Theory 01
Classifying Digital Products
Origin: Hui & Chau (2002); Session 3 slides
Hui & Chau (2002) argue that digital pricing strategy depends on the intrinsic characteristics of the product. A software utility, an e-book, and an interactive online service may all be digital, but they differ in how value is delivered, how divisible they are, and whether buyers can try them before paying.
Tools & Utilities
Downloadable software with specific functions (e.g., anti-virus). Generally low granularity and high trialability (Hui & Chau, 2002).
Content-Based
Information goods like news, music, or e-books. High granularity (divisibility), allowing for flexible packaging and pricing (Hui & Chau, 2002).
Online Services
Interactive, real-time solutions (e.g., online therapy). Medium granularity and often charged by usage time or subscription (Hui & Chau, 2002).
Intrinsic Characteristics (Hui & Chau, 2002)
| Attribute | Description |
|---|---|
| Delivery Mode | Downloadable (full product transfer) vs. Interactive (continual basis). |
| Granularity | Divisibility of the product. High granularity allows for vertical differentiation (e.g., selling chapters of a book). |
| Trialability | Ease of providing "free samples" without disrupting the core profit model. |
Theory 02
Price Discrimination in Digital Markets
Origin: Session 3 slides; Shapiro & Varian-style information goods logic
Digital products often have high fixed costs and near-zero marginal costs. That makes pricing less about covering unit cost and more about capturing willingness to pay across heterogeneous customers. The session distinguishes three major forms of price discrimination.
| Type | Definition | Digital Example |
|---|---|---|
| Personalized pricing | Charging each consumer close to their maximum willingness to pay. | Algorithmic offers, individualized discounts, or dynamic subscription offers. |
| Versioning | Designing versions so users self-select based on needs and willingness to pay. | Free, basic, pro, enterprise tiers. |
| Group pricing | Charging different groups different prices for the same or similar product. | Student discounts, enterprise contracts, regional pricing. |
The Digital Pricing Logic
Because the marginal cost of an extra user is often low, firms can profit by serving low-willingness-to-pay users without undermining higher-priced segments. The challenge is to prevent arbitrage and design product differences that make self-selection credible.
Theory 03
Versioning: Designing Self-Selection
Origin: Session 3 slides; Hui & Chau (2002)
Versioning means offering variants tailored to different customer needs. The point is not merely to offer "more choices"; it is to create a pricing architecture where users reveal their willingness to pay through the version they choose.
Market Analysis
Ask whether the market naturally subdivides into segments with different needs, behaviors, and willingness to pay.
Product Analysis
Identify dimensions to version: quantity, quality, speed, features, support, usage limits, ads, integrations, or data access.
Rule of Three
The slides emphasize three versions as a practical design: low-end entry, middle default, and high-end premium.
Installed Base, Rival Customers, and New Customers
Pricing can also vary by relationship status. Existing customers may be offered premium services, rival customers may receive switching discounts, and new customers may receive introductory offers. The design depends on lock-in, information about customers, and whether customers can arbitrage across offers.
Theory 04
Bundling: Reducing Dispersion in Willingness to Pay
Origin: Session 3 slides
Bundling is powerful for digital products because it can increase revenue without much additional marginal cost. The theoretical logic is that bundling can reduce variation in reservation prices across consumers, making total willingness to pay more predictable.
Price Bundles vs. Product/Service Bundles
- Price bundle: Separate products sold together at a discount. The bundle mainly changes price.
- Product/service bundle: Integrated products create additional value together, such as software plus storage plus support.
- Pure bundling: Only the bundle is sold.
- Mixed bundling: Products are sold both separately and as a bundle.
Theory 05
Platform Pricing: Subsidy Side and Money Side
Origin: Session 3 slides
In platforms, the relevant question is not only "what price should we charge?" but which side should pay? One side may be subsidized because its participation creates value for the other side. The platform then monetizes the side with higher willingness to pay or stronger value capture potential.
| Pricing Choice | Logic | Risk |
|---|---|---|
| Free user side | Attract users to make the platform valuable for advertisers, sellers, or developers. | Hard to transition from free to fee without backlash. |
| Charge producer side | Producers pay for access to users, tools, curation, or transaction infrastructure. | High fees can push producers to rival platforms. |
| Transaction fee | Platform takes a percentage when value is realized. | Participants may disintermediate the platform if governance is weak. |
| Premium tools | Basic access stays free, but advanced analytics, visibility, or support is paid. | The paid layer must add new value, not remove value users already received. |
Theory 06
Launch Pricing Under Network Effects
Origin: Lee & O'Connor (2003); Session 3 slides
When a product has network effects, launch pricing must help the firm reach critical mass. The slides contrast classical markets, where price can signal quality and skimming may work, with network markets, where penetration pricing and bundling often support long-term performance.
Skimming vs. Penetration
- Skimming pricing: High launch price to monetize early adopters and signal quality. Better for products where intrinsic value dominates.
- Penetration pricing: Low price or free access to grow the installed base quickly. Better when extrinsic value and network effects dominate.
- Lock-in then change prices: Once users are embedded, firms may shift pricing, but this risks trust and backlash.
- Bundling, compatibility, variety, and upgrades: These help build adoption and keep users inside the ecosystem.
Case Application
Novo Nordisk — Pricing and Capturing AI Value
Session 03 is less central to the Novo-OpenAI synopsis than dynamic capabilities, but it is useful for explaining how AI vendors and pharma firms can capture value from digital services.
- Enterprise versioning: OpenAI can price different tiers for API access, security, model quality, support, and fine-tuning.
- Bundling: Novo may bundle AI tools into research workflows rather than treat them as separate software products.
- Value-based pricing: In pharma, the value of faster discovery can be enormous, so pricing may be linked to strategic value rather than marginal compute cost.
- Switching and lock-in: If AI tools become embedded in R&D workflows, the vendor may gain pricing power over time.
Exam Preparation
Likely Oral Exam Questions
-
Core How does Hui & Chau (2002) classify digital products, and why is "granularity" important? ▶
- Classification: 1) Tools/Utilities, 2) Content-based, 3) Online Services.
- Granularity: Refers to divisibility. Content (like a newspaper) is highly granular — you can sell a single article. Tools (like anti-virus) are low granularity — you need the whole thing to work.
- Strategic importance: High granularity allows for versioning and capturing more consumer surplus by tailoring "slices" of the product to different segments.
-
Apply How does versioning help digital firms capture more value? ▶
- Different users have different needs and willingness to pay.
- Versions make customers self-select: free/basic users reveal low willingness to pay, while pro/enterprise users pay for higher quality, features, or support.
- The firm captures more consumer surplus without needing perfect personalized pricing.
-
Synthesis When should a digital firm use penetration pricing rather than skimming? ▶
- Penetration Pricing: Usually superior. You need to hit "critical mass" quickly to trigger the network effect. If you price too high (skimming), you may never build the necessary user base to make the product valuable.
- Connect this to the "Winner-Take-All" dynamic: being first to scale is often more important than immediate profit.