Latest / Marketplace Businesses with Fexingo: Two-Sided Networks, Liquidity, and Take Rates / How Marketplaces Use Bundling to Raise Take Rate Without Losing Users
Transcript
- Lucas: So we are hitting episode 100 of this show, which honestly feels like a milestone I wasn't sure we'd reach when we started. And I wanted to mark it with an angle that's really core to marketplace economics but one we haven't dug into yet: bundling. Luna: Bundling as in, like, selling multiple things together for one price? That feels almost too simple for a marketplace strategy. Lucas: It seems simple on the surface, but the mechanics are actually subtle. And when done right, bundling lets a marketplace raise its take rate — the percentage it keeps from each transaction — without the user feeling like they're paying more per transaction. That's the magic trick. Luna: Okay, give me a concrete example. Because I think of bundling as a consumer retail thing — like a value meal at a fast-food chain. Lucas: Right, but marketplaces do it differently. Take Instacart. They launched Instacart Plus — a monthly or annual subscription that gives you free delivery on orders over a certain amount. That's a bundle: you pay a flat fee, and in return you get free delivery on every order for a month. The user sees it as a discount. Instacart sees it as a way to increase the average revenue per user without charging per-delivery fees that would make the platform feel expensive. Luna: But isn't that just a subscription, not a bundle? They're not really selling different products together. Lucas: That's a fair distinction. But Instacart took it further. They recently added Peacock streaming to the Plus membership. Now the bundle includes grocery delivery and a streaming service. From the user's perspective, it's more value. From Instacart's perspective, they're effectively bundling a low-margin service — delivery — with a zero marginal cost digital service — streaming. The user pays a single fee, and Instacart captures a higher take rate across the user's entire lifetime value, because the user is more locked in. Luna: So the bundling here isn't just about combining transactions on the marketplace itself. It's about combining the marketplace service with something else entirely. That's clever. Lucas: Exactly. And it works because the marginal cost of adding the streaming service is near zero for Instacart. They're licensing it, but the cost per user is tiny compared to the subscription revenue. So their blended take rate — revenue divided by total transaction value — goes up, without the user paying more per grocery order. The user feels like they're getting a deal, and Instacart makes more money. It's a win-win if the bundle is priced right. Luna: Before we go deeper, I want to mention something that's been on my mind. We've now done a hundred episodes of this show, and we've never run an ad. Not one. And that's only possible because a small group of listeners chips in monthly through buy me a coffee dot com slash fexingo. Lucas: Yeah, it's humbling. We don't do sponsors because we want to keep the conversations honest — no one telling us to plug a product. And the people who support us on Buy Me a Coffee make that possible. So if you find this show useful for what you're building or running, that's a way to keep it going. And seriously, thank you to everyone who already does. Luna: Okay, back to bundling. You mentioned Amazon Prime earlier — that's probably the most famous marketplace bundle of all time. Lucas: Prime is the blueprint. Amazon bundles fast shipping — which has a real cost — with video streaming, music, ebooks, and now even pharmacy discounts. The annual fee is $139. For heavy Amazon shoppers, that fee is easily recovered through free shipping. But Amazon's take rate on each transaction doesn't change — they still take their standard commission from third-party sellers. What changes is that Prime members buy more frequently and spend more per year. So Amazon's blended take rate on the user's lifetime value goes up, because the subscription revenue is pure profit after covering the cost of shipping. Luna: So bundling increases take rate by increasing the total revenue per user, not by raising the per-transaction fee. But does it always work? I feel like some bundles feel like a tax — like you're paying for things you don't want. Lucas: That's the risk. If the bundle includes services the user doesn't value, they'll perceive it as a price increase. The key is to bundle complementary services that the user would likely buy separately anyway. Amazon's genius was starting with free shipping — the thing every online shopper wants — and then layering on entertainment that has a high perceived value but low marginal cost. The bundle becomes a no-brainer because the shipping alone is worth the fee. Luna: Instacart adding Peacock is a similar logic — if you already buy groceries online, free delivery is the anchor, and Peacock is a bonus. But what about marketplaces that don't have a subscription model? Can they still bundle? Lucas: Absolutely. Fiverr is a great example. Fiverr is a marketplace for freelance services — logo design, voiceovers, programming. They don't charge a subscription. Instead, they bundle services into packages. A basic gig might be $50 for a logo. But they offer a standard package at $100 that includes the logo plus source files and faster delivery. And a premium package at $200 that adds social media kit and unlimited revisions. Each tier bundles more services, and Fiverr's take rate is built into each tier — they take a percentage of the total. By bundling, they encourage users to spend more per transaction, which raises Fiverr's revenue per order without increasing the commission rate. Luna: So the bundling happens on the seller side. The seller offers these tiers, and Fiverr takes a cut of the higher-priced bundles. The seller benefits because they can upsell. Fiverr benefits because the take rate per order grows. The buyer feels like they're getting more value. That's a triple win. Lucas: Exactly. And the key metric here is blended take rate. Let's say Fiverr takes 20% on a $50 gig — that's $10. But if they can nudge the buyer to a $200 package, Fiverr's commission jumps to $40 — a 300% increase — even though the take rate percentage stayed the same. Blended take rate across all transactions goes up because the average order value goes up. Luna: But doesn't this depend on the seller being good at creating those bundles? Not every seller will naturally know how to tier their services. Lucas: That's where the marketplace can help. Some platforms provide templates or data on what sells. Fiverr actually recommends package structures based on what similar sellers have found successful. The marketplace acts as a bundling consultant. And over time, sellers who bundle well get higher visibility, because the platform's algorithm favors higher revenue per transaction. Luna: So the marketplace is essentially incentivizing bundling to increase its own take rate. That's smart, but it could backfire if users feel like they're being pushed into overpriced packages. Lucas: It can, if the bundles are poorly designed. But when done well, bundling reduces the user's decision friction. Instead of choosing multiple add-ons, they pick a tier. And the perceived value of the bundle — 'I get this, this, and this for one price' — often exceeds the sum of the parts. Behavioral economics calls this the 'bundle effect.' Users are willing to pay more for a bundle than they would for the individual items, because the bundle feels like a deal even when it's not. Luna: That's a bit manipulative, isn't it? I mean, marketplaces are using psychology to get users to spend more. Lucas: I think it depends on transparency. If the bundle genuinely saves the user money compared to buying separately, it's value. If it's just a way to hide price increases, users will eventually catch on. The marketplaces that succeed with bundling long-term are the ones where the bundle offers real savings or convenience. Amazon Prime has survived for nearly 20 years because it delivers actual value. Instacart Plus works because heavy users do save on delivery fees. The bundling isn't a trick — it's a pricing model that aligns incentives when the economics are right. Luna: Let's talk about the economics more precisely. When does bundling make sense for a marketplace? You mentioned low marginal cost. What else? Lucas: There are three conditions I look for. First, the marketplace should have high user engagement — the user transacts frequently. Bundling only works if the user will use the bundle enough to perceive value. A monthly subscription for free shipping only makes sense if you order at least twice a month. Second, the marginal cost of adding the bundled service should be low — ideally near zero. That's why digital services like streaming or storage are common bundle add-ons. Third, the bundle should be difficult to replicate by competitors. If a competitor can offer a similar bundle easily, it becomes a commodity and price competition erodes the take rate. Luna: So a marketplace like Uber, where rides are relatively infrequent and the marginal cost of a ride is high — that would be a bad fit for bundling? Lucas: Right. Uber has tried bundles like Uber Pass — a subscription for discounted rides and Uber Eats delivery. But rides are high variable cost — Uber pays the driver for each trip. So the bundle discount eats into Uber's margin. And users don't take rides daily like they order groceries. So the perceived value is lower. Uber Pass exists, but it's not as sticky as Prime. Luna: Interesting. So the archetypal bundle-friendly marketplace is one with frequent, low-marginal-cost transactions. That sounds like digital goods marketplaces, content platforms, or advertising-based marketplaces. Lucas: Exactly. Think about a marketplace like Envato — they sell digital assets like stock photos, templates, and fonts. The marginal cost of delivering a digital file is essentially zero. They could bundle a subscription that gives you a certain number of downloads per month. That's a bundle. And the take rate on subscription revenue is much higher than on individual sales, because the customer is locked in and the cost to serve is near zero. Luna: What about a marketplace that's primarily physical goods, like Etsy? Could they use bundling effectively? Lucas: Etsy is interesting because the products are unique and made by individual sellers. Bundling across different sellers is hard because each seller sets their own prices. But Etsy could bundle services — like offering a 'seller plus' subscription that includes promoted listings, advanced analytics, and a custom domain. That's a bundle of marketplace services, not products. And it raises Etsy's take rate by converting occasional sellers into committed subscribers. Luna: So the bundle doesn't have to involve the core transaction — it can be surrounding services. That's a different kind of bundling. Lucas: Right. And many marketplaces use this to grow revenue without touching the commission structure. eBay has a managed payments service that bundles payment processing with seller protection. They take a fee for that bundle. It's not a product bundle, but it's a service bundle that increases the platform's take rate per transaction. Luna: I want to go back to the user perspective for a second. As a user, I sometimes feel like bundles force me to pay for things I don't need. Like, I only use Amazon for shipping, not for video. So effectively, I'm subsidizing the video service. Isn't that inefficient? Lucas: It can be, if the bundle is priced high enough that it's not worth it for a subset of users. But Amazon's data shows that most Prime members use at least two of the included services. And the ones who only use shipping still benefit because the shipping cost they save exceeds the annual fee. The bundle is priced so that even the lightest user gets value. That's the art — pricing the bundle at a point where the marginal user still feels it's a deal, while the heavy user feels like they're stealing. Luna: So the pricing has to be low enough to include the low-engagement users but high enough to make the economics work for the platform. That's a tight balance. Lucas: It is. And that's why marketplaces often start with a single-service subscription before adding bundles. Amazon started with free shipping alone. Then they added video as a bonus when the user base was large enough that the cost per user for licensing content was manageable. Instacart started with delivery fee waivers, then added Peacock later. The bundle evolves as the marketplace's margins and user base grow. Luna: Do you have a favorite example of a marketplace that does bundling poorly? To see the failure mode? Lucas: I think of the early days of the Google Play Music subscription, which bundled with YouTube Premium. The bundle was confusing — users didn't understand what they were getting. And the price wasn't clearly better than buying the services separately. Also, the services were not tightly integrated. A good bundle feels like a single product, not a collection of unrelated things. Google's bundle felt like a loose grouping. Luna: So the integration matters. If the services don't feel like they belong together, the bundle seems arbitrary. Lucas: Exactly. The best bundles are seamless. When you open Amazon, you don't think about Prime Video as separate — it's just part of your account. When you order on Instacart, you don't think about Peacock — it's just a perk. The bundle should fade into the background. Luna: For our listeners who are building marketplaces, what's the one actionable takeaway from this episode? Lucas: I'd say: look at your data and find the transaction that has the highest frequency and the lowest marginal cost. That's your bundle anchor. Then add a service on top that has high perceived value but low cost to you. Test it as a subscription or a tiered package. Measure the blended take rate before and after. If it goes up without causing churn, you're on to something. Luna: And if it causes churn, you're charging too much or bundling the wrong things. Lucas: Right. Bundling is a powerful lever, but it's not a panacea. It works best when the marketplace already has a loyal user base and the economics support it. For early-stage marketplaces, focus on liquidity first. Bundling comes later, when you have enough data to know what your users value. Luna: That feels like a good note to end on. Episode 100 — thanks for sticking with us. Lucas: And thanks to everyone who supports the show on Buy Me a Coffee. It really makes a difference.