Duy D. Dao

Ph.D. Candidate in Information Systems, Rady School of Management

I am a fourth year Ph.D. candidate in Information Systems on the job market during the 2016-2017 academic year. My research broadly spans the economics of software and digital goods. In my job market paper, I examine how software vendors can use price as a lever to improve cybersecurity in a profitable way. In a second stream of research, I study pricing and release strategies in the entertainment industry.

Curriculum Vitae


Job Market Paper

Market Segmentation and Software Security: Pricing Patching Rights

(invited for Major Revision at Management Science, with Terrence August and Kihoon Kim) (pdf)

In my job market paper, I examine a new way to think about patching software that, if viewed in this way, could reduce cybersecurity risks for consumers, would be easily implementable in practice, and would be profitable for software vendors.

The reason I study this problem is because, over the years, large attacks such as Code Red, SQL Slammer, and Blaster have all exploited vulnerabilities for which patches were available prior to the attack. Even today, about two-thirds of exploits are still hitting vulnerabilities for which fixes were available years ago, according to HP Security Research. It’s clear that many users and organizations do not have the incentives needed to keep their machines up-to-date. For end users, patching may simply be an inconvenience. For firms, there is a barrage of patches to manage, and companies typically test patches extensively to make sure that the patches are compatible with their systems and applications.

The problem is that users' decisions to remain unpatched negatively impact other users of that software. Many users of a software being unpatched makes that software attractive to exploit. Furthermore, attacks that propagate across networks spread more rapidly when there are more unpatched users. As a result, unpatched users create risks for their networks such that the current patching approach has been less effective than desired.

Under the current patching approach, users and organizations are endowed with the right to decide whether or not to deploy security patches. A user’s ability to choose whether to patch is an implicit right upon buying the software. In this paper, we examine whether this right should be given freely to consumers. Together with Terrence August and Kihoon Kim, I study how this patching right should be priced appropriately. In the world we examine, users who value control over their patching process must pay for this. Viewed differently, users who are willing to forgo this patching right would actually receive a discount on the software. Paying this discounted price, these users would instead receive a version of the software that automatically updates whenever new security updates are released. Conceptually, unpatched consumers ultimately are the reason for the high number of attacks on patchable vulnerabilities. If patching rights are priced rather than endowed, then these users are forced to think about whether it is worth paying for this right so that they can remain unpatched.


Optimal Timing of Sequential Distribution: The Impact of Congestion Externalities and Day-and-Date Strategies

(with Terrence August and Hyoduk Shin), 2015, Marketing Science, 34(5), 755-774. (pdf)

My job market paper explores how negative externalities can be managed in the software industry in a profitable way. In my other stream of research, I study pricing, channeling, and information provision decisions in the context of releasing a product to congestion-sensitive consumers.

The motivation for this paper is that over the past fifteen years, we have seen dramatic changes to how movie studios manage film distribution. In recent years, more films have been skipping theatrical release entirely and going directly to home video as part of a direct-to-video strategy. In some cases, studios are even experimenting with “day-and-date” strategies (more noticeably in Europe than in the U.S.), distributing a new release across two or more distinct channels on the same day or within a few days of each other.

Historically, the industry has used a windowing strategy for theatrical releases, in which a movie is released on a primary channel for some time before being released on other channels. This strategy gives the studio and each distribution partner some time window to profit from the movie. By staggering the availability to each market, studios have hoped to tap into the full potential of each channel, without simultaneous competition from other distribution platforms. But is it profit-maximizing for a release window between sequential distribution channels to always exist?

Some argue that maintaining release windows reduces the impact of promotional spending, since in those months waiting for the video option, consumers might forget about the movie or, equivalently, have new options competing for their attention. On the other hand, cannibalization is a main concern when considering a smaller release window. In this paper, we view the problem from the whole supply-chain perspective and focus on the central question of how studios should time the release of a video option and price that option, given the key factors that drive different release strategies such as inherent movie quality, theater congestion costs, and movie durability driving multiple consumer purchases.

Working with Terrence August and Hyoduk Shin, I present a model of film distribution and consumption to gain insight into how studios should optimally price and time the release of video versions of their films given that consumers are making strategic decisions about how to consume the product. Our research objective is to develop an understanding of how a studio should coordinate the video release time and price for its film as part of a comprehensive strategy to manage the film’s total profitability across the theater and home video channels. Using our model, we characterize conditions under which direct-to-video, day-and-date, and delayed release strategies maximize profitability for a studio. Studios have begun experimenting with day-and-date strategies and benefit from understanding when these strategies would be most profitable.


Consumption Timing in Congestible Services: Strategic Smoothing

(with Terrence August and Hyoduk Shin)

Following this work, I characterize how consumers delay purchasing due to congestion costs during the release of a product, and together with Terrence August and Hyoduk Shin, I explore how studios and theaters can more effectively manage this externality.

A problem often faced by the entertainment industry is the impact of congestion on the release of a new product or service, such as a movie or an online game. For theatrical releases, consumers have learned to delay consumption to account for heightened congestion when a new movie is released but are forced to trade off this congestion cost with the loss of movie buzz as the movie fades in relevance over the release window. Some may even forgo purchase altogether because the movie has lost its relevance by the time they would have considered watching the movie. Furthermore, for online games, a surge of consumers logging on to play a massive multi-player online game often results in server issues when a game is initially released. This can induce consumers to post negative reviews for the game, and consumers may decide to not purchase games from that studio in the future.

In this paper, we model how consumers decide their timing of purchase when considering the congestion they experience, which arises as a result of others’ choices. One of the predictions our model offers is that the expected valuation of a consumer who purchases is non-monotonic in time. More specifically, higher valuation, low congestion-sensitive consumers will be first to purchase a new release. Subsequently, the expected valuation of a consumer will decrease in time initially (for example, over the first two or three weeks after a release), but after some time, the equilibrium congestion level will have reduced enough such that high valuation, high congestion-sensitive consumers find it incentive-compatible to purchase. As a result, there will be a jump in the expected valuation of a consumer, as those high valuation consumers who were congestion-sensitive delay purchasing in equilibrium. This has implications for a studio's decision of when to release the product on another channel. Understanding how consumers are spreading in time, we then explore the question of how a movie studio should time the announcement of a home video option. How long after a release should a studio wait before letting consumers know when the movie will be available on other channels? Making this announcement changes the information set of the consumers and, consequently, the resulting equilibrium consumption times. When to announce the time of product's availability on a secondary channel is a non-trivial decision, and we study the studio's decision problem in this paper.


I have had the opportunity to TA for the Rady School of Management's first-ever undergraduate business analytics course, as well as an MBA course on Technology and Innovation Strategy. In MGT 451 Technology and Innovation Strategy, I guided MBA students in developing their course projects on disruptive technologies in an industry of their choice. In MGT 153 Business Analytics, I led weekly discussion sessions to teach students the fundamentals of probability and statistics. By introducing students to a Shiny-based user-interface in R, I helped students learn how to explore data sets, conduct visualizations and regressions, and create decision trees. My teaching evaluations for these classes are given below.

Evaluations for MGT 153 Business Analytics, Fall 2016

Evaluations for MGT 153 Business Analytics, Spring 2016

Evaluations for MGT 451 Technology and Innovation Strategy, Winter 2016


Rady School of Management
University of California, San Diego
9500 Gilman Drive, MC 0553
La Jolla, CA 92093-0553