Recent articles
Customer-centric information architecture for efficient Customer insight
Google setting the clock on the web
Evaluation of User-intent on the web
Recency analysis on the web channel
Customer-centric information architecture for efficient Customer insight
Google setting the clock on the web
Evaluation of User-intent on the web
Recency analysis on the web channel
Since I started blogging, I have been reviewing my host’s weblog based analytics info and monitoring trends in traffic, most visited pages, traffic sources (referrer URLs), keywords which produce traffic (the q parameter in search engine organic traffic) and other stuff. Moreover I learned to focus on useful information (keywords, sessions, pageviews etc) while ignoring the useless (hits, files, Kbytes sent).
I have recently followed the advice of Avinash Kaushik (a major thought leader on the subject of web analytics) and installed Google analytics (hereafter GA) on my web pages. Actually, it is a very easy process copying and pasting a javascript piece at the end of the html code of each page. GA is a free page-tagging analytics solution, an approach which is considered by many as superior to web log based ones.
GA offers some very interesting reports. Some basic reports involve : • Sessions & Pageviews • Visitors vs new visitors: you can assess the level of repeat visitors on your site. Achieving a high percentage of repeat visitors should be a goal (site community building and visitor engagement analysis is carried out by the web marketing professionals community) • Highly visited pages (top 10) • Bounces (visits that involved a single page view) and exit rates per page • Average length of stay on each page and distribution . Pages on which length of stay is high are the most successful pages. Pageviews that lasted for few seconds (<10 sec). I have noticed a correlation between length of visit and pageviews on a page (of course this is because my pages are informative: the richer the content the higher the search traffic attracted). A very high percentage of such views is a bad sign: the page is not attractive to its visitors. A goal should be to reduce the percentage of bounces. • Depth of visit: number of pageviews per visit. • Traffic Segments by source (referrer reports): direct, search engine, affiliate. Segmentation is a powerful technique
In a next post, I will produce more detailed info and screenshots on GA.
I would like to pay tribute to Avinash Kaushik for the invaluable guidance he gives on his blog Occam’s Razor. He is proposing to avoid buying an expensive web analytics tool, before having reached a certain level of analytics maturity, gained by the implementation of a free solution like GA.
The value potential of web analytics, even with the use of a free tool like GA, has not been identified by many webmasters and businesses.
[Update on this post]
Some interesting analytics and relevant GA screenshots are presented below.
Visitor loyalty : Presents in a histogram the percentage for each category: new visitor, 2 visits, 3 visits etc in the selected time period. Visitor loyalty is an indicator of the visitor engagement achieved.
The graph below shows a low level of loyalty. The goal should be to move the histogram peak to the right (more repeat visits).
A related metric is ‘Absolute unique visitors’. The major goal is to increase the percentage of returning visitors. Returning visitors are users who have found value in the content or services offered.
‘Content drilldown’ screen provides a comprehensive report on the performance of the individual pages. Exit rate is an indicator of the success the page had in attracting the visitor’s interest. In the case of a content site (like a blog), the ability to provide contextually relevant links to other internal pages can affect this rate. In general the goal on content sites is to have a low exit rate on pages. On ecommerce sites the exit rate on initial pages (in the path towards an order) is very important; it determines the abandonment rate.
Average time on each page is a strong indicator of the attention applied by the visitor. Certain content pages may receive heavy attention (more than 3 minutes) while pages browsed for less than 10 seconds can be considered as unsuccessful visits (or bounces).
Content summary is an alternative view of content performance.
Click density (or Site overlay) allows you to navigate your site while viewing traffic and conversion data for each link. The bars are graphic indicators of clicks (percentage of all page clicks for the time range selected). Provides an indication on ‘sweet spots’ on the layout.
Referring Source: Analyses traffic segmented per source (visits per source, pageviews/visit per source), conversion rate (vis-a-vis goal pages set) per source.
The difference between weblog based analytics tools and GA, is that the page tagging approach of the latter provides more accurate results focused on actual users than all traffic, part of which usually relates to visits by search engine or other software bots. Moreover page tagging allows the isolation of certain PCs and exclusion of their traffic (for example the webmaster’s development and monitoring traffic) from the analysis provided.
On the other hand, page tagging and cookie based analysis may introduce a certain bias, given that many users do not allow javascript on their PC or delete cookies often.