I left a bag on the GO Train the other day (for those of you not in the Toronto area, the GO Train is our commuter rail system). After trying to chase the train down the highway (I didn’t quite make it to the next station in time), I resigned myself to hoping that someone else picked it up and turned it in.
The customer service rep at the train station told me that the lost and found was located at the main train station in Toronto, but if you’ve ever been to Union Station you know that it’s big and kind of disorganized. So I decided to check the GO Transit website to find out exactly where the lost and found is.
First try: Can I find information about it just using the site navigation?
Nothing jumped out at me right away (although upon further review, if you click on Union Station -> About Union Station and open the station map, Lost and Found is indicated on there), so on to option two.
Second try: Site search.
I clicked on the search box…
Hang on, that’s not a search box. It’s where a search box would normally go, looks like a text box, and even has the Google logo. But it’s actually translation (it says “Select language” in the box, but as a regular ol’ web user I didn’t actually read that).
Then I discovered the “Search GOTransit.com” link above that, which took me to a separate page that would let me search the site:
The surprisingly large block of text did catch my attention (after all, how much can you possibly write about searching a site?). It reads:
Use the form below to search for documents in this web containing specific words or combinations of words. The text search engine will display a weighted list of matching documents, with better matches shown first. Each list item is a link to a matching document; if the document has a title it will be shown, otherwise only the document’s file name is displayed.
Type the word you are looking for and click the Start Search button.
I think most people using the GO Transit website probably understand how a search engine typically works. There’s no need to say “weighted list of matching documents”. And of course “better matches [are] shown first”. It would be stupid to design a search engine that displays stuff that doesn’t match first.
“Each list item is a link to a matching document.” Yup, that’s how search generally works.
“If the document has a title it will be shown, otherwise only the document’s file name is displayed.” That’s just poor content design. Give everything a title. Oh, and most people using the site search probably won’t care.
So I typed in “lost and found” and got these results:
Where to start? Problems abound:
Site Search Design
- If you’ve used a search engine in the last 10 years, you’re probably used to seeing a linked title, a couple of lines of text, and then a URL (or some slight variation of that). You’re not used to seeing a four column table. I’d suggest you go for consistency with widely-accepted standards.
- Because they’ve chosen columns, typically long pieces of text (like filenames and page titles) are wrapped into 12-15 character line lengths, making them very difficult to read. The columns don’t take up any less room than displaying 3 or 4 lines of text, so why bother?
- Filenames? First of all, they’re not that helpful (e.g., result #4: buspassengersave.aspx). Second of all, they’re already in the URL. They’re really just taking up space.
- Size? Unless they’re downloading huge documents on a dial-up modem, most people don’t care. Oh, and what does this number represent? Bytes? Kilobytes? Characters? Words? Monkeys it took to write the content?
Site Search Functionality
- Didn’t they say the “better matches [are] shown first”? The first result is “copy of default.aspx”. That’s the homepage. Actually, it’s a random copy of the homepage. And I did a Cmd-F for “lost and found” on the homepage just because I was curious. It’s not there.
- I checked the other 9 pages – no “lost and found” on any of those pages either. Generally I’d recommend picking a search engine that will return helpful results.
- There are two copies of the homepage and three copies of the “citylocations.aspx” page in the first ten results.
- Probably your site search engine shouldn’t index things in the “_private” or “to_be_removed_sept302007″ folders. Most of these are blocked from Google by the GOTransit.com robots.txt file (although not all of them – they should check that), so why is the internal search allowed to index them?
- Something like “lost and found” should probably return one page. It’s pretty clear what I’m looking for. Tell me what the lost and found policy is and where I can go to find the lost and found. Pick a site search engine that lets you create priority pages for specific search terms.
Third try: Google
I gave up and Googled “go transit lost and found”. Google’s got it figured out – they returned the page I was looking for in the #1 spot:
Perhaps GO Transit should consider using Google site search as their internal search engine? Luckily the guy that I talked to at the Lost and Found counter was friendly and helpful, because I was relatively annoyed with the website.
My bag wasn’t there, but apparently I was speedier than they usually are – I’ll keep my fingers crossed and check again next week.