If you read enough trend reports, you start to notice they all sound the same — a confident voice asserting that "everything is changing" without saying much about how the work actually looks different on a Tuesday morning. So rather than write another one of those, I want to share what's genuinely shifted in our studio over the last twelve months, what clients are starting to ask for that they weren't asking for a year ago, and the bits of the hype cycle that have, predictably, not panned out.
I run visual production at a studio that ships product imagery to brands across the US, UK, and the Gulf. Most of what I'll say here is drawn from that work — what brand managers are briefing, what their merchandising teams are pushing back on, and where the budget is actually moving.
Human-led visual production
Want a visual system built around your product?
Share your brand, product references, and launch goals. We will recommend the right engagement model.
Personalised imagery is finally getting interesting
The pitch for personalised product imagery has been around for years: show every visitor a version of your hero image that matches their inferred taste. Until recently this mostly meant changing a background colour or swapping a model. The visual difference between "personalised" and "default" was usually too subtle to matter.
What's changed in 2026 is that the imagery being swapped in is now meaningfully different. A skincare brand can show a serum on a marble vanity for one segment and on a warm-toned wooden shelf for another, and both versions look like considered campaign shots rather than auto-recolours. The shift isn't that personalisation became possible — it's that the swapped-in images are now good enough that they don't undermine the brand.
I'd be careful about quoted conversion-rate uplifts you read on LinkedIn. Most of the numbers floating around are either internal to a single brand's test or measured against a deliberately weak control. What is fair to say is that for brands with a wide audience — say, a beauty brand selling into the US, Europe, and the Gulf simultaneously — being able to dress the same product in two or three regional contexts without commissioning two or three separate shoots is genuinely changing how teams plan their content calendar.
AR and product imagery have stopped being separate workstreams
For furniture, eyewear, and beauty especially, the AR try-on experience and the static product page are no longer produced by different teams using different reference assets. The same AI-generated 3D understanding of a product now feeds both — the marketing image and the AR overlay come out of the same pipeline.
The practical consequence is that brands launching a new SKU don't have to choose between "polished product photography" and "AR-ready model" anymore. They get both as a byproduct of the same brief. That's been one of the quietly significant shifts of the last year, and it's the one I'd watch most closely if you're in furniture, footwear, or eyewear.
The format-sprawl problem hasn't gone away — but it's better managed
A brand might need the same product shot as a 1:1 marketplace tile, a 4:5 Instagram crop, a 9:16 Reel still, a 16:9 banner, and a vertical email hero. A year ago the realistic answer was to shoot once and then either crop badly or commission additional formats at extra cost. Today most studios — ours included — generate the full format set as part of the original delivery, because the AI workflow doesn't care whether it's producing one canvas size or eight.
This is less of a "trend" than a quiet productivity gain. But it has reshaped what merchandising teams expect from a visual brief. The conversation has moved from "what's the hero shot?" to "what are the channels?" — which is a much healthier place to start.
Generated product video is real, but limited
Honest answer: video is the area where capability is moving fastest and where the hype is most ahead of the reality. We can generate a credible six-to-twelve-second product clip — a slow rotation, a soft pour, a fabric breeze — for most categories. That covers most of what an Instagram Reel or TikTok intro actually needs.
What's still hard, in my experience: anything longer than about fifteen seconds without an obvious join; anything involving complex human-product interaction; and anything requiring exact brand-accurate colour over a sustained shot. If a video studio quotes you a thirty-second AI-generated brand film for the price of a static shoot, look at their reel before you sign anything.
The cost gap has stopped being the headline
For a long time the conversation about AI product photography was dominated by the cost differential — studio rate of several thousand dollars a day versus a few hundred dollars per AI-generated set. That gap is still there, but I notice it's no longer what brand managers lead with when they brief us.
What they lead with now is speed and revision tolerance. A traditional shoot locks the look on the day. If the merchandising team decides three weeks later that the background should be warmer, you're either reshooting or retouching expensively. AI production lets a brand keep iterating on the same brief for the entire campaign window without re-mobilising a crew. The cost saving is incidental; the operational flexibility is what's changing how budgets get allocated.
Sustainability is a real conversation, not a greenwash one
I was sceptical about this when European clients first raised it. It's easy to wave the sustainability flag whenever you switch any process from physical to digital. But for brands shooting four seasonal campaigns a year across multiple markets, the physical cost of those shoots — flights for talent, set construction, freight of sample SKUs — is real, and the audit trail matters to the procurement teams asking the questions. AI photography genuinely removes a chunk of that footprint. Whether that's a primary reason to make the switch will depend on your brand; but if it's on your compliance team's list, the math is on your side.
Search visibility and AI-generated imagery
Google's position has evolved over the last year. The current guidance, in short: AI-generated product imagery is fine, provided the image is original (not scraped), relevant (it actually shows your product), and the page around it offers genuine value. There's no penalty for AI-generated visuals as a class.
What does help — and this has been true for a while — is properly labelled alt text, structured data on product pages, and image filenames that describe what's in the frame rather than a string of model output IDs. None of that is new. The change is that brands using AI photography now have the volume to keep their image library fresh week-on-week, which compounds in image search over time.
The bit that hasn't changed
Underneath all of this, the thing that determines whether a brand's visual content actually works hasn't shifted at all. It still comes down to creative direction — knowing what your audience is going to respond to, what a hero shot should communicate about positioning, and how a single image fits into the longer narrative the brand is building. None of that is replaced by AI. The tools have got better; the judgement still has to come from somewhere.
That's where the conversation goes wrong, I think, when AI product photography is framed as a replacement for the people doing the work. The brands getting the most out of these tools are the ones whose creative leadership has the strongest opinions about what their imagery should look like. The technology amplifies that point of view; it doesn't generate one.
What to actually do about it
If you're auditing your visual content plan for the rest of 2026, three questions are worth sitting with:
- Where are you reshooting the same SKU because the original shoot couldn't be revised cheaply? That's where AI production saves you the most operational pain.
- How many formats does each product image need to exist in, and how is that work currently getting done? The answer often reveals an entire hidden cost that an AI workflow makes disappear.
- Where is your team's creative judgement strongest, and where are you defaulting to the safest possible image because resources are tight? AI production frees up the second category so your team can spend more time on the first.
If any of that resonates, our team is happy to talk through it without a pitch. You can get in touch here, or browse our portfolio to see how the work translates across categories.