This summer, we will roll out several updates and enhancements to the university website that will affect all web pages. These updates are in response to consistent and frequent website user feedback, website accessibility requirements, user behavior evaluations and data analysis. The size of text will be increased in content areas making it easier to read stories and browse information. Visit Website Updates for detailed summary of changes, examples and review checklist.
“Disneyland will never be completed. It will continue to grow as long as there is imagination left in the world.” Walt Disney’s quote resonates well with technology as it is constantly evolving. Information provided in websites are a perfect example of this continuum where content, design and technology are constantly evolving, creating an interesting challenge and opportunity to ensure user satisfaction.
Websites are constantly updated with new content. Existing content is often modified or removed faster than it can be re-indexed by search engines and faster than referring inbound links can be fixed. This fast moving, non-linear momentum results in users visiting pages that don’t exist anymore (404 page), hurting the user experience. How can you stay proactive and minimize 404 pages? Below, using Google Analytics, is our implementation to generate a real-time automated “404 pages & referral pages” report.
STEP 1: Insert Google Analytics code
Here “var ref” has the location of the referring page and “var errorpage” has the location of the page not found. Using custom variables to capture 404 page and referring page location based on user click provides accurate information (posting on Google Analytics forum). Test your changes. Below is a screen shot of the Google Analytics Debugger extension for Chrome. Check if “_setCustomVar” is getting initialized with appropriate information.
STEP 2: Create Google Analytics Custom Report
Create a custom report with Dimensions ‘Custom Variable (Value 02)’ and ‘Custom Variable (Value 01)’. New to creating custom report? Article from Google Analytics about creating custom report.
STEP 3: Pages not found & referral pages report
The report has location of pages not found, location of referring pages and #visits. The report can be filtered and sorted to meet your needs. You can also create notification to receive regular updates in different formats. Amazing, isn’t it? In three steps you can generate automated real-time report so you may start taking steps towards minimizing 404 pages leading to higher user satisfaction.
Background
Last week we saw a spike in search traffic for a specific query term. Further investigation lead to a missing document, users were looking for. Ideas from Annie Cushing article to get full referrer and James Green article about tracking 404 pages lead to prototypes and after few iterations let to the current implementation.
We are pleased to announce the integration of Twitter’s typeahead.js with SJSU Search. typeahead.js is a “fast and battle-tested jQuery plugin for auto completion“ and is open sourced on GitHub under the MIT license. As you begin to type your query, Twitter’s typeahead.js suggests related search terms to choose beneath the search box empowering users to narrow down the context of the search (user intent) to generate relevant search results that increase user satisfaction.
We are also taking advantage of this auto completion feature to improve the user experience by mapping services and information users are looking for with internal terminology and units providing the service. This mapping makes it easy and convenient for users to find what they are looking for. For example when you type “transfer” you see terms related to “articulation” and “course-to-course articulation” in addition to “transfer” related search terms.
What are your thoughts? How can we improve the SJSU Search experience?
Enhancements to SJSU Search was presented at the ITS Open Forum on 3/21. Slides from the presentation are available for download [PDF].
6 weeks ago we started making incremental enhancements to SJSU Search Engine towards increasing user satisfaction. Incremental changes were coordinated in 4 major areas.
@ThinkWithGoogle 8 pillars of innovation that include, look for ideas everywhere, spark with imagination, fuel with data and think big start small have been the guiding principles.
SJSU Search Analytics: 3/2/2013 – 3/17/2013
A satisfied query is a search that ends with a user clicking on a result.
*A truly satisfied query is a search that ends with user finding what they were looking for (or stop searching for the same query after clicking on a result)
+An abandoned query is a search than ends with a user not clicking on any result
^An unsuccessful query is a search that start with user clicking on a result but ends with a user not clicking on a result
In the weeks ahead we will analyze abandoned queries and queries taking more than one click to identify solutions. We are are going to be making numerous incremental changes to www.sjsu.edu including primary navigation. We are also going to reach out to content editors and content owners so we can work together towards increasing user satisfaction and providing a consistent positive user experience.