As part of the work we’ve been doing for the Student Counselling Service, I’ve spent this week looking at some automated reports on user behaviour and advising what to do about broken links.

We have a paid subscription to Sitebeam, a third party site checker, and can produce automated reports for any site for anyone within the University, for free.

Site appraisal tools – Website Programme website

So we ran a report on the SCS site, to compare how it seems to be performing against the outcome they want from the web content.

Student Counselling Service – objectives workshop (blog)

Easy fixes

Some things, like the readability score, are useful nuggets of information in their own right, immediately offering up a fairly straightforward solution – copy edit the content, going back to some of the basics of writing for the web, as covered in our ‘Effective Digital Content’ training.

Effective Digital Content training – Website Programme website

Broken links – to fix or not to fix?

Other information is useful – but only in combination with other data. In the case of SCS, the Sitebeam report threw up broken links. Whereas the obvious solution here might be to fix the links, that could just be a patch. We needed to look at where these links are, and what they’re for.

Focusing on key pages

When we crossed over to look at data from Google Analytics, we saw that one or two of the broken links were on key pages, frequently viewed and linked to elsewhere in the University. Most of them, though, were on pages in the ‘long tail’ of page views – that is, pages viewed by very few people. The links themselves were to outside organisations, and could well change again.

Optimising your long neck

In the latter case, the solution might not be to fix the link, but to remove it, and thus remove a content overhead. Often the whole page in question can be removed.

This table gives a quick summary of the factors involved when deciding what to do with a broken link – it’s a sliding scale, with the extreme ends shown here.

Where factors score highly, it’s more worthwhile fixing the link. Fixing a link no-one really needs, on a page few people visit, may not be the best use of your time.