If you want to measure something happening on the web, you’re necessarily wrangling a moving target. New social networks might appear; new metrics might come into vogue, and users will want to see those reflected in their system. The best one can do when faced with dynamic technology is to build a system that allows for new modules to be plugged in and out. Future modules, however, will still have to be designed and coded depending on the latest input.
Another challenge is that many interesting questions are technologically unknowable with our current state of affairs. For instance, one popular request was to know why an article was doing well on Facebook. Due to the nature of that network, however, which is much less open than, say, Twitter, we simply can’t know the cascade of shares that lead to virality.vi Understanding the limits of analytics is an important starting point when framing the questions we want to ask about one’s content.