Latest / The DTC Marketing Podcast with Fexingo: Direct-to-Consumer Brands, Shopify, and E-Commerce Growth / How DTC Brands Use AI-Powered Dynamic Pricing
Transcript
- Lucas: So Luna, I wanted to talk about something every DTC brand deals with but few do well — pricing. Specifically, dynamic pricing powered by AI. Luna: You mean like Uber surge pricing, but for T-shirts? Lucas: Exactly, but hopefully less rage-inducing. The idea is that instead of setting a price and leaving it, you adjust in real time based on demand, inventory levels, competitor moves, even time of day or weather. And AI makes that actually feasible for mid-size brands now. Luna: I feel like consumers have a love-hate relationship with this. When Amazon changes prices throughout the day, it feels efficient. When a hotel does it, it's just normal. But when a small brand does it, people can get suspicious. Lucas: Right, that's the key tension. Let's look at a specific case. I was talking to the founder of a mid-tier DTC apparel brand — let's call them 'Coastal Threads' — they do about $15 million a year on Shopify. They started using an AI pricing tool last year. Before that, they had maybe four price tiers: basic, premium, sale, and clearance. The tool now adjusts prices on about 60 percent of their SKUs based on signals. Luna: Sixty percent? That's aggressive. What kind of signals? Lucas: Demand velocity — if a particular style starts selling fast, the price creeps up a few percent. Inventory depth — if they have too many of a certain size, that size gets a small discount. Competitor price changes — if a similar brand drops a comparable hoodie, the algorithm might match or slightly undercut. And seasonality — a jacket in July is cheaper than in November. Luna: And the results? Lucas: Over six months, their gross margin went from 52 percent to 58 percent. Revenue stayed flat — they didn't juice sales — but they made more money on each sale. And they didn't get a single customer complaint about pricing. Because the changes were small — three to five percent moves — and they never changed the price after someone added an item to cart. Luna: That last point is crucial. The worst thing you can do is change the price during checkout. That's how you lose trust. Lucas: Absolutely. They made sure the price shown on the product page is the price you get. No surprise hikes. And they actually tested a version where they showed a little badge: 'Price adjusted based on demand — 20 units left at this price.' That version had slightly higher conversion. People saw it as scarcity, not gouging. Luna: So there's a transparency angle that actually helps. But what about the flip side — brands that do this poorly? Lucas: Oh, plenty. I saw a beauty brand that jacked up prices on their best-selling moisturizer by 20 percent during a shortage. People noticed on Reddit. A thread called them out. They had to apologize and freeze prices. The algorithm didn't have a 'reputation guardrail'. Luna: So you need to set rules. Like a maximum price increase per day, or a floor so you're not going below cost accidentally. Lucas: Exactly. The smart brands layer in business rules on top of the AI. For example, never increase price more than 10 percent above the 30-day average. Or never drop below a 40 percent margin floor. The AI suggests, the brand decides the guardrails. Luna: What about the competitive angle? If every brand in a category uses dynamic pricing, doesn't it just lead to a race to the bottom? Lucas: It can, but only if everyone's algorithm is optimizing for the same thing — usually market share. The smarter algorithms optimize for profit per customer over lifetime. So you might see a brand charge a returning customer a slightly higher price because they know loyalty is high, while offering a discount to new customers. That's price discrimination, but it's common — think 'new subscriber' discounts. Luna: That feels sketchy if it's not transparent. But I guess the industry is moving that way anyway. What tools are brands like Coastal Threads actually using? Are we talking about custom models or off-the-shelf Shopify apps? Lucas: Both. There are Shopify apps like Prisync, Omnia, and Price2Spy that do competitor-based repricing. But for true AI dynamic pricing with demand forecasting, you're looking at something like Competera or Reactev. Those integrate with Shopify and use machine learning to predict optimal price points. Some brands build their own if they have data science teams, but most don't. Luna: And the cost? Is it accessible for a brand doing, say, one or two million a year? Lucas: It's getting there. The tools start around $500 a month, which is meaningful but not insane if it lifts margin by a few points. For a two-million-dollar brand, a 3 percent margin lift is sixty thousand dollars a year. So the return is real. Luna: Let's talk about the ethical boundaries. Some critics say dynamic pricing exploits loyal customers who don't comparison shop. If a repeat buyer sees a higher price than a new visitor, that feels punitive. Lucas: I think the line is drawn at deception. If you're charging different prices based on browsing history or device type without disclosure, that's bad. But if you're adjusting based on clear signals like inventory or time to expiry, and you're transparent about it, most customers accept it. The key is framing — say 'limited stock at this price' rather than 'we're charging you more because you're on a Mac.' Luna: Have you seen any brands successfully use dynamic pricing for subscription boxes? That model usually has fixed prices. Lucas: Interesting question. I know a coffee subscription brand that offers a 'dynamic' price per bag based on the origin's harvest yield. If a single-origin Ethiopian bean was scarce that season, they'd charge a premium, but they communicated it — 'This year's crop is smaller, so the price is up 15 percent.' Subscribers actually appreciated the transparency. It felt like they were getting the real story. Luna: So it's about narrative. If you just change a number, it feels arbitrary. If you explain why, it feels fair. Lucas: Exactly. And that brings me to a key recommendation for any brand considering this: start with a pilot on a small segment. Maybe just your top 10 SKUs, or just during a specific time window like Black Friday weekend. Measure customer feedback, conversion rates, and margin. Then iterate. Luna: And what about the technology itself? Are we seeing AI models that can factor in things like weather or social media sentiment? Lucas: Yes. More advanced tools now ingest weather data — if it's going to be rainy, a raincoat brand can increase price because demand will rise. Or if a celebrity wears a certain style, the algorithm can detect the spike in social mentions and adjust accordingly. It's getting pretty sophisticated. Luna: I also wonder about the psychological effect on the brand team. If the algorithm changes prices every hour, do they lose a sense of their own brand positioning? Lucas: That's a real risk. I've talked to founders who felt like they lost control. Their brand was supposed to be 'premium' but the algorithm kept discounting to chase sales. So you need a pricing philosophy that the algorithm can't override. For example, 'we never discount more than 20 percent off MSRP' or 'our core collection is always full price.' Luna: So it's a tool, not a replacement for strategy. Lucas: Yeah. The best use I've seen is for inventory management and seasonal clearance. A fashion brand I know uses dynamic pricing to automatically reduce prices on slow-moving sizes and colors, while keeping bestsellers at full price. That cleared out dead stock 40 percent faster than their old manual markdown process. Luna: That's a concrete win. And it doesn't feel scammy because it's just smarter inventory management. Lucas: Exactly. And if these marketing conversations have sparked something you've actually used — a strategy, a tool, a way of thinking — we'd love it if you considered supporting the show. A couple of dollars a month is genuinely what keeps these going. Buy me a coffee dot com slash fexingo, if you've gotten something out of them. Luna: Yeah, it really does make a difference. Every contribution helps us keep the episodes ad-free and focused on what matters. Lucas: So back to dynamic pricing — the big takeaway for me is this: the technology is ready, the adoption is rising, but the winners will be the brands that combine AI with a clear ethical framework and a strong brand narrative. Don't let the algorithm dictate your values. Luna: And don't forget to test, measure, and iterate. It's not a set-it-and-forget-it thing. Lucas: Right. Next time, we'll look at how DTC brands are using AI to personalize the entire shopping experience — not just pricing, but product recommendations, site layout, email content. Luna: Looking forward to it. Thanks, Lucas.