Every Shopify brand watching their product catalog outgrow their photography budget is staring down the same wall right now: it’s already Q3 and back‑to‑school offers, fall lineups, and BFCM production are all fighting for the same shoot calendar.
If your team is still betting on a single studio day or a stack of DIY lightbox tutorials to keep pace, you’re going to fall behind the brands that have already figured out how to scale visual content without scaling their shoot schedule.
Sam Thomas Davies is Head of Content at Dreem, the AI‑powered content platform spun out of Creative Force, the enterprise production software that runs content operations behind brands like Calvin Klein, Columbia, and SanMar. Dreem takes that same enterprise‑grade production logic and hands it to SMB and mid‑market Shopify merchants: upload a single product photo (even one shot on your phone) and generate a full content kit of pack shots, on‑model imagery, lifestyle scenes, and video—no studio and no extra shoot required.
In this conversation, Sam breaks down the real economics of AI‑generated product content: where traditional photography still wins, where Gen‑AI content closes the gap in hours instead of weeks, and the crawl‑walk‑run framework brands are using to test it without blowing up their existing workflow. Whether you’re a one‑SKU brand testing your first Gen‑AI shots or an established operation weighing a full content refresh before Q4, this is your playbook for modern product content.
Let’s dive in. 👇
✅ Why Q3 is the real deadline for holiday content: How fall lineups, back‑to‑school, and BFCM offers all fight for the same shoot calendar—and why brands that wait until Q4 to plan their content are already behind.
✅ How a single product photo becomes a full content kit: Dreem’s process for turning one phone shot into pack shots, on‑model imagery, lifestyle scenes, and video—and what’s actually happening behind the scenes to make that work.
✅ The real economics of AI‑generated content vs. traditional shoots: What changes when you compare a lightbox‑and‑DIY setup against an enterprise studio budget, and where the time and cost savings actually show up.
✅ The crawl‑walk‑run framework for testing Gen‑AI content: The three buckets brands fall into (internal‑only testing, small pilots, and full adoption) and how to choose the right starting point for your team.
✅ Why “human in the loop” still matters: How art direction, talent selection, and brand style guides stay in your hands, even as more of the production work gets automated.
✅ The sweet‑spot merchants winning with Dreem today: Why both one‑SKU brands testing the waters and growing catalogs that are outpacing their shoot schedule are seeing the biggest impact right now.
If you run a Shopify store, here’s a number that should bother you: 1-2% of your revenue can quietly disappear to chargebacks months after you won the sale. You paid to acquire that customer, earned the conversion, and then a dispute four months later takes the money back. Most operators just eat it, because fighting disputes by hand isn’t worth the hours.
That’s what Chargeflow fixes. It’s a 2-click Shopify install that automates the entire dispute process. AI builds and submits the evidence for you, with zero manual work. And you only pay when Chargeflow actually recovers the money. Win, and it’s worth it. Lose, and it costs you nothing.
See what you’re leaving on the table at https://app.chargeflow.io?plid=140894
Sam Thomas Davies starts with Dreem’s origin story. Creative Force first earned its reputation untangling content chaos for enterprise retailers like Calvin Klein, Columbia, and SanMar—teams running multiple shooters, stylists, and full commercial studios just to keep catalogs current. When leadership across those accounts started asking the same question (“what do we actually do with AI?”), Creative Force spun up Dreem as an experimental sandbox. What began with simple color changes and model swaps has evolved into a self‑serve content engine built on frontier models like Nano Banana 2 and Kling 2.5, wrapped in a production harness purpose‑built for e‑commerce.
Sam then walks through what actually happens behind the scenes. A merchant uploads a product photo, chooses (or creates) a talent from Dreem’s library, and the platform builds out a full content kit—on‑model shots, pack shots, and video—while keeping a human in the loop at every step so the output still reflects real brand direction, not a generic AI look. That “human in the loop” principle comes up again and again in this conversation, and it’s the piece Steve keeps returning to: this isn’t about removing the art director; it’s about giving them far more raw material to work with.
The economics are where this gets real for merchants at every stage. Sam breaks down why SMB brands—the ones shooting product on an iPhone in the driveway—see the fastest payoff by turning existing product shots into video for a fraction of what a traditional shoot costs, while enterprise teams are wrestling with a different challenge entirely: scaling consistent content across tens of thousands of SKUs without losing brand fidelity. He also lays out the crawl‑walk‑run framework Dreem uses with customers, from internal‑only experimentation to full‑scale adoption, and makes the case that there’s no wrong place to start as long as you’re testing against a clear hypothesis.
This isn’t theory. It’s a working blueprint for closing the gap between your product catalog and your content capacity—whether you’re a one‑SKU brand testing your first AI‑generated shot or a Shopify operation pushing thousands of new products into Q4.
👉 Treat your content calendar like a Q3 problem, not a Q4 scramble. Back‑to‑school, fall lineups, and BFCM offers all compete for the same production window, so brands planning holiday content in November are already behind. Folding AI‑generated content in as a supplement, not a replacement, gives you slack in that calendar without adding headcount.
👉 The economics favor AI content earliest for ad‑hoc SMB setups. If your current photography stack is a phone, a lightbox, and whoever’s available that day, turning those existing shots into video and additional looks is almost pure upside. The savings show up first in what you stop paying for—extra shoot days, talent bookings, and reshoots—rather than in a single new budget line.
👉 “Human in the loop” is the differentiator, not a drawback. Brands that stay hands‑on with talent selection, styling, and brand guidelines get output they can confidently put their name on. Teams treating AI as fully hands‑off end up with generic, forgettable results; the tools are only as strong as the direction you give them.
👉 Use a crawl‑walk‑run framework instead of all‑or‑nothing decisions. Start internal with mood boards and styling exploration, then move to a small pilot with a clear hypothesis (for example: does on‑model outperform pack shots on this PDP?), and only then consider a fuller rollout. This keeps risk low while you learn what actually moves conversion for your specific catalog.
👉 Unisex and low‑SKU catalogs have a big, underrated upside. If you’ve only ever shot one gender for a unisex product, or you’re a one‑ or two‑SKU brand debating whether a full shoot is worth it, AI‑generated variations let you test demand and visual directions before committing budget. It’s a cheap way to answer questions that used to require full production.
👉 Enterprise brands should optimize for scale and consistency, not just speed. Once catalogs stretch into the tens of thousands of SKUs, the hard problem isn’t producing one great asset—it’s maintaining brand fidelity across all of them. That’s where workflow, style‑guide enforcement, and the production harness around the AI matter just as much as the underlying models.
Sam Thomas Davies
Head of Content, Dreem
Sam Thomas Davies leads content at Dreem, the AI‑powered content platform Creative Force spun out after years of running enterprise production for retailers like Calvin Klein, Columbia, and SanMar. That enterprise pedigree shapes how Dreem approaches AI‑generated content: not as a replacement for craft, but as a harness that gives merchants of every size access to production discipline that used to require a full commercial studio.
Sam has been close to the product since its earliest days as an internal R&D sandbox, watching it evolve from simple color changes and model swaps into a full content engine capable of generating pack shots, on‑model imagery, lifestyle scenes, and video from a single photo. He regularly hosts live demos and webinars that walk brands through building talent, testing workflows, and getting started with Gen‑AI content, and his hands‑on experience with real merchant questions shows up in how clearly he talks about the technology’s limits as well as its potential.
With deep exposure to enterprise‑scale production challenges and daily conversations with SMB brands testing their first AI‑generated assets, Sam brings a rare perspective that speaks credibly to both ends of the Shopify merchant spectrum.
Over 9 seasons, I’ve been incredibly fortunate to chat with some of the brightest founders building amazing Shopify brands, as well as the partners shaping the app and marketing ecosystem. Every conversation has taught me something new, and I’m grateful for the chance to learn alongside you.
What matters most is that this podcast helps you solve real challenges and discover new ways to grow. Your support, feedback, and stories have made this journey truly special. Thanks for tuning in, sharing your wins and losses, and being part of the eCommerce Fastlane community.
Stay Connected: Leave an Honest Rating/Review on Apple Podcasts or Spotify. Follow & Subscribe on YouTube for new episodes.
Steve Hutt:
Welcome back to eCommerce Fastlane, I’m your host. Today’s conversation is going to be all about product content—the photos and videos that power every product page you publish on your Shopify store. If your catalog is starting to outgrow your ability to shoot it, you’re not alone. This happens a lot.
Steve Hutt:
Just think about seasonality, especially with fashion. I’ve already seen back-to-school campaigns and I’m just getting into summer right now, but those offers are already out there. It’s interesting because, even though I don’t have a brand myself currently, I find it fascinating how forward-thinking brands need to be—literally planning in quarters. We’re in July right now as this is being recorded; we’re already in Q3. People are already thinking about back-to-school, some are thinking about BFCM offers, getting production in order, and lining up fall collections.
Steve Hutt:
My guest today, and his company, have solved a lot of these production problems brands are facing. His name is Sam Thomas-Davies. He’s the Head of Content at a company called Dreem, spelled D-R-E-E-M. Dreem is a merchant-facing product from a larger company called Creative Force, which we’ll talk about in a few minutes. Creative Force is a production platform that sits behind a lot of enterprise eCommerce studios and, from what I understand, turns a single product photo—even one you take on your phone—into a full content kit: pack shots, on-model imagery, lifestyle scenes, and video. It’s pretty amazing—no studio and no extra shoots required. It’s a really interesting solution.
Steve Hutt:
My hope today is that we’re going to unpack the economics around AI-generated product content. I think that’s the key: understanding traditional photography—there are lots of tutorials out there about how to DIY your own product shots—and then looking at the potential cost savings of having a tool like Dreem take care of all that for you. So, hey Sam, welcome to the show.
Sam Thomas-Davies:
Thank you very much, Steve. Very happy to be here.
Steve Hutt:
This is great. Before we jump into the Dreem product, can you set the stage a bit about Creative Force? Who are the brands you’re working with at the enterprise level? I want to understand what that company is and how it’s assisting these Fortune 500-style companies. Once we set that stage, I want to talk about the offshoot with the Dreem product.
Sam Thomas-Davies:
Yeah, so before we talk about Dreem, we need to go back and talk a little bit about the problem Creative Force solved a few years ago and continues to solve today. A lot of high-volume retailers like Calvin Klein, Columbia, and SanMar—big retailers you’ve definitely heard of—have a lot of frustrations around creating content at scale. They’re often working in fragmented tools, with internal communication spread across multiple systems, and there are many steps in the process. As you can imagine, it becomes a very complex operation over time.
Sam Thomas-Davies:
Creative Force is a software-as-a-service platform for enterprise brands that allows them to manage their workflows end-to-end within one platform. They’re able to integrate with tools like Capture One, and this is a good segue into how we got started with Dreem. In the last 12 months or so, there’s been a lot of development in AI, and many companies have been trying to figure out, from a positioning perspective, how to factor that into their business. In the last six months, we’ve seen a lot of our big customers come to us saying, “Leadership has told us we need AI. We don’t know what that means.” They’ve put together small teams to try and figure it out, but it’s a very broad term that can mean many different things for different companies.
Sam Thomas-Davies:
We had Dreem on the back burner. Originally, it was our research and development tool, where we were experimenting with AI—simple things like color changes, putting an image on a model, and very rudimentary image-to-video workflows. We started to take it very seriously in the last six months and allocated more resources to it. Now, as you said in your introduction, there are so many things you can do. One of them is: imagine capturing an image of a product on your phone, maybe even a sample in the warehouse before it’s shipped to your distribution center, and turning that image into what we call a content kit. It’s multiple pieces of content for a campaign: front and back on-model shots, a product or pack shot, and image-to-video.
Sam Thomas-Davies:
We’ve iterated a lot on that over the past few months to the point where you can create your own talent, and think about scale. We just launched our MTP, so you can actually talk to Dreem via Claude, and you’re able to integrate Dreem into different tools, including Shopify on the SMB side. It’s a really exciting time for the company, and hopefully we can dig a little more into Dreem on this call.
Steve Hutt:
Yeah, it’s amazing. You talk about these frontier models—I follow a lot of influential people in this space. I’m on the media company side, more in content production. These frontier models, and the intelligence level that’s available now, are incredible. Companies like yours are creating a harness around that intelligence to understand how to use these models specifically for eCommerce businesses and their use cases.
Steve Hutt:
You could say, “Why can’t I just upload a photo of myself into Claude or Claude Design and let it run?” It doesn’t actually work like that. There’s a lot of underlying technology—the harness in the background that sits on top of these models—that allows all this magic to happen. I think that’s the unique problem you’re solving.
Steve Hutt:
So let’s talk about what happens in practice. You mentioned the hoodie example: someone has a photo they took on their phone of a hoodie. What happens between uploading that photo, what’s happening in the background, and what’s ultimately delivered to the marketer or the brand side?
Sam Thomas-Davies:
To your earlier point, there’s a lot going on behind the scenes, and I feel that’s a strong competitive advantage for us because we’re completely transparent. The models we use—and when I say “models,” I’m differentiating from talent—are AI models like Nano Banana 2 and Kling 2.5. You can see these models visibly in the tool, but there’s a lot happening in the background.
Sam Thomas-Davies:
We just launched a new chat, and it’s able to identify, based on the image you upload, whether it’s male or female. We’re looking deeply into aesthetics, fit, and fidelity. We get a lot of questions like, “How does the AI model understand texture and drape?” That’s something we’re actively exploring. So, to your question, when you upload an image—or pull one from a source like Shopify—you go through multiple steps.
Sam Thomas-Davies:
You’re able to choose a talent, and, as I mentioned earlier, you can upload your own talent if you wish. You can have a digital twin if you have consent from a model you like using. This is important because it’s not a binary box where you’re either using Gen AI content or you’re not. It’s a spectrum—how much of it you use. We can come back to that later. With regard to what’s actually happening, the “human in the loop” part is crucial, and it’s something we believe in strongly.
Sam Thomas-Davies:
We see a lot on LinkedIn and elsewhere about people scaling their content production, and it can sometimes feel like you’re behind when you see a super-advanced workflow someone has built in Nanonets or similar tools. They’re pushing assets between different platforms. But it’s really important to keep humans in the loop. Just like in a traditional setup, you want to be hands-on with choosing talent. An art director wants control over how an outfit is built for a particular look, and how poses are selected for the model.
Sam Thomas-Davies:
Often, brands have style guides—very specific looks they need to adhere to. We want you, the human, in the loop so you’re able to give input at every step. That way, the output is something you’re happy with and can confidently put your stamp on, whether that’s on your website or any other channel.
Steve Hutt:
Yeah, it’s amazing. When I think about choosing talent, I think about brands with brand guidelines and style guides—they’re differentiating themselves from competitors. They know their ICP, their ideal customer profile, and why customers are moving along their journey.
Steve Hutt:
It’s interesting to imagine feeding that information into Dreem: “These are our brand guidelines, here are our style guides, here’s some live photography we’ve taken that got us to where we are today, and we’re looking to expand on that.” I love the human-in-the-loop aspect because that’s taste and art direction, but now the tool can make choices based on those frameworks.
Sam Thomas-Davies:
Yeah, absolutely. I think this is something many people overlook. We had a webinar recently where we made a strong point that this is not about replacing anyone. There’s sometimes a danger of thinking “no photo shoots, no models, no talent.” That’s more relevant on the SMB side, where if you can’t afford traditional production, you now have an advantage: you can compete with bigger companies by using a tool like Dreem to create video, for example.
Sam Thomas-Davies:
I spoke to a small business owner a few months ago who said, “It’s so costly for me to shoot video, but I can now do it with Dreem because I already have my own model shots.” Brands are starting to dip their toe into Gen AI content by taking existing assets they’ve spent years shooting and experimenting with turning them into video. If they have the model’s consent, they can experiment with different looks without reshooting for different seasons.
Sam Thomas-Davies:
I hope we can dig more into this on the call because I see so much opportunity not just in Gen AI content, but in AI in general across creative operations. There are so many things you can do, and I don’t think we’ve begun to scratch the surface.
Steve Hutt:
Yeah, and think about it: just a year ago we were wondering if the quality would hold up. Now, every new generation of models is coming out almost weekly. What was possible six months ago is nothing compared to what’s available today. The acceleration is wild.
Steve Hutt:
Dreem’s harness on top of these models, plus customer feedback wanting to expand the product and improve quality, is powerful. When you talk about drape and fabric, it’s a really interesting question: how do we simulate that? You do your own experimentation, but a lot of it is built into the harness itself—that’s why the product is so good. I’ve watched some videos produced by Dreem and a few case studies, and it’s unbelievable how good it is at scale.
Sam Thomas-Davies:
Yeah, you touched on an important point about feedback. When we were in private beta and showing Dreem to people, the reaction was often, “The generation looks good, but what about this?” They’d ask how the model understands drape, or how to ensure consistency across hundreds or thousands of views.
Sam Thomas-Davies:
A good learning for us was: show it warts and all, and don’t be afraid to show the product. Sometimes we’d run webinars and a generation would fail live, but it didn’t really matter—we showed that this is possible and something we’re actively working on. We’re only limited by where the models are at any given time. A couple of months ago, you couldn’t make a video without a back shot because the back would change when the model turned around. That’s no longer the case. We’re now at a point where you can have multiple reference shots as the model turns, adding more detail to the outputs.
Sam Thomas-Davies:
It’s extraordinary how fast things are accelerating. If you’re committed to continuous improvement and consistently getting feedback from users, you end up with a phenomenal product that can really stand out.
Steve Hutt:
Let’s talk about the economics. This is always interesting to me. We’re comparing AI generation—that’s what we’ve been discussing—against a traditional photo shoot.
Steve Hutt:
Some SMBs listening right now are in that traditional tutorial world: how to use a lightbox and a camera, going through those motions in the early days. Then you migrate to a more professional setup. When I was at Shopify, I worked with a fashion brand that had a corner of their warehouse set up as a studio. That’s where they brought in live models for product photography because their fashion lines changed a lot. They had a separate section, lighting, the whole thing. A lot of that has changed now.
Steve Hutt:
From a cost perspective, how do you see traditional photo shoots versus how things work at scale now with Dreem?
Sam Thomas-Davies:
It’s a good question. I think there are two ways to answer it, depending on whether you’re SMB or enterprise. On the SMB side, we sometimes speak to companies where their models are employees or even their children. If they want a lifestyle shot, they just go into the street and take pictures on their phone. For them, “10x-ing” their content is simply turning existing product photography into video.
Sam Thomas-Davies:
Other times, it’s about creating their own talent and generating more content using Dreem. For them, traditional photography is basic and ad hoc. On the enterprise side, if you visit some of the big companies that have their own commercial studio in-house, you see ghost mannequin shooters, on-model photographers, stylists, and a lot of crew on set. There’s substantial logistics and cost involved.
Sam Thomas-Davies:
So, it really depends on the company’s goal. As I mentioned earlier, you don’t fall into a camp of “you use Gen AI content or you don’t.” We have customers who aren’t ready for Gen AI on their site—sometimes it’s against the ethos of the company—but they love Dreem for outfitting and styling. Using the Outfit Builder in Dreem is much more fun and interactive than a pin-up board or a manual process.
Sam Thomas-Davies:
There are many different use cases for tools like Dreem. It’s about finding the work you don’t enjoy doing, how you can outsource that to AI, and how you can get more out of what you’re already doing.
Steve Hutt:
What I find interesting is the offshoot into social ads. If you’re creating content through Gen AI from one photo, and you get a lot of derivative content, you can share that with the paid social team and say, “Here are a bunch of assets for the new collection—feel free to use some in your ads.”
Steve Hutt:
The merchandisers working on PDP or collection pages are often different people than those working on paid social. They talk and work together, but it’s a different craft. Has there been a good workflow bridging content production for PDPs and social ads?
Sam Thomas-Davies:
Yeah, it’s a great question and something we’ve thought a lot about. In Dreem, depending on your workflow or goals, you can export assets in different dimensions and file types, from 2K to 4K. We’re also looking at pushing these assets to different platforms.
Sam Thomas-Davies:
As I mentioned, we have an integration with Shopify. Some people have reached out saying, “I love what you’re doing with Dreem, but I don’t want to log in. I want a workflow where, when I publish a new PDP on my Shopify page, it sends a request to Dreem, Dreem creates an asset, sends it back to Shopify, and then it gets pushed to my social channels.”
Sam Thomas-Davies:
We’re very close to enabling a lot of these scenarios, especially now with MCP and similar technologies. The workflow aspect is really exciting. We also recently launched a connector with Creative Force. Creative Force has a workflow engine where you’re able to define multiple steps, push assets through different production phases, and assign them to different people.
Sam Thomas-Davies:
Some teams push assets to Dreem for certain tasks, then back to Creative Force. Being able to pull in social channels as part of that workflow is an option too. Many of us know the pain of working in paid marketing—creative is key, and you’re often waiting on a designer. The ability to do this yourself while maintaining high quality is really exciting for a lot of teams.
Steve Hutt:
You mentioned some brands following a crawl-walk-run framework. Maybe they’re not ready for full Gen AI production yet. Some people push back on AI, talk about “AI slop,” and that’s often based on older models without the harnesses you’ve built.
Steve Hutt:
When you talk about traditional setups on a PDP—traditional photography, Gen AI, and potential video—that seems like a strong one-two-three punch: traditional photography where needed, using that as a reference image for AI-generated content, and then a video option. That gives site visitors more ways to engage with the product, stay longer on the PDP, click through images, and imagine the product on themselves. What’s your view on combining traditional, Gen AI, and video?
Sam Thomas-Davies:
To answer this, I’d look at it through the lens of a webinar we recently ran about how to get started. I really think brands fall into one of three buckets.
Sam Thomas-Davies:
The first bucket is: “We’re open to testing, but we want to keep it internal.” That could mean using Dreem for outfitting and styling, or exploring localization for a new market. We’ve spoken with people in creative ops who already use ChatGPT as inspiration for different looks and campaign angles. That’s bucket one.
Sam Thomas-Davies:
The second bucket, which ties to your video point, is running a small pilot. For example, if you’re an SMB and only have product or pack shots on your website, you now have an opportunity to put items on-model because we know on-model often converts better than pack shots. Or, if you already have on-model, you can take some low-converting pages and try turning existing assets into video.
Sam Thomas-Davies:
The important thing is to create hypotheses: “I want to increase the conversion rate on this PDP. How might I do that? Let’s test on-model versus video.” One of my favorite use cases is unisex products where you’ve only shot female or male; now you can show the other without a new shoot. Have some fun, build hypotheses, and test. It’s very easy to do.
Sam Thomas-Davies:
The third bucket is what we’re seeing with some big companies: they’ve gone all-in on AI. There’s a lot happening right now, and more coming soon, around regulations and transparency in Gen AI. On the Creative Force side, we’re releasing guidance to help brands stay compliant. Some companies are fully leaning into AI, but my point is there’s no single “right” way. It’s about understanding who you are as a company, how you feel about AI, and your stance. Then you run small experiments—internal or external. Often, just testing more video is a very easy early win.
Steve Hutt:
I see something similar with brands when we talk about platform creep or app creep. They’re using so many different tools, and when you audit everything, it’s clear they’re experimenting everywhere.
Steve Hutt:
What Dreem seems to do is create a source of truth for content production—image or video—and makes it easy for the brand to produce that content and know where it resides and what’s being worked on. Versus people getting ideas in ChatGPT or Claude or Gemini and scattering them. Am I on the mark in thinking Dreem is aiming to be the source of truth for eCommerce brands’ image and video content?
Sam Thomas-Davies:
I think so. I think it will be similar to what we see at Creative Force now—very much an end-to-end workflow, whether you’re SMB or enterprise. We already have multiple entry points, like Shopify or our Apple app. The aim is to push content to multiple platforms: social, your ESP, and other tools, while you interact with Dreem’s chat interface.
Sam Thomas-Davies:
We just launched our new V3 chat, and you can talk to it like any other LLM. You can say, “I have this product and I want an Australian-looking model. Who would you suggest from the talent library?” We have 400+ talents, and if we don’t have what you’re looking for, you can create a talent via a prompt. I can see teams working primarily inside Dreem, similar to how they work inside Claude or ChatGPT today, and pushing content out to the right places.
Steve Hutt:
Amazing. Let’s talk about the Shopify app. If someone goes to the App Store, types “Dreem,” finds it, and installs it, what’s happening in the background when they connect? I’m trying to understand the pipes between Shopify and Dreem—what’s happening, and how does the user actually use the platform? Is it from within the Shopify admin, or is the app more of a data pipeline with users working directly in Dreem.ai?
Sam Thomas-Davies:
We have a two-way synchronization. When you download Dreem, you can bring all of your products from Shopify into Dreem. As you create assets, you can assign them to specific products and push them directly to your Shopify pages.
Steve Hutt:
That’s cool—so there’s a one-to-one sync between the two. When you’re working on something, those assets are created and saved, but you don’t have to publish them live immediately. You can still view them within the Shopify admin for that particular product.
Sam Thomas-Davies:
That’s exactly right.
Steve Hutt:
Amazing. What do you see as the next steps from a merchant perspective? I’m always interested in the “sweet spot.” You’re serving enterprise through Creative Force, but it seems like you’re moving downstream to help SMBs and mid-market merchants with more of a DIY solution.
Steve Hutt:
Can you talk about the sweet-spot merchant that’s getting the best value out of Dreem today? I’m thinking in terms of economics: is there real value beyond testing and experimentation?
Sam Thomas-Davies:
When we speak to eCommerce brands, the one metric they think about more than anything is time to market. How do they reduce it—from weeks to days, or even hours? I truly believe Dreem can help with that.
Sam Thomas-Davies:
One problem we’re working hard on solving right now is scale. How do you scale this? We’ve talked a lot about product shots, pack shots, video, and on-model, but what about lifestyle assets? What about infusing your brand guidelines and aesthetic into the content? What do you do when each SKU might have 10–20 assets, and you’re launching hundreds, thousands, or tens of thousands of SKUs each year?
Sam Thomas-Davies:
We’re looking deeply into scaling while maintaining consistency. This isn’t unique to AI—traditional photography has post-production, revisions, and humans in the loop. That’s the big problem we’re trying to solve now, and we’re very excited about it. We see clear paths forward, especially as the models keep improving.
Steve Hutt:
When we talk about traditional photo shoots, I want to go back to that, because it ties into crawl-walk-run. When do you think a traditional photo shoot might still be the best option for a brand today? And what eventually motivates them to move toward Gen AI production?
Sam Thomas-Davies:
I imagine on the SMB side, as the business grows and you have more products, there will be a need for traditional shoots. The more and better content you have for your products, the higher your conversion rate, sales, and revenue, and the more opportunity to scale.
Sam Thomas-Davies:
Typically, SMBs come to us when they simply can’t afford video. They can put a dollar amount on what video costs, and Dreem delivers that at a fraction of the price. That’s often what lights a fire under them—the ability to see how much more they can allocate to their marketing budget. They can spend more time acquiring customers instead of paying for studio time.
Sam Thomas-Davies:
That unlocks a lot of opportunity on the SMB side.
Steve Hutt:
I also think of brands with one or two SKUs, or maybe an expensive product. It might not be in their current workflow to use AI. They might have annual versions of the product with enhancements, but they’re still willing to spend money on traditional photo shoots because they only have a couple of SKUs.
Steve Hutt:
That happens a lot in consumer electronics or crowdfunding products on Indiegogo and Kickstarter. They’ll invest in a traditional video and photo shoot, and that’s fine. But there are still options on the experimentation side. If you’re investing and have one SKU—say you’re a luggage brand with a single set of luggage—you can take the original photography yourself, then use your phone with the Dreem app to see what comes out. It feels like a no-brainer.
Sam Thomas-Davies:
Yeah, especially in fashion and apparel, which is our specialty, it’s not enough to have simple front and back shots, depending on the sophistication of the product. You want strong fidelity. You want to show the material and colors, close-up shots, what we call detail shots, and video.
Sam Thomas-Davies:
Luxury brands often have ten or more assets per SKU. On the SMB side, they might not afford that or haven’t thought in those terms. They might not have the workflow sophistication to think that way. But now, when it’s as simple as uploading a front and back shot and maybe a detail shot, and generating as many poses as you want, that’s very exciting.
Steve Hutt:
So cool. We chatted before recording and I want to be mindful of everyone’s time. I really think this is one of those tools you don’t realize you need until you try it. You don’t know what you’re missing.
Steve Hutt:
You have an offer for our listeners. I’m looking at the App Store right now—pricing is already very affordable on its own, so it’s definitely worth a test. I’ll hand the baton back to you: what’s the offer, and what do you see as next steps for those listening today?
Sam Thomas-Davies:
Dreem is free. We offer 100 monthly credits, which is more than enough to get started. It’s credit-based, so different asset types require different numbers of credits depending on quality and output. You can try Dreem for free.
Sam Thomas-Davies:
If you’re listening to this podcast, we’d love to give you some bonus credits so you can really test Dreem. If you go to dreem.ai/fastlane, those will be auto-applied to your account.
Steve Hutt:
Perfect. I’ll make sure that link is in the show notes. I’ve written a lot of notes today and learned a ton. Some of my takeaways—and I’m sure listeners would agree—are that “human in the loop” is key. The crawl-walk-run framework is important because it doesn’t mean you have to change everything overnight. It’s about experimentation, trying these frontier models, and seeing what happens.
Steve Hutt:
The fact that you can get video from a couple of shots on your phone and quickly build a full roster of content is incredible. When I look at your case studies, the quality is phenomenal, and I know it’ll keep improving.
Steve Hutt:
I want to thank you, Sam, for coming on the show and sharing. You’re clearly building something impactful for brands, and I think they need to try it. Again, that link is dreem.ai/fastlane—it’ll be in the show notes. Thanks for coming on.
Steve Hutt:
Any parting words for those listening?
Sam Thomas-Davies:
I’d say: think bigger about what you have today and know that it’s within reach. You can try a tool like Dreem and see how easy it is to get more and better content for all your channels and products. And thank you for listening.
Steve Hutt:
My pleasure. Have a great afternoon. Well, that’s it for today’s episode. I’d like to thank you personally for being a loyal listener of eCommerce Fastlane. My hope is that this podcast offers you a ton of value through growth strategies, tactics, and exclusive insider tips on the best Shopify apps and marketing platforms—all with my goal of helping you build, manage, grow, and scale a successful and thriving company powered by Shopify.
Steve Hutt:
Thanks for investing some time today and listening to the show. I’m proud and excited that you have a growth mindset and are a constant learner. I truly appreciate you and your entrepreneurial journey. Enjoy the rest of the week and keep thriving with Shopify.
