Versioning and Quality Distortion in Software? Evidence fromE-CommercePanel Data

We present a framework for measuring software quality using pricing anddemand data, and empirical estimates that quantify the extent of qualitydegradation associated with software ver- sioning. Using a 7-month,108-product panel of software sales from Amazon.com, we document theextent to which quality varies across different software versions,estimating quality degradation that ranges from as little as 8% to asmuch as 56% below that of the corresponding flagship ver- sion.Consistent with prescriptions from the theory of verticaldi¤erentiation, we also find that an increase in the total numberof versions is associated with an increase in the difference in qualitybetween the highest and lowest quality versions, and a decrease in thequality difference between 'neighboring' versions. We compare ourestimates with those derived from two sets of subjective measures ofquality, based on CNET editorial ratings and Amazon.com user reviews,and discuss competing interpretations of the significant differencesthat emerge from this comparison. As the first empirical study ofsoftware versioning that is based on both subjective and econometricallyestimated measures of quality, this paper provides a framework fortesting a wide variety of results in IS that are based on related modelsof vertical differentiation, and its findings have importantimplications for studies that treat web-based user ratings as cardinal data.