Local search: Using site search in regional press websites - Part 3
This week I've been writing a series of posts looking at the site search facilities provided by some of the UK's leading regional newspapers.
The EDP24 site has a very small search box located in the very top right-hand side of the homepage. This is labelled with an image that says 'News Database', and is titled 'Story search'.
When the results page initially loads, it is hard to see any results, as there are three large elements on the page above them.
The first of these is a large promo for the Google News Search service. Clicking this sends the user through to a Google News search scoped to just content from the Norfolk Eastern Daily Press.
The second element is a link suggesting that the user "can also get help from our newspaper library". This does not actually perform a search, but leads to a promotional page for the 'Archant Norfolk newspaper library'
"Interested in old photographs of Norwich and Norfolk? Looking for an article from a past newspaper? Want to check how your company has been reported in the local press? Why not try the research team at Prospect House library! For a small fee experienced staff will undertake your research. If you prefer to do your own research you can visit the newspaper archives by appointment"
The third feature is a large advanced search selection box, that allows the user to enter a new search and refine their results according to date ranges.
It doesn't surprise me the the site tries to direct users to other search services, as their own seems pretty poor. For example, on the 4th January, a search for 'norwich' produced a lot of results. On the first page of results, however, 8 of the 10 displayed were all the same article, a match report of the previous day's F.A. Cup tie against Charlton.
Portsmouth News & Sports Mail
Yorkshire Evening Post
The results generated are Google results rendered within an appropriate template. The sites don't seem to offer any additional features beyond a standard Google search restricted to their own content.
Access to search on the Belfast Telegraph site is via a box at the top right-hand side of the page. The page displays 20 results by default, and the results have a couple of pieces of metadata associated with them. As well as the headline and standfirst, users can see the date of publication and the main 'section' in which the story was published.
Less helpfully, there is also a percentage score for relevancy. Users are very aware that they are expecting the top results to be the most relevant, and so this kind of measure is pretty meaningless.
I did spot one 'Gotcha!' with the interface. When you search for a term, you are invited to refine your search by adding to a search box that has been pre-populated with your previous search terms and a boolean style 'AND'.
This happens even if your original search query has generated a null result set. Searching for 'anything' AND 'something that produces null results' is just going to lead the user to another dead end.
Tomorrow I'll be publishing a couple of table which sum up the search features across the 20 sites I've been reviewing.