Because the majority of RNW Media’s work takes place in an online environment, many of our measurement methods are focused on capturing online data. We use a mix of indicators and methods to capture the various aspects of our performance through websites and social media channels, each with their own opportunities and limitations.
Our primary indicator for measuring our reach is website sessions. This indicator counts how many times a user visits our website. Sessions, even when added over time, are more reliable to report on because we count how many times users access our websites, rather than implying that they are individual people. In addition to sessions, we also measure website page-views. This measures how many times our pages were opened and loaded by users. As a proxy indicator for people reading our content, we have started measuring how many people stay longer than 25 seconds on a page.
In 2019, we have invested significantly in upgrading our PMEL system, which has come with a critical review of how we measure and evaluate our (digital) performance. As part of that review, some indicators have been adjusted to provide a more accurate picture of our performance for both internal learning and external communications. This is also why we are now more and more talking about the total of content impression instead of the total content views as we reported in the last years.
Our primary reach indicator on social media is post reach, which counts how many users have had a certain post displayed on their screen, in combination with video viewers. For good measure, we also keep track of post impression, which is how many times a certain post has been displayed in anyone’s feed – regardless of whether they actually click or even see it. We measure social media followers, which counts every individual that likes or subscribes to a social media page or account. Because subscriptions are linked to unique accounts, this is a reliable number to use when communicating how many people we engage on social media. However, different followers on different channels are not necessarily different people. Therefore, when wanting to add up across social media, we can speak of the number of followers across social media. We collect engagement data: how many times users react, comment and share our posts. Reactions in this context means adding an emoji to a post. Furthermore, we have adapted our metrics to fit the unique Chinese social media landscape. We measure post/article views and post/article viewers rather than post impressions and reach.
*In the 2018 report, this number was 466,3m because it included Chinese video views, which were counted double (they were and are still also part of the total number of video views). **In the 2018 report, this number was 19,7m. This number has been revised in 2019 due to changes in the metrics and a correction made for the numbers related to This is Africa.
In 2019, we saw the impact of the reported changing behaviour around internet use. Young people consume content through apps and social media more and more rather than through websites. China is a case in point. The short-video platform TikTok, known as “Douyin” in China, has rapidly become a major player in the Chinese social media ecosystem. With over 500 million monthly active users, TikTok is currently China’s most downloaded video-streaming app.
We also experienced our own particular set of circumstances. Love Matters India and Hablemos de Sexo y Amor in particular have seen declines in web use. In India this was due to Google’s several major algorithm changes and the many shutdowns. India leads in the number of days of internet shutdowns. For Hablemos de Sexo y Amor, traffic dropped after the redesign of the platform in September 2018. Since September 2019, traffic has started to increase again but is taking some time to reach the previous level.
INTERESTING ENGAGEMENT NUMBERS
Apart from the data that we have reported on for the last few years, we have some more interesting engagement numbers through which we communicate with young people.