A substantial body of theoretical literature indicates that network effects may hinder the entry of higher-quality products into markets in which network effects are impor tant. However, Tellis, Yin, and Niraj (2009) provide com pelling evidence that, in general, higher-quality offerings win out in software markets after a short time lag. Because software markets are commonly believed to be susceptible to network effects, this finding provides important empiri cal evidence against the hypothesis that network effects impede entry. Because Tellis, Yin, and Niraj obtain their results across a large number of product categories and because their analysis holds up across various methods, their evidence that high quality trumps network effects is impressive. However, in the final section of the article, Tellis, Yin, and Niraj are careful to provide a set of limitations for their research. Because I believe that their results must be quali fied in the light of these limitations, I elaborate on some of these in my comment. Because the authors have gone about as far as possible with the data at their disposal, this com ment is intended to stimulate further research on the topic of network effects and quality. Consistent with Tellis, Yin, and Niraj's research objec tives, their conceptual model focuses on the demand side and factors that might affect consumer response, but the supply side is also important. In particular, it is not known whether the firms refrained from developing or marketing products because they judged that network effects were too difficult to overcome. Thus, the results are subject to a sam ple selection problem, in which only products that suppliers believed to be worthy of introduction on the market were selected. Because the sample is limited to cases in which suppliers believed that introducing the product on the mar ket was justified in the face of any network effects, this cre ates an unknown bias toward showing that quality can over come network effects. A related consideration is that suppliers have ways of dealing with network effects or even using them to their advantage. One is to make the higher-quality product com patible with its predecessor, such as making Excel compati ble with Lotus. Another is to arrange to have software bun dled with the sale of new computers, thus forcing its acceptance in the market. For example, a current buyer of a Windows computer must either accept Vista or have some one uninstall this software in favor of an older version, a time-consuming and expensive process. This bundling may have facilitated the adoption of Windows, Word, Excel, Internet Explorer, PowerPoint, AOL, and possibly other software types. Supplier actions to mitigate network effects, such as compatibility and bundling, do not invalidate the general findings in Tellis, Yin, and Niraj's article. Rather, they may help explain how and why network effects can be overcome.