Measuring continuity

Whenever we speak of measuring web performance, we always refer to measuring static events, like page load or time to first tweet. A performant user experience is much more than that. Philip Tellis and Nic Jansma explore methods of measuring web performance as it relates to continuous interactions between the user and a page.

Whenever we speak of measuring web performance, we always refer to measuring static events, whether the old page-load model, the controversial above-the-fold time, or the product-specific “time to first tweet.” Each of these have their pros and cons ranging from ease of implementation to relevance to the actual user’s experience. But a performant user experience is much more than these static events, and no methods measure the experience the user has interacting with the page after that initial event. We seldom measure the experience of the user when they spend the most time on the site (i.e., between load and unload, when they’re reading and interacting with content on the page, etc.) or ask questions like: Was scrolling smooth or janky? What about animations? Did the user click expecting something to happen but nothing did? Are too many reflows happening? Drawing on their work on boomerang, Philip Tellis and Nic Jansma explore new methods of measuring web performance as it relates to continuous interactions between the user and a page.