CS 171 Spring 2009 Final Project:
Pricing Microsoft Office

by Steven Zhu

Source code: fp Table

Motivating Questions

- How should Microsoft price its Office products to maximize revenue?
- Should Microsoft bundle its Office products or sell them separately?
- How do network effects change the prices that maximize revenue?

In my economics tutorial, Economics of Information Technology, we learned about the effects of bundling and network externalities on information technology goods. We were asked to survey twenty people about their willingness to pay for Microsoft Office and calculate the revenue-maximizing price for several scenarios. While the calculations gave clear results, someone who wanted to see how changing prices affected revenues would have to pore over multiple tables. Therefore, I believed these questions would be much better answered through interactive visualization.

Data

I created an online survey that asked 1) How much you would pay for Microsoft Word, Excel, and PowerPoint if no one else used Office; and 2) How much you would pay for those programs if Office already had an established user base. The responses to the first set of questions would determine revenue-maximizing prices without network effects, and the responses to the second set of questions would determine revenue-maximizing prices with network effects.

Design

At the top left, there are two tabs that allow the user to view demand curves either for separate products or for pure bundling. Under the "separate products" view, the bars of each graph are in the colors of their respective products (blue for Word, green for Excel, red for PowerPoint). A gray line indicates the chosen price point for each product. The total revenue is indicated by the shaded gray bars. Below each graph, a box shows the price of the product, the quantity demanded at that price, and the revenue that would be generated at that price and quantity. There are two buttons at the top right. The user can switch between the dataset that accounts for network effects and the dataset that does not. The user can also immediately set all of the prices at the revenue-maximizing level. The chart on the right side displays the total revenue generated at the chosen price points, as well as the total consumer surplus and deadweight loss.

Use

The user can switch between bundling and no bundling by clicking on the tabs at the top left. The user can toggle on/off network effects by clicking the "Network Effects" button. Clicking within a graph will set the price at that point. The user may also drag each price line. When the mouse is hovering over a graph, the user can press the up and down keys to adjust the price by $1 increments. To immediately reach the optimal prices, the user can click the "Maximize Revenue" button.

Answers/Insights

I assume that Microsoft is a monopoly, so it can set prices wherever it likes. For clarification, I also assume that marginal costs are 0 because the products are software, so revenue equals producer surplus.

To maximize revenue, Microsoft should sell its products separately. With network effects, Microsoft should sell each Office component at $50. Wihout network effects, Microsoft should sell Word and Excel at $45 each, and PowerPoint at $30. Thus, network effects incresae the optimal prices.

For separate products, maximum total revenue with network effects is $1850, and maximum total revenue without network effects is $1230. For pure bundling, maximum total revenue with network effects is $1650, and maximum total revenue without network effects is $1080. Thus, pure bundling does not increase Microsoft's revenue, but network effects do.

The visualization also revealed that network effects increase deadweight loss.