Nine Months In: What We Learned After Launching ACi Studios
This article has been written by Giles Mosley

Nine Months In: What We Learned After Launching ACi Studios
When we launched ACi Studios nine months ago, we thought we were entering a conversation about Ai and technology.
We weren’t.
We were entering a conversation about standards.
Yes, Ai is the headline. It’s the thing everyone talks about. It’s the novelty that gets attention. But after nine months of real projects, real deadlines, real product ranges and real brand teams signing things off, the pattern is obvious: people don’t buy “Ai imagery.” They buy imagery and video that performs. They buy work that feels premium, believable and on-brand. They buy assets that they can put live with confidence.
The technology is rarely the deciding factor. The outcome is.
Because in retail, no one gets rewarded for how a picture was made. Customers don’t care. They don’t pause on a PDP and think, “This was generated affordably.” They either trust what they’re seeing, or they don’t. They either feel desire, or they don’t. They either click “Add to basket,” or they keep scrolling.
And that’s the first thing we learned: the market isn’t hunting for cheap Ai. It’s hunting for superb creative at speed.
The uncomfortable truth about the current Ai wave
The second thing we learned is a little more uncomfortable.
Most people using Ai to create imagery are making a mess with it.
Not always in an obvious way. Sometimes the images look impressive at a glance. The lighting is dramatic. The scene is “cool.” The output has that glossy, attention-grabbing sheen that generative tools are good at.
But retail isn’t judged at a glance. Retail is judged in repetition.
It’s judged in grids. It’s judged next to last season’s photography. It’s judged beside competitor products. It’s judged across marketplaces, social, email, category pages, hero banners and detail sections. And when you look at Ai imagery in that real commercial context, you see the problem: it often doesn’t hold. It drifts. It changes its mind. It produces a different “version of the brand” every time someone types a new prompt.
What you end up with isn’t a visual system. It’s a visual slot machine.
And established brands can’t afford that. Not because they’re old-fashioned, but because they’ve spent years and millions building recognition. Their advantage isn’t that they can produce images. Their advantage is that customers know them when they see them. Consistency is equity. Style is memory. Coherence is trust.
Ai, left to itself, isn’t built to protect that.
It’s built to generate.
What people actually want (and what they keep asking for)
In nine months at ACi, the request has been remarkably consistent, even when it’s phrased differently.
People want images and video that feel like their brand, not like a tool.
They want the product to be truthful. Materials to read correctly. Proportions to feel right. The scale to make sense. The styling to signal the right price point. The environment to match the customer they’re selling to. The tone to be familiar enough that it doesn’t jar the existing brand world, but elevated enough that it feels like progress.
They want speed, but not chaos.
They want savings, but not compromise.
They want Ai’s flexibility, but not Ai’s randomness.
Most of all, they want confidence. They want to know that when an asset goes live, it won’t create a quiet credibility leak. Because that’s how brands lose ground now. Not through one dramatic failure, but through a hundred small inconsistencies that make the brand feel less intentional over time.
So yes, budgets matter. Of course they do. But what we’ve learned is that **quality is more important than price**, and it always has been. AI hasn’t changed that rule. If anything, it has made the rule more visible, because it has flooded the market with “good enough.”
And good enough is everywhere.
Ai has levelled execution, not judgement
This is the real shift.
Ai has brought the cost of execution down. It has made creation accessible. It has raised the baseline. A smaller brand can now produce passable content at volume, quickly. That’s healthy for the market. It’s exciting.
But it means the old advantage, simply being able to produce more, faster, with bigger budgets doesn’t protect you the way it used to.
Now the advantage is judgement.
Judgement is the part that decides what belongs in frame. What should never be in frame. What “premium” means for a specific retailer. What lighting language fits a category. What kind of styling feels aspirational versus fake. What kind of composition sells form versus distracts from it. What a customer needs to feel in half a second for the image to do its job.
Ai can’t own those decisions.
And that’s why so much Ai imagery ends up looking like content, rather than commercial work.
Where ACi Studios sits (and why our approach works)
This is exactly why ACi Studios exists.
We didn’t build ACi to become another platform where anyone can generate anything. We built it to solve a retail production reality: brands need scale, but they also need control. They need speed, but they also need standards. They need the flexibility of Ai, but with the discipline of a studio.
The thing that has surprised people most in these nine months is not that Ai can generate impressive results. Everyone knows that now. The surprise is how different the output becomes when Ai is treated like a production engine inside an expert-led process.
Because we don’t start with “What can we generate?”
We start with “What must be true?”
What must be true about the product. What must be true about the brand. What must be true about the customer. What must be true about the commercial purpose of the asset.
Then we direct the generation with the same mindset we’ve used for decades in high-volume retail production. The same discipline. The same eye. The same insistence on consistency. The same focus on what works when the image is repeated across hundreds of SKUs and dozens of channels.
That’s the part people don’t get from “Ai imagery.”
They get it from experience.
The real lesson after nine months
The biggest thing we’ve learned since launch is simple: the market doesn’t want Ai. The market wants positive outcomes.
They want imagery and video that looks superb, engages customers and strengthens a brand rather than diluting it. They want work that holds up under scrutiny and repetition. They want content that doesn’t just look good in isolation, but builds a recognisable visual world over time.
Ai makes it possible to produce faster and at scale.
But the thing that makes it valuable is direction.
In a world where everyone has access to the same technology, technology stops being the advantage.
Judgement does.




Share this article