Through Google Analytics, Could We Establish Best Practices for Arranging Website Content?
- Is there a systematic way to pull reports for each major section of the website and compare interaction?
Pull and compare reports for pages with different kinds of content (text-only pages vs text-plus [video/social media links/etc.]).
The Web Team regularly conducts site-wide reviews of analytics, watches for anomalies in the data, runs usability and other tests, and references the work of experts in the field to create a body of best practices for arranging website content. We try to distill the essence of what we know into easy-to-follow guidelines which are posted at Standards & Styles for UUA.org (for UUA-specific stuff) and Creating a Congregational Website (which includes information helpful to congregations and others).
Some questions have answers that are more qualitative than quantitative. From usability studies and experts in the field, we know that people prefer web pages to PDFs or other “specialty” formats (and we know that the web-based version of “The Real Rules” saw 203 unique pageviews during July 2012, while downloads of the PDF version happened 3 times). If a critical piece of information is only offered as a PDF file, are there people who dislike PDFs enough (or are missing the necessary software) to avoid clicking on that link? The answer probably depends on how motivated the user is: the Settlement Handbook is probably safe. A pamphlet for seekers probably shouldn’t take the risk.
It would be of limited value to compare very different parts of the website, likely to have different audiences and perhaps even different times of activity (for example, board members might be interested in polity during the Fall, and growth committees might be interested in demographics during the Spring). Without knowing all the variables in play, comparisons can be misleading.
To determine whether one type of page design performs better than another, an “A/B” test could be devised where two different versions of a page are created. Each visitor to a particular URL would see either one version or the other, but never both (web cookies ensure they’ll keep seeing the same version on subsequent visits to the site). With a clear idea of what constitutes “success” (length of time on page or site, clicks to a particular link, etc.) it may be possible to determine a winning design.
Often, the answer to a particular question can be answered by reviewing similar studies that have already been done.