Latest / Marketplace Businesses with Fexingo: Two-Sided Networks, Liquidity, and Take Rates / How eBay Uses Reserve Prices to Protect Liquidity
Transcript
- Lucas: If you've ever sold something on eBay—maybe an old guitar, a collectible, a vintage camera—you've probably faced the moment where you have to decide: do I set a starting price, or do I let the market decide? Luna: And if you set it too low, you're scared someone will snatch it for a fraction of what it's worth. Too high, and nobody bids. Lucas: Exactly. And eBay's answer to that dilemma is a mechanism that's been around for decades but is surprisingly subtle: the reserve price. A hidden minimum that the seller sets privately, and if bidding doesn't reach it, the item doesn't sell. Luna: So the buyer never sees the reserve—they just see 'reserve not met' if they try to win below that threshold. Lucas: Right. And what's interesting is that reserve prices are a liquidity tool, not just a seller-protection tool. They solve a chicken and egg problem that marketplaces face when the item is high-value or unusual. Without a reserve, many sellers wouldn't list at all—they'd fear a lowball outcome. So the marketplace would lose supply. But with a visible reserve, many buyers would feel like the game is rigged—why bid if you don't know the floor? Luna: So the hidden reserve is like a compromise: the seller gets their floor, the buyer gets to hunt for a deal, and eBay gets the listing. Lucas: Exactly. And this is critical for marketplace liquidity. Let's look at a real example. Say someone lists a vintage Leica M3 camera—a classic rangefinder from the 1950s. In good condition, those can sell for anywhere from $800 to $2,500 depending on lens, wear, provenance. If the seller sets a starting bid at $500 with no reserve, they risk selling a $2,000 camera for $500. So they either start high—say $2,000—and risk zero bids, or they set a low start and add a reserve at, say, $1,800. Luna: And the buyer who sees a starting bid of $500 thinks 'bargain opportunity.' They bid early, drive momentum. Other bidders see activity and join. The reserve stays invisible until someone crosses $1,800. Lucas: Yes. And eBay's data shows that listings with a reserve price sell at a higher average price than those without, because they attract more bidders early. But there's a catch: reserve prices also reduce the overall probability of sale. A study from a few years ago found that reserve-price listings on eBay had about a 15 percent lower sell-through rate than no-reserve listings. So it's a trade-off: higher price conditional on selling, but less certainty. Luna: Fifteen percent—that's meaningful. So the marketplace has to design the feature so that sellers use it wisely, not aggressively. Lucas: Exactly, and eBay does that by charging a fee based on the reserve amount. If you set a reserve above $50, eBay charges a non-refundable fee—something like one percent of the reserve value, capped at a certain amount. That fee disincentivizes sellers from setting unrealistic reserves that waste bidders' time. Luna: So it's a friction point that aligns incentives. If you set a reserve that's too high, you pay for the privilege and still might not sell. Lucas: And that's a clever bit of marketplace design. Because without that friction, sellers could flood the platform with 'fishing expeditions'—listings where they'd only sell at a pie-in-the-sky price. That would frustrate buyers, who'd learn to avoid reserve listings, and liquidity would suffer. Luna: I've definitely been on the buyer side of that. You bid on something, think you're close, and then 'reserve not met.' It feels like a bait and switch. Lucas: It can be. And that's why eBay has iterated on the feature over the years. They used to display the reserve price on certain listings—you could see 'Reserve: $1,800'—but they moved to the current 'reserve not met' messaging because they found that showing the actual number discouraged bidding. Bidders would see a high reserve and think 'why bother?' So hiding it preserves the energy. Luna: There's a behavioral economics angle there: the endowment effect on the seller side, and the sunk cost on the buyer side. Once you've placed a bid, you're more likely to come back and bid again. Lucas: Exactly. It's the same logic that makes auction houses like Sotheby's and Christie's use estimates instead of reserves. They publish a range—say $1,500 to $2,500—and the seller has a confidential reserve that's typically the low end of that range. The published estimate anchors the bidders, and the hidden reserve protects the seller. But the difference is that at Sotheby's, the auctioneer might start the bidding below the reserve to warm up the room. Luna: Right. So the mechanism varies by context. eBay is a digital mass market, Sotheby's is a curated high-end room. Both use reserves, but the presentation is tailored to the audience. Lucas: That's a great point. And it raises a broader question for marketplace builders: when should you use a reserve price, and when should you go no-reserve? The general rule is: if the item has a wide valuation range and the seller is risk-averse, a reserve helps. If the item is commoditized—like an iPhone case—a reserve is unnecessary and actually hurts. Buyers expect a low starting price and quick sale. Luna: So reserve prices are really a tool for the 'long tail' of the marketplace—the rare, the unique, the hard to price items. That's where liquidity is hardest to create. Lucas: Exactly. And that's the part that I think operators sometimes overlook. They focus on the high-volume categories, but the long tail is where marketplaces build their defensibility. eBay's moat isn't in selling new electronics—it's in selling vintage Leicas, used car parts, collectible stamps. And reserve prices are one of the mechanisms that make that long-tail liquidity possible. Luna: If today was actually useful to you, the way these stay ad-free is listener support—buy me a coffee dot com slash fexingo. It's a small gesture that keeps the conversations going without any sponsor noise. Lucas: Yeah, we really appreciate that. And honestly, knowing people find value in these deep dives is what keeps us doing them. So thank you to anyone who's contributed. Luna: Alright, back to reserves. So we've got eBay and Sotheby's. Any other marketplaces that use this mechanic differently? Lucas: Yes—Catawiki, the European auction platform for collectibles, uses a variant. They let sellers set a reserve, but they also have a team of experts who set a 'market value estimate' that's displayed to bidders. The reserve is typically 80 to 90 percent of that estimate. So the bidder sees a price anchor, and the seller gets a floor. Catawiki's take rate is around 12 to 15 percent, higher than eBay's typical 10 percent, because they're curating the supply side. Luna: So the expert valuation substitutes for the trust that a visible reserve might erode. The buyer trusts the expert, not the seller. Lucas: Exactly. And that's a different design choice. eBay's approach is more laissez-faire: let the market decide, but protect the seller with a hidden floor. Catawiki's approach is more interventionist: we'll tell you what it's worth, and we'll enforce a floor near that. Luna: Which approach works better? Is there data on conversion rates? Lucas: Catawiki doesn't publish granular data, but their sell-through rate for reserve listings is reportedly around 70 percent, which is high for collectibles. eBay's overall sell-through for reserve listings is lower, maybe 50 to 60 percent, but eBay's volume is so massive that the absolute number of transactions is huge. The key metric for a marketplace is not just sell-through, but also the ratio of listings to completed transactions. If you have too many unsold reserve listings, buyers get frustrated and leave. Luna: So it's a delicate balance. And the take rate—the fee the marketplace charges—is also affected. If the reserve is too high and the item doesn't sell, the marketplace earns zero. But if no reserve leads to a low price, the seller might not come back. Lucas: Right. And that's where the marketplace has to think about the lifetime value of the seller. A seller who gets burned once—feels they undersold—might never list again. So a reserve price is a cheap insurance policy for the platform to retain supply. eBay's reserve fee is a small price to pay to keep a seller happy and active. Luna: That's a really good framing. It's not just about the single transaction; it's about the flywheel. Happy sellers bring more listings, which attract more buyers, which drives more bids, which drives higher prices, which makes sellers happier. Lucas: That's the marketplace flywheel in action. And reserve prices are one of the small mechanical choices that can either accelerate or brake that flywheel. Get it right, and you build a healthy, self-sustaining ecosystem. Get it wrong, and you get either a ghost town or a bazaar where nobody trusts the prices. Luna: So what's the next innovation? Are we going to see dynamic reserves that adjust in real time based on bidder activity? Lucas: I think so. Some newer platforms like Whatnot, the live-streaming auction app for collectibles, already have a variant. The seller sets a starting price, but the auctioneer—the host—can adjust the reserve on the fly based on chat engagement. If there's a lot of interest, they might raise the reserve mid-stream. It's a more fluid version of the same concept. Luna: Interesting. So the reserve becomes a real-time signal, not just a static number set before the auction. Lucas: Exactly. And that's where marketplace design is heading: more dynamic, more responsive, more personalized. The core tension—protecting the seller without scaring the buyer—remains, but the tools are getting smarter. Luna: Well, and with AI, you could imagine a system that predicts the optimal reserve for each listing based on historical data for similar items, seller reputation, and buyer demand forecasts. Lucas: Totally. eBay already does something like that with their 'suggested start price' feature, but it's not yet a dynamic reserve. I suspect within a few years, we'll see marketplaces automatically set a reserve that maximizes expected revenue or liquidity, and sellers just opt in. That would remove the guesswork and potentially increase overall transaction volume. Luna: That would be a game changer for casual sellers—people who just want to clear out their closet without becoming auction strategists. Lucas: Exactly. And that's the ultimate goal of marketplace design: make it so easy that the mechanism fades into the background, and the transaction feels natural. Reserve prices, when done right, are invisible. You only notice them when they fail. Luna: Maybe that's the sign of good design: the friction you don't see. Lucas: Yeah. The reserve price is a perfect example—a tiny piece of code that has outsized effects on trust, liquidity, and retention. Worth paying attention to.