Latest / The CMO Podcast with Fexingo: Marketing Leadership, Budgets, and Executive Strategy / How CMOs Are Using AI for Real-Time Pricing Strategies
Transcript
- Lucas: So we talk a lot on this show about AI for creative work, for targeting, for personalization. But there's one area where AI is actually changing the P&L in a very direct, measurable way, and it's not getting nearly the same hype. Luna: Let me guess — pricing. Lucas: Exactly. Real-time pricing powered by machine learning. Not just 'let's run a coupon code' dynamic pricing, but algorithms that adjust prices every few minutes based on demand, competitor moves, inventory, even weather. CMOs are starting to own this because pricing is increasingly a customer experience decision, not just a finance one. Luna: I think a lot of people hear 'dynamic pricing' and immediately think of Uber surge pricing or that time Coca-Cola tried a vending machine that raised prices in hot weather. Both of those backfired badly. Lucas: Right, and that's the cautionary tale. But the technology has gotten a lot more sophisticated since 1999. Let's take a concrete example — a major US airline. They've been using a machine learning model that ingests hundreds of variables: how many seats are left on a flight, what competitors are charging for similar routes, historical booking patterns, even the day of the week and time of departure. The model then recommends price adjustments every 15 minutes for each fare class. Luna: And the CMO's role in that? Because traditionally that would be revenue management. Lucas: Traditionally, yes. But what we're seeing is that as pricing becomes more personalized and more frequent, it intersects directly with brand perception. The CMO at that airline told me they now have a veto over any pricing change that deviates more than 20 percent from the published fare, because they need to protect the brand from looking predatory. So the algorithm proposes, but marketing has the final say. Luna: That's a fascinating governance model. So the AI is optimizing for revenue, but the CMO is optimizing for trust. Lucas: Exactly. And it's not just airlines. I spoke with a mid-size retailer — think home goods — that implemented ai driven markdowns last year. Previously, they would run seasonal sales based on calendar dates. Now, their system looks at sell-through rates, competitor promotions, and even local weather forecasts to decide when to discount and by how much. They saw a 12 percent lift in gross margin on clearance items in the first quarter. Luna: Twelve percent is huge. But how do customers react when they realize the price they paid might be different from someone else's? There's a real transparency problem. Lucas: That's the number one concern I hear from CMOs. And the solution so far seems to be transparency with guardrails. Some companies are experimenting with showing customers a 'price confidence' indicator — like 'this price is guaranteed for the next 24 hours' — so people don't feel cheated. Others are using AI to personalize discounts upward rather than downward. So instead of charging a loyal customer more, you offer a new customer a lower price to acquire them. Luna: That's a much smarter framing. It's not about charging some people more; it's about charging some people less. Lucas: Right. And the AI makes that scalable. You can test thousands of price points simultaneously across different segments and see which ones drive conversion without damaging brand equity. A travel company I know runs continuous A/B tests on hotel room pricing — they show different prices to different user cohorts and measure not just booking rate but also repeat visit rate and review sentiment. Luna: So the CMO is essentially becoming the chief experience officer for pricing. They're responsible for how the price feels, not just what the number is. Lucas: That's a perfect way to put it. And it requires a completely different skill set than the old model of setting prices once a year and then running promotions. Now you need data scientists, behavioral economists, and UX designers all at the table. Luna: What about the B2B side? I imagine real-time pricing is trickier when you have negotiated contracts and volume discounts. Lucas: It is, but it's happening there too. A software company I talked to uses AI to optimize their tiered pricing. The algorithm looks at usage patterns, feature adoption, and customer health scores to suggest personalized upsell offers at the exact moment a customer is most likely to upgrade. Their CMO said it doubled the conversion rate on expansion revenue. Luna: That's a smart use case because it's value-based, not just demand-based. The customer gets a higher price when the product is delivering more value. Lucas: Exactly. And that's the ethical line a lot of CMOs are drawing. If the price increase is tied to demonstrable value, it feels fair. If it's just because you have low inventory and the customer has no alternatives, that's where you get into trouble. A couple of CMOs told me they've hard-coded 'fairness constraints' into their pricing models — like never pricing a necessity above a certain multiple of the cost, or never increasing price by more than 10 percent in a single transaction. Luna: You know, if these marketing conversations have sparked something you've actually used — maybe a pricing test, a new way of thinking about customer trust — we'd love for you to support this show. A couple of dollars a month is genuinely what keeps these going, and it means we can keep them ad-free. Lucas: Yeah, it's a small thing that makes a big difference for us. If you've gotten value from the podcast, you can head to buy me a coffee dot com slash fexingo. That's buy me a coffee dot com slash fexingo. And honestly, it's listener support like that that lets us dive deep into topics like this without worrying about sponsors. Luna: So back to those fairness constraints — how are companies actually enforcing them? Is it a technical thing or a policy thing? Lucas: Both. Technically, you can code the model to reject any price recommendation that violates a rule. But the more interesting layer is the policy — some CMOs are creating a pricing ethics board that meets monthly to review the algorithm's decisions and adjust the rules. One retail CMO told me they actually publish an annual 'pricing transparency report' that explains how their dynamic pricing works and what safeguards are in place. Luna: That's a bold move. It turns a potential PR liability into a trust-building asset. Lucas: Exactly. And I think that's where the smartest CMOs are heading. They're not just deploying AI to optimize revenue; they're building the governance and communication around it so customers feel informed rather than manipulated. The technology is moving fast, but the competitive advantage might actually come from how you manage the customer's perception of fairness. Luna: So for a CMO listening who wants to start exploring this — what's the first step? Lucas: Don't start with the algorithm. Start with the data. You need clean, real-time data on transactions, inventory, and customer behavior. Without that, the AI will just give you bad recommendations faster. Then run a small pilot — pick one product category or one customer segment — and measure the impact on both revenue and customer satisfaction. If the satisfaction metric doesn't move in the right direction, you're not ready to scale. Luna: That's a pragmatic approach. And it avoids the Coca-Cola vending machine problem. Lucas: Exactly. Learn from history. The technology is better now, but the human psychology hasn't changed. People still want to feel like they got a fair deal. If AI helps you deliver that at scale, you've got a real advantage.