Every Figure Tells a Story

I originally set the contest to run for only a week, but the students were getting into it so much that I extended it another seven days. As you'll see from the screenshots of the Google Analytics below, the contest wasn't really a contest after the first day or so. However, the students were fascinated (and a bit creeped out) by the amount of data I could pull from their marketing efforts. To whit: I was able to pinpoint that after one student sent out an email blast to her friends in China, I was able to track the individual page visits from her friends using iPhones on a wi-fi network from a bookstore in Jiangsu Province via the Unicom network.

After one day, this is what we were seeing:

Obviously, one student has shot out ahead of the pack in terms of visits to her page. But check out some of the other numbers that are showing up. For instance, check out the bounce rate - not so bad, when you consider that a lot of the people coming to these pages do not speak English as their first language.


The time spent on page - OK, OK, I know that this is a tricky and unreliable stat, but still. By no means are the stats telling us that people were performing "pro forma" clicks to these pages, just to satisfy a call to win a contest.


No, they were actually sticking around and reading the content. Which speaks to just how "sticky" eyecandy infographics are, and how they can engage readers long enough to get them to start scrolling around the page to see what else is going on there.

By Day 2, the traffic was really starting to pour in (especially to a student project):

In the space of one day, we've seen the pageviews basically triple. The line on the chart looks like the famed "hockey stick" from an Al Gore presentation. We're also seeing that there is a big disparity between pageviews and unique pageviews. Which tells me that some of the friends of my students are trying to "game the system" by loading up the page and clicking "Refresh" over and over.


Google is too smart for that, kiddies. Unless you can build a bot that opens a new Incognito window where the tracking cookies from the previous browser sessions aren't already resident and in effect, that there Unique number ain't gonna budge.


The Bounce Rate has gone up a bit - but so has the Time on Page.

By Day 3, other lessons are starting to emerge - such as the importance of mobile:

This one really stunned my students. They are so used to thinking in terms of appealing to readers who are accessing sites via the regular old "wired desktop," that they never really considered how their friends and contacts were going to be checking out their creations. This, despite the fact that they all have their mobile devices permanently welded to the palms of their hands.


These numbers are no doubt skewed by the framework of the experiment. My students are reaching out to people in their same age range. Who overindex for mobile usage. Much of the traffic to the site is being driven by two students from China - where, again, the user base overindexes for mobile usage.


I have to admit that I was surprised by how low the number is for tablet usage. I had thought that tablet market penetration that we hear so much about would have drive those numbers much, much higher. I guess starving students and the parents who support them really don't have much extra in the old bank account to plunk down on an iPad.

One week in: we start to see the global audience, and the concentration in China

This composite panel shows that we are getting visits from all over the world -- showing what a global network my students have. The next stat that jumps out is the Visits by Browser. Not only is this heavily weighted towards mobile - but specifically to Safari. Which tells me that the users are on their iPhones. Almost 100% more than visits from the Android system.


There is a doctoral thesis waiting to be done here, about the spread of iPhones amongst international students who are studying journalism or mass communications, particularly in China.


It's also interesting to see that we're getting traffic from Ukraine (probably from my students there, who have seen this contest on my Facebook feed), and from New Zealand, Finland and Saudi Arabia, which I don't even.


The intensity of the traffic is settling into the familiar bell-shaped curve. The bounce rate is staying about the same.

Almost done: email trumps social

This may be the most important lesson I imparted to the students. Despite all the hype about how to use social media to drive traffic and create "viral hits" the main source of traffic is a direct link. That is, something that was sent out via email blast.


Email is often overlooked by journalists, because it seems to old school. But that old dog has a real bite. About 500% more effective that using Twitter or Facebook to plead for clicks. I tried to get my students to use bit.ly or other link shorteners that would allow them to track what they were doing on their own. Unfortunately, it seems that these services are blocked in China and other countries (or so they complained).


I would have liked to have them run campaigns where they were able to do more A/B testing and where they were empowered more to monitor their own progress. Maybe some future iteration of this exercise can incorporate tools from real heavy-duty analytics solutions, like Nielsen or Omniture.

Final grades: more than 6,000 page views, but a big reveal in uniques

The final positions in this contest really didn't change that much from the positions after the second day. Scarlett's extensive email tree allowed her to jump into an early lead that she never relinquished. However, in terms of pure pageviews, Jingyi was steading narrowing the gap.


What really jumps out at me is how the graph started trending upward after bottoming out over the weekend. When I looked into the analytics deeper, it showed that a lot of this traffic was starting to come via social, rather than direct. What that means is that people were being encouraged by my students to share the link with their friends via social platforms.  However, the discrepancy between total visits and unique visits also reveals that there were a lot of users who were just clicking on the link again and again and just refreshing the page to try to drive up the total (check out the numbers in "Entrances."


This is not necessarily a bad thing. In the context of this contest, it shows that the students have managed to motivate some of their friends and family to click on and re-load their pages at least three times (if not far, far more). Now go on over to the What We Learned page to check out some of the publicity stunts and other creative means that students came up with to try to come out on top in this contest.

All items and materials copyright 2014 David LaFontaine

All rights reserved