How easy is your catalogue to search?

One of the eternal questions, and worries, for Libraries is the user-friendliness of their search mechanism. For most of us that means the OPAC provided by our library system vendor and traditionally Libraries have been prone to simple 'out-of-the-box' implementations or (perhaps even worse...) turning 'on' all of the catalogue's added-value features.

Recently, however, there have been more attempts to provide a coherent - and contextual - approach to the OPAC. Let's face facts: No one reads help pages or FAQs. It's true - check your own web logs. So how do you help your users to get the most out of their search (and most importantly, not just walk away)? Here's some ideas.

Little and Often 

Ok - so the big chunks of help don't work but there's nothing wrong with a nudge in the right direction. Check out any popular website and you'll see succinct clues to how and why results appear as they do. The best trick I've found is to try and write a short sentence to explain a function and then cut it down to half the time. 

Reuse! Reuse! Reuse!

Take a look at the University of Huddersfield Catalogue (the work of Dave Pattern) - data's been pulled from search and usage logs to create neat little features like the Tag Cloud on the front page and the Amazon-style 'other people also searched with' feature on the results page.

Don't Fight Google

It's done. People expect your search box to work like Google. In fact, not just like Google but better than Google. We can moan about the 'dumbing down' of researchers but as soon as people see that empty white box they expect to stick a string of (misspelled) keywords into it and get a result they like within the first page of hits. The sooner we learn to work within these parameters, and not fight them, the sooner we can build better mechanisms for search. And don't even think about making the default search anything except keyword.

Know Your Data (& Fix Your Indexing)

In any kind of searching consistency is everything so we all need a firm grip on our data. A beautifully constructed catalogue record is one thing but if your search parameters and indexing are so complicated that complex combinations are required to achieve useful search results perhaps it's time to reflect on the value of that data. A few coherent indexes with consistent data will always beat a hundred 'correct' ones.

Accept the Complexity 

Sometimes, however the gap between user expectations and the data is just too vast. Say, for example, that you're predominantly a science library but you've also got a small audio collection - how do you provide a straightforward service to your 90% of science users but also support the 10% of audiophiles? Sometimes it comes down to accepting that you're going to hit issues and making innovative use of zero results pages, 404s, "Email a Librarian" functions and hey - maybe even a live 'search help' facility.

A fair number of libraries are starting to look faceted and other 'innovative search solutions' (Primo, Aquabrowser, VUFind, and so on) to try and give a kick start to their user experience. These products are great but we still have to remember to be clear in our approach to tailoring that user experience.

The final approach is to embrace the major search engines and to focus on exposing your catalogue and Search Engine Optimisation. I have to admit to having some sympathy for this approach - rather than trying to constantly improve your own search it should be possible to just have a first class landing page and leave the rest of the work to Google, Yahoo! and the next great search engine through a site-search. Of course, this all depends on how open your catalogue is to indexing - or how up you are on XML site maps.