A-Z Indexes to Enhance Site Searching : Comments
January 5, 2005
Comments
Despite having built websites for three years it has never crossed my mind to build an A-Z for the site. However, if I look at some of the site maps I’ve done then it’s easy to see they are half way there anyway. A good article for me as it has focused my mind a little bit on an area that I’d always ‘just done’ rather than thought hard about.
Heather, I can’t believe that you have written an article about A to Z s and haven’t mentioned UsabilityViews, or the fantastic Alertbox A to Z.
Or Digital Web A to Z hehe… point being, there are plenty of examples, some better than others.
Nick, I would have expected to find Jakob under J, rather than D for Doctor.
At work we have put a lot of effort into producing a useful index on our intranet that doesn’t have 500 entries under T because of the dreaded ‘The’ by being very strict about the names of sites that appear in our index which contains over 1,500 entries.
Our A to Z is displayed on the home page with each letter appearing on the homepage in a 9×3 grid. We have found that 12% of navigation is via “A to Z”, 26% via the search results with the remaining 61% of homepage navigation being made via the specific homepage navigation elements.
We have found that the A to Z is much more effective if you display a one ot two line summary with the link name.
I think it’s essential to display a 1 or 2 line summary for the system that Chris mentions (I work for the same company). The reason being, there are so many pages on the Intranet and so few people who maintain the pages that are careful/dilligent with the naming of the page (as in what appears in the TITLE tag). Personally, I’m always thinking of what people are likely to search for and what the page is gonna give them based on that title.
But to go back to the UsabilityViews page mentioned (an excellent site, by the way), I would suggest that there could be a little more work done to refine this. Namely, there are a lot of pages that begin ‘A review of’ etc. It would be a doddle to convert such titles so that they appear alphabetically minus the leading ‘A’ (so ‘A review of …’ appears in the R listing as ‘[A] review of …’ or similar). Well, Chris, you did mention the dreaded ‘The ‘ problem, so I had to throw it back at you ;-)
Heather – brilliant. Absolutely brilliant. After all the time spent on creating a site map, category listings, pushing old articles back up to the front for a fresh review, why did I, a long time writer, think of this. Thank you for bringing this to our attention. Now, if we can just get brilliant software products to handle this more efficiently….
This is a good article and I can see how we can use this for our web survey tool(especially in the User Directory section). I think another way this could be used efficiently is to have the index indicate more than just the alphabets. For example:
For denoting the number of entries for each alphabet,
A(3) | B(2) | C(1) | D(9)....
usage Justification: This could be useful when we are trying to get a quick count of the total number of users without actually intervening in each aplhabet entry list.
or
A New! | B | CNew! ....
Usage Justification: This is more of notification about new entry to the corresponding alphabets.
A long-time Technical Writer, this was a no-brainer when tasked with designing, creating, and managing a repository of knowledge assets within the division. I’m pleased to read others believe the method — however traditional it might be — does in fact add value.
While I agree that an A-Z index is not a bad thing to have, I fail to see how it is much better than a good search function. Would all synonyms be included in the index? What about different word orders (“french revolution” vs. “revolutions in europe”)? Boolean search fixes both of these features – something like “france|french revolution|uprising” where “|” is the “or” operator might work well to find the information I want.
Ultimately, an index like this is too one-dimensional for the amount of information we have on today’s intra- and internets. Also, indexes have to be updated when new content is added – a search is based on information that is already there. There is a reason that the main way of finding information in most online resources is the search. It just works.
Very good point tom. The amount of work involved maintaining the index for a very large site would almost certainly be too much.
You can use this
feed to keep up with the comments made on this article.