Price Discrimination within the Blockchain
Firms that sell digital goods and services must differentiate their products and obtain some market power, in order to be able to price discriminate and charge above marginal cost, which in the digital world is very close to zero. In this note, we explore the most common types of price discrimination and discuss the opportunities and restrictions that arise when sales occur within the blockchain.
When firms formulate pricing strategies for digital goods and services, they are bound by the following cost structure: the fixed costs are high, because the first unit is expensive to produce, but the marginal cost is almost zero, as digital goods are very cheap to reproduce. For example, developing the new version of Office 365 may be very expensive, but selling one extra unit costs essentially nothing.
In an environment with perfect competition, this cost structure is unsustainable for firms, because the price must be equal to the marginal cost, which means that prices will eventually go to zero. Therefore, firms that offer digital goods and services need to differentiate their products and obtain some market power, in order to be able to charge over and above their marginal cost. Their competitive advantage, as compared to brick-and-mortar stores, is that they are much more efficient at collecting information about the digital presence and online behaviour of their customers, hence they can better predict their willingness to pay and differentiate prices accordingly.
In this note, we describe the three forms of price discrimination, and we elaborate on how they can be tailored for digital goods and services that are offered within the blockchain.
Price discrimination refers to the practice of charging customers different prices for the same good or service. The degree of price discrimination depends on the firm’s effectiveness in predicting how much each customer is willing to pay and its ability to segregate customers into different markets. We consider three degrees of price discrimination.1
First Price Discrimination
First price discrimination is personalised pricing, as the firm can charge each customer a different price. E-commerce firms have the ability through cookies to track the behaviour of costumers in the digital world and therefore gather a lot of information about their tastes and willingness to pay. Moreover, they can customise prices instantly, depending on who is buying, therefore segregating customers into different markets. The advent of the blockchain poses both risks and opportunities for these pricing strategies. On the one hand, a customer can connect anonymously in the blockchain through their wallet in order to buy a digital good or service, thus restricting the information that is available to firms. On the other hand, because the transaction history of any wallet is public, the seller has more direct information about their customer. In the future, the NFTs that a wallet holds may reveal a lot about the customer’s spending habits, communities they participate in, and therefore their willingness to pay, in the same way that now information about which supermarket they shop. Hence, it may be that firms have even more information about their customers. The main difference is that the customer will have more control over what information they share with firms, which now is not the case with cookies.
Second Price Discrimination
Second price discrimination refers to the practice of offering a menu of different versions and allowing customers to choose the one they want. A classic example of versioning is offering an economy, business and first class for airplane tickets. The key to being successful is that the extra cost of offering the premium services is much less than the price difference from the basic service. In the blockchain, an early and successful example is NFT collections, which release multiple NFTs at once.2 Although each NFT costs the same to produce, they differ in terms of their characteristics, establishing a rarity index for NFTs that is easily checkable by the customer. This differentiation creates value for consumers. For example, one can check through the marketplace OpenSea that the Bored Ape Yacht Club collection has 10000 NFTs but only 46 have solid gold fur. The floor price for the collection is currently 69 ETH, but for solid gold fur apes it is 800 ETH.
This example provides a useful insight. An important distinction between digital goods that are not on the blockchain and those that are, like NFTs, is that the latter are actively traded on marketplaces. The implication is that an NFT issuer does not have to predict which characteristics will appeal the most to customers and what is their willingness to pay. Instead, they can release the collection using a uniform price and then benefit from the different prices across NFTs that emerge naturally in the marketplace, through the royalty fees that they set. This means that price discrimination can be more effective as the market gradually reveals the private information of customers about their own tastes, through trading. It is notable that the earnings for the most successful NFT collections are predominantly through royalty fees, rather than from the initial sale.3 Intuitively, because NFT customers are owners, they have a strong incentive to reveal their value when they resell it. On the contrary, a customer who buys a monthly subscription to a music service that cannot be resold, will have an incentive to hide her willingness to pay.
Price discrimination can become even more effective when NFTs are connected with different real-life benefits, for example access to certain events or meeting the artist. In the future, an NFT can give access to certain services and allow for greater customization, for example listening to Classical music on Spotify. This could open up a new avenue for price discrimination and increased revenues for firms that are currently selling digital goods off the blockchain and rely on uniform pricing and no resales. Finally, dynamic NFTs can change their metadata automatically, as a response to events that are communicated in the blockchain. In the future, this characteristic could be used as a way of customizing the NFT according to the buyer’s preferences, allowing for further price discrimination.
Another common practice when second price discriminating is offering discounts for buying multiple items. For example, Office 365 is a bundle of Word, Excel, PowerPoint, and other applications. The price of the bundle is always lower than the price of the individual products. However, bundling can increase revenues if the willingness to pay for the bundle is less dispersed among consumers than the willingness to pay for the components. To provide a simple example, suppose that Ann’s willingness to pay for Word is $120 but only $50 for Excel, whereas Bob’s willingness to pay for Word is $50 but $120 for Excel. If Microsoft prices the two apps individually, the maximum they can get is $240. However, if they create a bundle with both apps they can price it at $170, hence earning $340 in total. The extra revenues from bundling arise because the two consumers are very different in terms of how they value the individual apps, but identical in terms of how they value the bundle. A monthly subscription with unlimited consumption is another example of bundling.
An early form of bundling is emerging in DeFi, where customers can compose their own bundle of financial services from different providers, such as exchanging, lending and borrowing, and staking.4 This is different from the traditional bundles of digital goods, that are generated and controlled by the firms. However, in DeFi the services are currently priced individually. In the future, one can imagine that the financial services providers will enter a revenue sharing agreement that offers discounts to consumers who choose specific bundles.
Third Price Discrimination
Third price discrimination occurs when the firm charges a different price according to group identity. The main justification is that groups differ in terms of their price sensitivity. For example, students and senior citizens are more price sensitive than the average consumer, hence it makes sense to offer them discounts for goods and services. Another dimension is geographical location, usually a good predictor of price sensitivity. A necessary condition for third price discrimination is that resales are prevented between groups.
In the blockchain, characteristics such as age and location may be more difficult to establish, because wallet addresses are usually anonymous. However, the transaction history and the NFTs that a wallet holds may be equally good at predicting price sensitivity. The main difficulty is that if the firm sells a non-differentiated digital good on the blockchain, then they will not be able to prevent resales, thus negating the usefulness of third price discrimination. A successful pricing strategy would involve selling differentiated goods that can be resold between members of the same group but not across groups.
Price discrimination is inevitable for firms that offer digital products and the blockchain offers both risks and opportunities for the pricing strategies of firms. On the one hand, anonymity of wallets may make it harder for firms to gather information in order to predict their willingness to pay, whereas customers may have better control over the information they share about themselves. On the other hand, the tradability of NFTs implies that the firms may have more opportunities to differentiate their products and let the market reveal the customers’ willingness to pay.
1 For a more detailed discussion of pricing digital goods and services, see Shapiro, Carl and Varian, Hal R. “Information Rules: A Strategic Guide to the Network Economy”, Harvard Business School Press, 2010, and “Digital Innovation and Revenue Models”, Prysm Group, 2022, available at https://www.prysmgroup.io/articles/digital-innovation-and-revenue-models.
2 See https://coinmarketcap.com/nft/collections/ for a list of the most successful NFT collections, in terms of value and transactions.
3 Royalty fees are usually between 2-10%. For a simple model of how price discrimination can be more efficient through royalty fees, see “Hemenway Falk, Brett and Tsoukalas, Gerry and Zhang, Niuniu, Economics of NFTs: The Value of Creator Royalties”, 2022, available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4284776.
4 For more details on the composability of DeFi services, see “Decentralized Finance: Efficiency vs. Risk”, Aaro Capital, 2022, available at https://en.aaro.capital/Article?ID=080b7c54-eb49-45ea-b350-f7692b988467.