Experience PageDriller
In the past few months we have been working together with Dagens Nyheter AB to create PageDriller, and we are very proud to announce that we have made a real-time tool for frontpage monitoring which is easier than ever to access, fun to use and gives an incredible depth to intelligent news publishing online.
- DN.se's goal has been to push for a report that is as fast and easily accessible as possible. It's the only way editors will adopt it and give them value on a daily basis. Less is more, is for real in this case. OnSite has shown an excellent understanding for our specific needs for real-time measurements.
Bjørn Lilja, Dagens Nyheter AB
Curious to know how we did it?
First we took the data that we already have available to us in LinkPulse. Namely, information about clicks, where they happened, where they lead to and what position on the page the link was in.
All this data was already there in real-time in LinkPulse. With the use of tabular report called Editor's Eye, we have utilized this to give editors an idea of what is going on on their frontpages for years. We compared traffic on stories from a short-term perspective (15 minutes) to a long-term perspective (3 hours) to tell the editors in big capitals how the story was performing. The editor could then deduce where the story happened to be in its life cycle and make necessary changes to it, place it more prominently or remove it altogether.
And even with our popular Toolbar, we have been able to visualize this information, on top of the frontpage (also known as heat maps, but we hate that expression).
In PageDriller, not only did we combine Editor's Eye with the Toolbar, we also did something about the intelligence.
We figured that what we already had didn't compensate for traffic fluctuations during the day. Most news sites (and other web sites) have specific traffic patterns throughout the day. In Editor's Eye, editors had to correct for these fluctuations in their brains, by knowing that just before lunch, everybody surfs the news so all the stories point upwards at that time.
We wanted to give the editors a gift; the gift of knowing how a story performs independent of such fluctuations. So what did we do?
We simply took the short-term data and compared them to similar time-frames in the past. For example, if today is Wednesday and its about noon, we compare the short-term data to the same time-frame around noon in the previous few Wednesdays.
The optimal configurations of how many historical days to consider, as well as the size of the short-term window, varies from site to site, so we made these into parameters that can be manipulated by each end-user.
Visually, the information is presented as an over-lay on the frontpage, with colors signifying the relative performance of each position or story, in addition to a "speedometer". At the top of the page, we display some key metrics about the overall performance right now, today and the current week.
Tags: