Barclaycard has probably one of the largest resource libraries for those new to credit. You can find a detailed explanation of how credit cards work, what is your credit score and how the purchase rate is different from other types of interest – but only if you look hard enough.
All the guides for beginners are spread across the site almost randomly. The global navigation is also full of surprises:
In other words, a gold mine for someone looking to practice their IA skills.
Overview:
Planning the research activities
Setting up a tree testing study
Research methods used:
First things first, we need a plan. After having spent on the website a little longer I've realised that the issues with the navigation are much deeper than some labels being duplicated or misplaced, and at least a brief content inventory would be needed.
Next, we would need to sort the content and group it in a more sensible way (hello card sorting, it's been a while); test the results (a perfect time to try out tree testing studies and remote usability testing), and finalise the new menu.
I started with replicating the global menu in Google Sheets and highlighting all duplicated items. The result was fascinating: I found 13 pairs of duplicates; although I realised later that one of them, 'Balance transfers', was not a duplicate but a decoy – if you click on the 'Balance transfers' link that sits under 'Existing customer', you'll end up on this page, however, if you go to Get Credit Confident > Credit Cards > Balance transfers, you'll end up here.
So I checked all the other menu items to make sure they lead to the same pages – just in case – and removed one of each 'twins' out of my content inventory document.
Next, I went through the Sitemap and added all the pages from there that didn't appear on the main menu. Later on, I noticed that some of the website pages are missing even from the sitemap, but at that point, it felt like a decent start.
I've ended up with 85 pages that needed to be sorted and grouped somehow.
Sorting 85 cards in an open format requires quite a commitment. Knowing that some of the items have very similar labels, I decided to do the first round of sorting with myself as the only participant: the goal was to find pages with similar labels and check their content to see if some of them could be merged or made redundant.