In his article, Wixon argues that everything can be measured, but something about the object of measurement is always lost. Therefore, we need multiple measurements to get more accurate measures of the real thing. Still, the most important thing in Wixon’s opinion is that the ones doing and using the measurements agree that the metrics are meaningful and sufficient to work with.
Wixon’s argument might seem a bit bold – or naive – at first, but his analog to measuring marathon route is quite good. The length of the marathon does not tell very much of the route, so the ones really running the marathon need more metrics, such as altitude and profile of the route. Still, many are quite satisfied just with the length of the route. As with the marathon route, the team developing a new system could agree with one or two major metrics, and then complement these with new metrics if needed. There are plenty of ways to study people’s reactions to systems – for example asking about it. About asking the users, Wixon reminds that it is important to ask the users in many ways about the likings and reactions – not just one question, but many things around it as well.