How do you find something unless you know what it’s called?
When researching the CRB article I used the government site’s search engine to answer what I thought was one simple question:
how much does it cost?
I typed those exact words into the search box and this was the result:
Not only was I intrigued to find I had stumbled upon a sole result, I was pleased that it was 447% likely to be the answer I wanted, despite the article in question containing a list of hundreds of questions posed by members of the public on such fascinating topics as how much the government Christmas party cost and whether the CRB uses bottled water at its meetings.
I did eventually find the answer on that page by using Firefox’s search-in-page feature and stepping through answer after answer until I found the one. Afterwards I also found a shorter, much better document specifically on the associated costs of a CRB check by typing in a single search term:
And therein lies the problem with search: everyone approaches it in different ways. The people that programmed the software clearly didn’t bank on one result coming back with a huge number of articles in it that yielded fanciful percentages. Likewise, they didn’t bank on someone asking a question in the search box. And they assumed I already knew their internal processes and terminology. That isn’t much use: if members of the public need to use the correct government terminology to find stuff, it begs the question what’s the point of the search feature at all.
Until cognitive and semantic search becomes a more widespread science (and the standards/browsers catch up) it seems we’re doomed to find little relevant content on the Internet’s cubby holes unless we climb inside the mind of the creator of the meta data.