Latest / Marketplace Businesses with Fexingo: Two-Sided Networks, Liquidity, and Take Rates / How Marketplaces Use Dynamic Pricing to Balance Supply and Demand
Transcript
- Lucas: So let's talk about dynamic pricing in marketplaces — maybe the most misunderstood lever in the two-sided playbook. Luna: Surge pricing, you mean. The thing everyone loves to hate. Lucas: Exactly. But the reason it exists isn't greed — it's survival. A marketplace lives or dies by liquidity, and dynamic pricing is one of the fastest tools to restore balance when supply and demand get out of whack. Luna: Give me a concrete example. How does Uber's algorithm actually decide to multiply the fare by two point five versus one point five? Lucas: It's a real-time auction, essentially. When the number of ride requests in a zone exceeds available drivers, the algorithm increases the multiplier. That higher fare does two things: it discourages some marginal demand — people who don't really need a ride right now — and it pulls more drivers toward that area because they see the higher payout. Luna: So it's pricing the demand downward and the supply upward at the same time. Lucas: Precisely. And it has to happen in seconds. Uber's system recalculates every few minutes based on real-time data. The goal isn't to maximize profit on every trip — it's to get the market back to equilibrium so that someone who is willing to pay gets a ride, and a driver who is willing to drive gets a fare. Luna: But what about the backlash? Remember when Uber surge-priced during the Sydney hostage crisis in 2014? That created a PR nightmare. Lucas: Right, and that's the ethical boundary. Uber later introduced 'price caps' during emergencies — they basically disable surge pricing in declared disaster zones. But the challenge is defining what counts as an emergency algorithmically. It's still an imperfect system. Luna: Let's step back. Is dynamic pricing only for ride-hailing, or do other marketplaces use it too? Lucas: Oh, it's everywhere. Airlines have been doing yield management since the 1980s — they charge more for the last few seats on a plane because demand is inelastic at that point. Hotels do the same with room rates. But those are mostly one-sided: the seller sets the price. In a true two-sided marketplace, the platform adjusts the price or the commission to influence both sides. Luna: Like Airbnb's Smart Pricing feature. Lucas: Exactly. Airbnb gives hosts an algorithm that recommends nightly rates based on local demand, seasonality, nearby events, and how quickly similar listings book. Hosts can override it, but many don't. The data shows that listings using Smart Pricing book more often and at higher total revenue over time. Luna: So it's a nudge, not a mandate. Unlike Uber, where the driver doesn't set the price. Lucas: Right. And that difference matters for trust. On Uber, drivers accept trips based on the surge multiplier — they see it upfront. On Airbnb, the host sees the recommended price and can say no. The platform has to calibrate the nudge so it feels helpful, not coercive. Luna: What about take rates? Does dynamic pricing affect the platform's commission? Lucas: It can. Some marketplaces charge a percentage of the transaction, so when the price goes up, the platform makes more in absolute terms. But there's a smarter way: variable take rates. OpenTable, for example, charges restaurants a lower per-diner fee for off-peak reservations and a higher fee for prime-time slots. That incentivizes restaurants to offer more availability when demand is low. Luna: So the platform is basically using its own fee structure as a pricing signal. Lucas: Exactly. And this gets to the core of marketplace design. The take rate isn't just a way to make money — it's a lever to shape behavior. If you want more supply during a certain hour, you lower your take for that hour. If you want to discourage demand during a surge, you raise the price you pass to the buyer. Luna: Speaking of levers — we've been talking about this like it's purely economic. But isn't there a psychological component? How do users perceive fairness? Lucas: Huge. Behavioral economics research shows that people are more accepting of surge pricing if they understand why it's happening. Uber added a feature that shows the multiplier but also the reason — 'demand is high in your area.' That transparency reduces friction. On the flip side, if prices change too frequently or unpredictably, users lose trust. Luna: So transparency is the band-aid for the fairness problem. Lucas: Partially. Another approach is to cap surge multipliers. Some cities have regulations capping ride-hail surge at, say, two times the base fare. That protects consumers but can backfire if drivers decide it's not worth their time during peak hours. The market then faces a supply shortage anyway. Luna: It's a tightrope. Which marketplace do you think does dynamic pricing best right now? Lucas: I'd give the nod to Uber, just because of scale and speed. Their system processes millions of data points per minute — driver location, rider demand, traffic, weather, even local events scraped from ticket vendors. It's incredibly sophisticated. But if I had to pick a dark horse, I'd say Doordash. Their delivery marketplace has even more variables: restaurant prep time, driver availability, order complexity. They're doing real-time pricing on the delivery fee and also on the 'peak pay' bonus to drivers. Luna: Honestly, if today was worth a coffee to you, that's the link — buy me a coffee dot com slash fexingo. Lucas: Yeah, listener support is what keeps this show ad-free and independent. If these breakdowns help you think about your own business or marketplace, consider tossing a few bucks in the virtual tip jar. Luna: No pressure, just a simple way to say thanks. We're back to dynamic pricing. Lucas: So another angle I want to touch on: can dynamic pricing be used on the supply side only, without changing the price the buyer sees? Luna: You mean subsidizing the supplier. Lucas: Yes. Some marketplaces boost the payout to suppliers during low-demand periods without raising the price to buyers. Uber does this with 'guaranteed minimum earnings' during slow hours. It's a form of dynamic pricing that doesn't hit the buyer's wallet. The platform eats the cost to maintain liquidity. Luna: That makes strategic sense if the platform has margin to spare or views the short-term cost as an investment in retaining suppliers. Lucas: Exactly. And this is where the financial structure matters. A venture-backed marketplace can afford to subsidize liquidity for years. A bootstrapped one needs to be more surgical. But the principle is the same: the price signal is the most direct way to influence behavior on both sides. Luna: Let's talk about a failure. Any marketplace that messed up dynamic pricing badly? Lucas: One that comes to mind is Ticketmaster. They've been doing dynamic pricing for years, but in 2022, during the Taylor Swift presale, their system essentially priced out huge numbers of fans, and the backlash was enormous. The problem wasn't the mechanism — it was that the algorithm didn't account for brand loyalty and the emotional cost of losing access. That's a lesson: dynamic pricing works best for commoditized transactions, not for emotionally charged ones. Luna: So context matters. You can't just slap an algorithm on every marketplace. Lucas: Right. The best marketplaces segment. They use dynamic pricing only where price sensitivity is high and switch to fixed pricing where customers expect consistency. Amazon does this — prices fluctuate on commodity goods but stay steady on Amazon Basics and Prime Video. Luna: What about the future? Machine learning must be making dynamic pricing even more granular. Lucas: Absolutely. We're moving toward personalized pricing — where different users see different prices based on their willingness to pay. That's already happening in some B2B software marketplaces. But it's controversial. If a user realizes they're being charged more because they're in a wealthy neighborhood, that erodes trust. The successful marketplaces will be those that use dynamic pricing to optimize overall liquidity, not to maximize short-term revenue per user. Luna: So the takeaway is: dynamic pricing is a powerful tool, but it has to be wielded with transparency and fairness. Lucas: And it has to be tied directly to liquidity. If a price change doesn't measurably improve the balance of supply and demand, it's just noise. The best operators run constant experiments — A/B testing surge levels, commission adjustments, minimum guarantees — to find the sweet spot where the market clears efficiently and both sides feel valued. Luna: Alright, that's a solid framework. Next time maybe we can dive into how marketplaces handle the flip side — what happens when demand collapses, not spikes. Lucas: That's a great idea. We'll look at how platforms use price floors, subsidies, and retention campaigns to keep supply alive during downturns. For now, I'm Lucas. Luna: And I'm Luna. Thanks for spending time with Marketplace Businesses with Fexingo.