I started a small independent fashion brand four years ago. Slow fashion, made-to-order, mostly women’s wear in natural fabrics with a quiet aesthetic that does not pull off well in fluorescent lighting. We do maybe two hundred orders a month, which is enough to pay me and one part-time pattern maker, and the brand has held together through the kind of small monthly disasters that come with running a clothing line that does not really fit into any major sales channel.
The single line item on the brand’s annual budget that has caused me the most agony, year after year, has been photography. Not the basic product shots — I have a corner of the studio set up for those, and a decent camera. The lookbook. The seasonal, location-based, model-styled, mood-rich photo sets that fashion brands are expected to produce twice a year minimum, sometimes four times, and that small brands like mine can almost never actually afford to produce well.
This article is about how I have, over the last year, mostly gotten around that problem.
Why Fashion Needs Lookbooks Differently Than Other Industries
Lookbooks are not the same thing as product photos. A product photo is documentary. It shows the item as it is, often against a clean background, so the customer can see exactly what they are buying. A lookbook is a mood. It shows the item being worn, in a setting, by a person, in a way that communicates who the brand is and who its customer might want to be.
For mainstream e-commerce, product photos are doing most of the conversion work and lookbooks are a nice-to-have. For independent fashion, the balance flips. The lookbook is doing the brand-building work that makes a customer choose your dress over an objectively similar dress at a department store. Without strong lookbook imagery, an independent brand reads as either generic or hobbyist. With strong lookbook imagery, the same brand reads as intentional and desirable.
This is why every successful indie fashion brand pours so much money and time into seasonal shoots. The shoots are the brand. They are not optional in the way other marketing might be.
What a Real Fashion Lookbook Shoot Actually Costs
I priced out a full lookbook shoot in 2022, in the second year of the brand. The numbers nearly killed the project.
A modest spring lookbook shoot, with one model, one photographer, one stylist, one hair and makeup person, one location, and a single day of shooting, ran about six thousand dollars total. That number assumes everyone involved is mid-career and not famous, and that the location is something we could negotiate access to rather than rent at commercial rates.
For two seasonal shoots per year, that is twelve thousand dollars annually, on a brand doing maybe two hundred thousand in revenue. Six percent of revenue going into photography alone is a number that gets discussed in board meetings at larger brands and considered borderline at brands like mine.
And that is the modest version. A more ambitious shoot — multiple models, a destination location, multiple looks, video coverage — runs twenty to fifty thousand dollars without much trying. Those numbers are not available to most small brands. They are available to the brands that get written about in fashion magazines, which is part of why those brands are the ones that get written about.
The result is that small independent fashion brands historically either spent money they did not really have on shoots, or produced lookbooks that quietly looked like home photography, which signaled to customers that the brand was not serious.
Where Nano Banana Entered the Picture
The first time I tried using Nano Banana for a lookbook image, I was about three weeks from the launch of a new collection and well past the point where booking a real shoot was feasible. I had clean product photos of the new pieces, taken in my studio against a neutral background. I had no model, no location, no stylist, no time.
I uploaded one of the product photos — a long linen dress in a soft terracotta color — and asked Nano Banana for the same dress, in the same color and silhouette, worn by a woman in her early thirties, standing in the doorway of an old stone farmhouse in late afternoon light, with the dress catching a slight breeze. I described the woman, the light, the location, the mood.
What came back was a lookbook image. The dress was recognizably my dress — same drape, same neckline, same color, same fabric texture. The setting was rich and specific. The woman looked like a real person, not a stock photo. The lighting was warm and intentional. The whole image read as something that could have been shot on a real production day with a real budget.
I sat there at my desk for a long time, not quite sure how to feel about it. Then I started generating the rest of the lookbook.
How a Full Lookbook Comes Together From One Photo
The workflow that emerged over the next few weeks is the workflow I still use. It starts with a clean product photo — the cleaner the better, ideally against a neutral background, with the garment laid flat or shown on a mannequin so the silhouette is clearly visible.
From that single product photo, I can generate the same garment in any number of contexts. The terracotta linen dress worn in the farmhouse doorway. The same dress on a sunlit beach in early morning. The same dress in a tiled courtyard with potted lemon trees. The same dress against a soft gray studio backdrop with directional lighting. Same dress, same drape, same color, different worlds.
The key is consistency. The dress has to look like the same dress across every shot, or the lookbook reads as chaos rather than as a coherent visual story. Nano Banana handles this surprisingly well, as long as the original product photo is clean and the descriptions of the variations are specific enough.
I usually generate eight to twelve images per piece, then narrow down to the four or five strongest for the actual lookbook. The discards are not wasted — they often become social media content, email header images, or website background imagery that I would otherwise have had to source separately.
Model Consistency Across the Collection
The other thing I did not expect to get from this workflow is model consistency across an entire collection.
Most independent fashion brands cannot afford to book the same model for an entire seasonal campaign. The result is that lookbooks often feature multiple different models, which dilutes the visual story. Compare this to bigger brands, which can keep a single model across an entire campaign and build a stronger brand association. Nano Banana closes that gap for small brands in a way that did not really exist before.
With Nano Banana, I can generate an imagined model — described in enough detail that she comes out the same way each time — and use her across the whole collection. Same woman, same general look, same coloring, different outfits and settings. The lookbook gains a consistency that used to be reserved for brands with proper budgets.
I want to be careful about how I describe this, because it touches on a real question about fashion and authenticity. The model in my lookbook is not a real person. The customer is not buying clothing that has been worn by an actual human in those exact circumstances. I have made the choice to use generated imagery in my lookbook, but to use real customer photos and real fit photos elsewhere on the site — the kind of imagery where customers can see how the clothing actually looks on a real human body. The lookbook is the mood. The fit photos are the truth.
Seasonal Updates Without Production Cost
The other shift this enables is the cadence of lookbook updates.
Real photoshoots happen twice a year for most small brands because they cost too much to do more often. But customer attention works on a much faster cycle than that. Six months is a long time to ask people to keep looking at the same lookbook imagery.
With Nano Banana in the workflow, I can refresh the lookbook in any direction the season takes the collection. Holiday-themed imagery in December. Spring-themed in April. A mid-summer “wearing the same pieces in warmer settings” set in July. The same core garments, generated into new contexts as the cultural moment shifts.
This kind of frequent updating used to be reserved for brands that could afford monthly shoots, which is a small number of brands. It is now available to anyone willing to put in the time to generate and curate.
The Honest Conversation About AI Imagery in Fashion
I want to address this directly, because I think it is the most important conversation in this whole area.
Fashion is an industry built on aspiration. Customers know that lookbook imagery is constructed. They know the model has been styled and lit. They know the garment has been steamed and pinned to fit perfectly for the camera. The aspirational quality of lookbook imagery is not a deception. It is part of the product.
What changes when the entire image is generated rather than photographed is that the line between “constructed reality” and “imagined reality” shifts. I think small brands using Nano Banana imagery have an obligation to be transparent with their customers about how the imagery was made, and to pair the aspirational lookbook with truthful fit photos that show the garment on actual bodies. Anything less starts to feel like deception rather than aspiration.
I have started writing a small note on my About page that acknowledges some of the lookbook imagery is generated through Nano Banana, and pointing customers to the fit gallery where real people are wearing the pieces. The customer response has been more positive than I expected. People appreciate the transparency. They also continue to buy the clothes.
What Nano Banana Cannot Quite Do for a Fashion Brand
There are limits, and being clear about them matters.
Garment fit on a body is still something AI gets approximately right but not perfectly right. The generated image of a dress on a model will show the dress falling in a way that is close to correct, but a real photographer with a real model can capture the specific way that this specific fabric drapes on this specific body. For technical fit communication, real photography wins.
Detail close-ups — the specific texture of a fabric, the exact hardware on a button, the seam construction — are not really lookbook territory anyway, but worth flagging. Those still need real product photography, often shot by me with a macro lens.
Video is mostly outside the Nano Banana workflow. A lookbook video — short clips of garments moving, models walking, fabric catching the light — still requires a real shoot.
And cultural specificity matters. Generated models default toward certain demographics unless explicitly directed otherwise, and brand owners working with AI imagery should be deliberate about who appears in their lookbook and what that says about who the brand is for.
Why This Quietly Levels the Field for Small Brands
For most of the industry’s history, the fashion brands that looked the most expensive were the ones that could afford the most expensive photography. The visual gap between the small brand and the large brand was a budget gap, not a taste gap. Lots of small brands had excellent taste and lookbook imagery that did not communicate it.
That gap is narrower now. A small brand with strong taste and access to tools like Nano Banana can produce lookbook imagery that holds its own next to brands ten times its size. The clothes still have to be made well. The fit still has to be right. The customer service still has to be human. But the visual presentation of the brand no longer requires a six-figure photography budget to look serious.
For independent fashion specifically, where the brands that survive are usually the ones that look more interesting than they technically should given their size, this shift is genuinely meaningful. The next decade of small fashion is going to look a lot more visually ambitious than the last decade did, and the budgets are not going to be the reason.
