been doing.” One is taste, the other is direction.

Then send us one or two you actively dislike. This is the part people skip and it’s genuinely valuable. Knowing that you find dark moody product shots too cold, or that a competitor’s style feels cheap to you, stops us from spending two days walking down a road you were never going to approve.

And a note on AI-generated references, since they show up constantly now: they’re useful for mood and composition, and we use them ourselves for pre-visualization. Just be aware that an AI image is not a promise. It can show a bottle with a shape that can’t be manufactured, or a lighting setup that ignores how glass works. If you send one, tell us what you want us to take from it — the mood, the angle, the color — rather than expecting a one-to-one match.

5. Who approves, and when

The least creative item on this list and possibly the most expensive one to get wrong.

We need to know two things: who has final sign-off, and at which stages we’re getting feedback. Our process has natural checkpoints — concept, 3D blocking, lighting and look development, animation, then final grade. Each one exists because it’s cheaper to change direction there than after.

What breaks projects is feedback arriving out of sequence. A marketing manager approves the animation, then a regional director sees it for the first time at final delivery and wants the camera angle changed. That’s not a revision; that’s going back to the blocking stage with a finished film’s worth of work already done on top of it.

So put the approval chain in the brief. If there’s a stakeholder who will definitely have an opinion, get them into the concept review, not the final one. If there’s a hard external date — a launch, a trade show, a Ramadan window — say it in the first email, not the third.

The short version

If you take nothing else from this: purpose, final artwork, material reality, honest references, approval chain. Five things. Most of them you already know; they just tend not to make it into the document.

None of this is about making the client do the studio’s work. It’s that CGI front-loads its decisions. Unlike a photo shoot, where you can try six angles in an afternoon and pick one, a render is expensive to reverse and cheap to plan. The hour you spend on the brief is the highest-leverage hour in the entire project.

And if you’re not sure how to answer some of these, that’s fine — say so. A brief that says “we don’t know yet, help us decide” is far more useful than one that guesses. That conversation is part of the job.