Internet traffic data in Sweden and neighbouring countries during the COVID-19 crisis – insights and comparisons

By Anders Näsman, Senior Solution Engineer at Akamai Technologies

Inspired by colleague Martin McKeay’s blog post I set out to find the answer to the question: Are Swedes really behaving that differently during the pandemic? One way to look at behaviour, is to analyse the vast amount of data that we have in Akamai about Internet traffic locally and globally.

This is the data for February and March 2020:

The numbers represent all internet traffic, expressed in bytes served, delivered by Akamai into Sweden (not peaks); comparing February with March 2020. The blue colour shows the daily difference and the orange dotted line shows the 3 day rolling average trend. The red vertical line represents the day when Sweden’s authorities recommended people to be working from home.

In general, the traffic growth in Sweden has slightly decreased during February which is a natural effect of programming and event changes due to COVID-19, albeit with occasional spikes for software downloads.

What about the neighbours?

Here's Denmark

In Denmark we can see the traffic was marginal until the authorities locked down the country including schools and shops. After that traffic has grown. Note, that as in the case of Sweden, Mondays are low traffic days.

Moving on to Norway

In Norway the pattern repeats, but with an even stronger internet traffic increase after the lockdown.

And Finland

Finland went into lockdown a few days later, on March 16, but the situation here is very similar to Denmark.

Answers?

So, can I find the answer to my question "Do Swedes behave differently during the COVID-19 crisis" in this data?

Well, at least partially. Clearly the trend of traffic in Sweden is more even over the months and in line with the content available. Traffic "should" decrease somewhat as live events are fewer and people are still working daytime. Which it does. At the same time traffic has been higher in March, even in Sweden but not due to the lockdowns as in neighbouring countries. Not quite the "wall" of traffic that my colleague Martin McKeay has observed in other countries.

A cultural reflection:

1) It is spring and the weather in Sweden is really nice. As we are allowed by authorities to choose to leave the house – while maintaining social distancing - Swedes continue to enjoy the outdoors and not in front of their computers 18 hrs a day. (There seem to be more people in the trails running tracks and forests, though not shopping)

2) Swedish schools and pre-schools are still open. So people working from home can still do this without interruption.

What about capacity then?

Looking at our own metrics we can see that end user throughput is hardly impacted by the over all increased Internet traffic in Sweden.

When we're talking to ISPs in the Nordics it is clear that some are under pressure (we also see this looking at other available data, but that's for another article), but very few of them have recommended media companies to actually reduce their bitrates although this was suggested by the European Union**.

The Internet in the Nordics is built for far larger traffic spikes than we see with the more broad traffic increase of today, especially since most live sport eventsare on hold.

We are sure living in exciting times and it's going to be super exciting monitoring the progress of Internet traffic patterns over the coming days and months!

External references:

BBC reporting on Swedes behaving differently: https://www.bbc.com/news/av/health-52075983/coronavirus-sweden-s-life-as-normal

BBC reporting about EU recommendation on lower traffic usage: https://www.bbc.com/news/technology-51968302

Denmark: https://en.wikipedia.org/wiki/2020_coronavirus_pandemic_in_Denmark

Sweden: https://en.wikipedia.org/wiki/2020_coronavirus_pandemic_in_Sweden

Norway: https://en.wikipedia.org/wiki/2020_coronavirus_pandemic_in_Norway

Finland: https://en.wikipedia.org/wiki/2020_coronavirus_pandemic_in_Finland