Here’s something most retailers already know but don’t think about enough: product images do the heavy lifting in e-commerce. Customers can’t pick anything up. Can’t feel the fabric or check how heavy it is. They’re making decisions based almost entirely on how something looks in a picture.
Which raises the question — what’s the best way to create those images?
Traditional photography has been the default forever. But CGI has caught up in ways that surprise people. The gap in quality has shrunk dramatically, and for certain use cases, CGI actually comes out ahead. That said, photography isn’t going anywhere either.
It’s not really an either/or situation. More like figuring out which tool fits which job.
Traditional photography does some things really well
There’s something about a real photograph. Light hitting an actual object, captured through a lens. It feels authentic because it is. And for certain products, that matters.
Anything where texture is the selling point — chunky knit sweaters, artisanal bread, skincare with that dewy finish — tends to photograph better than it renders. The subtle imperfections, the way fabric catches light slightly differently in each fold. CGI can approximate this, but it takes serious skill and time to nail.
Photography also doesn’t require explaining to anyone. Your team knows how it works. Your suppliers know how it works. There’s no learning curve, no new vendor relationships to figure out.
But the downsides are real, too. Shoots take time to organize. You need the actual product in hand, which rules out pre-launch marketing. And if you’re selling, say, a chair in fifteen fabrics? That’s fifteen separate shots. Multiply that across a whole catalog, and you’re looking at significant costs and logistics headaches.
Where CGI starts making sense
The pitch for CGI is flexibility. Build a 3D model once, use it forever. Need that chair in fifteen fabrics? Swap the texture file, re-render. Takes a fraction of the time and cost of reshooting.
And you’re not limited by physical reality. Want the product floating? Done. Want to show a sofa in a room that doesn’t exist? Easy. Want marketing materials for something that’s still in prototype? CGI doesn’t care that the product isn’t manufactured yet.
Retailers with large catalogs have figured this out. The upfront investment in 3D modeling services pays for itself once you’re reusing those assets across seasons, campaigns, and channels. The per-image cost drops the more you use them.
Quality-wise, CGI has come a long way. For hard goods — furniture, electronics, homewares, accessories — a well-executed render is genuinely hard to distinguish from a photo. Sometimes it looks better because you can control every single variable. No dust. No weird reflections. Perfect lighting every time.
The catch is that “well-executed” part. Bad CGI looks bad in a way that’s immediately obvious. It requires genuine skill to pull off.
The cost question is complicated
On the surface, a single photoshoot might seem cheaper than commissioning 3D models. And for one or two products, that might be true.
But photography costs are mostly linear. Ten products cost roughly ten times what one product costs. CGI is front-loaded — expensive to set up, cheap to scale. The crossover point depends on your specific situation, but it comes faster than most people expect.
There’s also the hidden costs of photography. Shipping products. Coordinating schedules. Rebooking when something goes wrong. That time your hero product got stuck in customs and you had to push the whole campaign back two weeks. CGI doesn’t have these problems.
Speed works differently, too
Photography timelines are fixed by logistics. You can’t shoot faster than products can ship. Rush fees exist, but they’re painful.
CGI can move faster once systems are in place. Last-minute background change? An afternoon. New angle for a social campaign? Probably a day. This kind of agility matters when you’re running a lot of campaigns or reacting to trends quickly.
But — and this is important — CGI isn’t instant. Building quality 3D models takes time upfront. The speed advantage kicks in after that initial investment.
Some products just photograph better
It’s worth being honest about CGI’s limitations. Anything organic or highly textured remains challenging. Food photography is still dominated by actual food. Fabrics with complex weaves or delicate draping. Cosmetics where the whole appeal is sensory.
These categories aren’t impossible for CGI, but they require top-tier artists and more time to get right. Sometimes a camera is still the easier path.
Most retailers end up doing both
The smartest approach is usually hybrid. Let CGI handle the heavy lifting — the catalog images, the variations, the endless SKU photography that nobody wants to organize fifty shoots for. Save traditional photography for hero products, flagship campaigns, or product categories where authenticity matters most.
Some brands go further and composite CGI products into photographed environments. Real room, digital sofa. It sounds complicated, but the results can be seamless.
There’s no prize for purity here. Use what works.
What actually matters
The CGI vs photography debate is less important than the quality of execution. A mediocre photoshoot won’t outperform excellent CGI. Lazy renders won’t beat skilled photography. The method matters less than doing it well.
Figure out what your products need. Consider your catalog size, update frequency, and budget. Then invest in quality execution, whichever direction you go.
