Ready to test out new techniques
When I first transitioned to this department in March 2010, I attended two workshops in Ann Arbor at Merit Network. One was about Information Architecture and another was about Site Search Analytics. I learned a lot in these two days and I'm anxious to test out some of the techniques I've learned!
Here are some things I wanted to share based on what I learned:
Information Architecture
IA is all about helping people find what they need. Using a combination of statistics and field research will help you come up with different categories of content. But before you figure out how you are going to present this content, you need to figure out what you are presenting. Ask questions about an organizations goal, why they want a Website and what they want people to find there. Talk to the people who answer questions (e.g., switchboard secretary or helpdesk operator) about the organization.
Card Sorting
Use index cards to represent the content pieces. Lay them out to see what kind of pattern forms and what content relates to other content. Categorize things and see if there is commonality between items. This will help you form a good outline to start with.
Many organizations don't choose to make connections between related content. So they set up with a site and as time goes by they don't know where to add things and things end up in maybe not the best place.
When you feel you have a decent navigation together, try this navigation stress test (opens new window). It makes you look at your content differently and helps you spot anything you could have missed during your initial analysis.
Site Search Analytics
This was my favorite portion of the workshop -- insight into human behavior. First we analyzed logs and tried to determine a user's session and what they were looking for on a specific site. Did things change from line to line in the context of a session? Was the person at the library or were they at home?
What people are searching for on a site is what they want. So you can get an idea of what people are looking for in your site, regardless of whether it's obvious in the navigation or not. You may need to roll your logs into spreadsheets in order to get value from it.
If you keep looking at your data you will come up with question after question. And this is good since it's questions about your own data. Specific search terms will occur at certain times of the year sometimes. You will be able to spot this pattern and be prepared in the future.
Here are a few techniques to try:
- Review top ten and top 20 search term results every week and chart it out over a period of weeks.
- Review percentage of search exits -- number of people who left the site after a search. Is it a specific term? Search this term and see what might be holding people up. Were they looking for something else perhaps?
- Review low search terms. It may not indicate a problem but instead shows that people are finding what they are looking for.
- Analyze queries that returned zero search results. Why are people looking on your site for this content and what can you do to improve their results? Failure gets a lot of attention.