Metricas de la experiencia de usuario a gran escala

Solo disponible en BuenasTareas
  • Páginas : 15 (3508 palabras )
  • Descarga(s) : 0
  • Publicado : 11 de marzo de 2011
Leer documento completo
Vista previa del texto
Measuring the User Experience on a Large Scale: User-Centered Metrics for Web Applications
Kerry Rodden, Hilary Hutchinson, and Xin Fu Google 1600 Amphitheatre Parkway, Mountain View, CA 94043, USA {krodden, hhutchinson, xfu}

More and more products and services are being deployed on the web, and this presents new challenges and opportunities for measurement of userexperience on a large scale. There is a strong need for user-centered metrics for web applications, which can be used to measure progress towards key goals, and drive product decisions. In this note, we describe the HEART framework for user-centered metrics, as well as a process for mapping product goals to metrics. We include practical examples of how HEART metrics have helped product teams makedecisions that are both data-driven and user-centered. The framework and process have generalized to enough of our company’s own products that we are confident that teams in other organizations will be able to reuse or adapt them. We also hope to encourage more research into metrics based on large-scale behavioral data.
Author Keywords

ability to use web server log data to track product usage on alarge scale. With additional instrumentation, it is also possible to run controlled experiments (A/B tests) that compare interface alternatives. But on what criteria should they be compared, from a user-centered perspective? How should we scale up the familiar metrics of user experience, and what new opportunities exist? In the CHI community, there is already an established practice of measuringattitudinal data (such as satisfaction) on both a small scale (in the lab) and a large scale (via surveys). However, in terms of behavioral data, the established measurements are mostly small-scale, and gathered with stopwatches and checklists as part of lab experiments, e.g. effectiveness (task completion rate, error rate) and efficiency (time-on-task) [13]. A key missing piece in CHI research isuser experience metrics based on large-scale behavioral data. The web analytics community has been working to shift the focus from simple page hit counts to key performance indicators. However, the typical motivations in that community are still largely business-centered rather than user-centered. Web analytics packages provide off-the-shelf metrics solutions that may be too generic to address userexperience questions, or too specific to the e-commerce context to be useful for the wide range of applications and interactions that are possible on the web. We have created a framework and process for defining large-scale user-centered metrics, both attitudinal and behavioral. We generalized these from our experiences of working at a large company whose products cover a wide range of categories(both consumer-oriented and businessoriented), are almost all web-based, and have millions of users each. We have found that the framework and process have been applicable to, and useful for, enough of our company’s own products that we are confident that teams in other organizations will be able to reuse or adapt them successfully. We also hope to encourage more research into metrics based onlarge-scale behavioral data, in particular.

Metrics, web analytics, web applications, log analysis.
ACM Classification Keywords

H.5.2 [Information interfaces and presentation]: User Interfaces—benchmarking, evaluation/methodology.
General Terms

Experimentation, Human Factors, Measurement.

Advances in web technology have enabled more applications and servicesto become web-based and increasingly interactive. It is now possible for users to do a wide range of common tasks “in the cloud”, including those that were previously restricted to native client applications (e.g. word processing, editing photos). For user experience professionals, one of the key implications of this shift is the
© ACM, 2010. This is the author’s version of the work. It is...
tracking img