On November 22, I added some code to our homepage that tracks when someone clicks on one of the waveform bars to open a story and when they click on the links within that story to read more about it. This doesn’t count the story that is initially open when you load the page. This tracking is done using the Google Analytics* Event Tracking code, which we can use to measure user interactions with the site. If you have suggestions on other activities that we can measure, please let me know.
Since November 22, 67,888 story bars have been opened on the home page, 20,973 of which have been “unique events”. Unique Events is the number of visits during which one or more events occurred. This means that an average visitor to the site who opens at least one story will open 3.24 stories during their home page visit. The people who use the waveform, like to click around and view multiple stories.
Around 1,000 stories are opened every day on the site. The day with the most opened stories was January 9 when 1,457 stories were opened. The day with the fewest opened stories was December 25 when only 461 stories were opened (Thanksgiving was a close second here with 464 stories opened). The most opened stories was A Green Tour of Campus which was opened 1,641 times since November 22. The least opened story (7 times) was Field Hockey Falls in NCAA Title Game, though that story was published shortly before we began tracking and may not have been featured for very long.
The average story was opened 411 times. In all, 161 different stories were opened on the home page. This is significant because, we generally have around 100 stories on the homepage at any given time; currently there are 90. This means that every story in the waveform is getting attention from some number of people visiting the site. There are usually a half dozen “featured” stories that are in the rotation of stories which will automatically open when you visit the site. These are always in the center of the waveform. The position of the remaining stories is random, changing once every five minutes or so when the cache for the homepage is regenerated.
One thing to note is that the waveform serves the primary audience of the homepage: external visitors. Of the 67,888 stories opened since November 22, only 12,491 were opened by people browsing on the Middlebury campus. Local users are better served by the new portal, which contains these stories alongside other information we hope is useful to the community. Also since November 22, the average time on page for an external visitor** was 3 minutes and 7 seconds, showing that visitors do stay and play with the waveform feature.
Of the 20,973 visits where someone opened at least one story, 8,344 resulted in the person clicking to read one of the stories they opened. The most read story was an AP article about a Middlebury alum winning Survivor: South Pacific. Other popular stories were A Green Tour of Campus, One Thing to Remember for Finals Week, and a Thank You to Donors for the Holidays.
These are my conclusions from this data:
- People who visit the set “get” the waveform and click to find multiple stories.
- The waveform is most interesting to external visitors and we should continue efforts to promote the new portal for the local community’s needs.
- We should look at what stories got the most clicks and which got the fewest to figure out what type of content is the most compelling for our external audience and produce more of that content.
* See the Middlebury Privacy Policy for our data collection policy. The Middlebury website does respect the Do Not Track header, if you have that turned on in your browser.
** The average time on page for an local visitor was 5 minutes and 43 seconds, but these stats get skewed by lab and walk-up machines where the homepage is their default page in the browser and are left on all day.
Hi Ian– this is all very interesting! I’m curious, what’s the percentage of people who visit the homepage that end up clicking on at least one waveform bar?
3.67% of unique pageviews result in someone clicking at least one story bar.
Ian, great stuff here. Conclusion number three is of particular interest in that it suggests a potential stumbling block. Creating more of a certain type of content requires either dedicating more human resources to content creation or realizing a simultaneous decrease in other types of content. The latter usually requires significant discipline to achieve.