sueddeutsche - leserzahl

News websites satisfy society's need for current information. In the absence of an editorial deadline they try to outpace each other on current events with live tickers and breaking news. I wondered to what extent the intensity of use of those websites follows the daily routine of their readers.

For several months now, the Süddeutsche Zeitung publishes the current traffic to their website. It is absolutely a good thing that a great news portal dares this step towards transparency and I hope that other sites will follow suit over time. Especially the combination of these datasets would create interesing evaluation options.

With the help of a small Python script I collected and evaluated these figures for the last month. Even though I had expected the overall behavior of the curve, some of the results still came showed a few interesting details.

For a better readability of the diagrams please click on the "Zoom" symbol with the arrows on the top right!

The average hourly user figures for August show quite clearly the regular "heartbeat" of news consumption: In the night users will read less than by day, on weekends less than on weekdays. This rhythm can clearly be seen in other data sets such as car traffic volumes throughout the day.

A closer inspection, however, shows that the numbers depend to some extent also on the news reports themselves. The announcement of the release of Gustl Mollath for example, is certainly responsible for the slight peak at 6.8. . To verify this assumtion, it helps to compare the numbers to search volumes shown on Google Trends. The lower numbers of 15 and 16.8 can be explained with a bank holiday and a bridging day.

Calculting an average week helps to better understand the "heartbeat pattern" of the dataset. The slightly lower traffic on Thursday and Friday could be explained by the holiday mentioned above. This shows the weak point of the dataset: A single month simply produces not enough data.

The average daily curve generated by work and weekend days shows several interesting characteristics: 

  • On weekday peak periods twice as many people access Sü as on weekends. Either readers satisfy their needs on weekends through other media types (like Sunday newspapers) or the interest in news is lower on weekends. In view of the good August weather, the second option seems very likely.
  • User counts are rising right after people are getting up in the morning (or probably even after waking up). Smartphones and tablets seem to be accessible in the vicinity of the bed. On weekends, the curve is delayed of course. The 5000 users mark is reached on workdays at 7:30 'clock in the morning and on weekends at around 9:00 'clock.
  • On weekdays, the influence of the lunch break can be very well seen: At 12:00 'clock we see a distinct rise in the user counts of about 1,000 pageviews.
  • Both curves begin to converge significantly from ca 19 'clock. Readers seem to go to bed mostly at around 22:00 'clock .