Revisiting Web Usage Metrics
Posted by Brian Kelly on 17 March 2008
I recently wrote a post on The UK Government and Web Metrics in which I described potential ambiguities in reporting on the usage of Government Web sites. In a comment on the post Phil Wilson oberved that
This extract from Hansard only really tells me one thing: there isn’t a government-wide standardised hit-tracking/visitor analysis scheme.
That’s true – and the temptation would be to recommend the adoption of an industry standard, such as that provided by ABCE. As this page says:
The ABC international standards working party (IFABC, International Federation of Audit Bureaux, http://www.ifabc.org/) has developed a set of rules and definitions that are the effective world-wide standard for Web audits. Definitions and rules specific to the internet industry in the UK and Ireland are controlled and developed by JICWEBS, the Joint Industry Committee for Web Standards. ALL current Industry agreed metrics are listed below (in alphabetical order):
Great, we have a standard which can be used for measuring Web usage.
The problem is, what if the content of a Web site is syndicated? What if users don’t visit the Web site to read the information, but expect the information to come to them, via their preferred RSS reader?
This struck me when I viewed the usage statistics for my initial post on The UK Government and Web Metrics. At one stage all I could view via the administrators interface on the WordPress.com service was the overall hits on pages on my blog. But some time ago WordPress provided a display of syndicated accesses to blog posts, as can be seen in the image.
Now what would I report on the day the post was published if I was making use of the ABCE’s standard for Web site usage? Less than 40 page views on the day the post was published, and a drop in views after that. The statistics showing the much higher syndicated views of the post would fail to be reported.
OK, so the usage data is flawed – but everyone knows that. The danger, of course, if usage data becomes competitive, with services failing to be funded if the usage levels as recorded by Web site visits doesn’t reach acceptable levels. And what will providing RSS feeds to services do – it may provide a richer and more personalised ervice for the end user, but the Web usage figures as reported by tools which comply with the ABCE standard will drop.
Here’s an example of how use of an agreed international can potentially result in a failure to develop richer service for the user community. Now I’m not saying that we shouldn’t have an agreed baseline for usage statistics. Rather the Web site usage needs to be analysed in conjunction with an understanding of alternative ways in which users may access the data. And I don’t know if there’s a standard available for this.