RSMS – The Real-Time Social Media Sentiment Indicator
Germany is a global player in the economic market. Its economic success comes from internationally developed companies like Daimler, VW and Adidas. Companies, as well as the Ministry of Finance, rely on economic forecasts which predict the future condition of the economy, specifically GDP growth rates, using a combination of various indicators. Businesses and the government base their decisions regarding investments, hiring, spending, and other important policies that impact the economic activity of a country on those forecasts.
The problem is that economic forecasts are prone to failure in many people’s eyes.
To improve economic forecasts, we focused on the target variable private consumption, which is predicted by the consumer sentiment. The private consumption is of uttermost importance for the health of our economy as it makes up 60% of the Gross Domestic Product.
The current capture of data for economic forecasts is done by surveys which are carried out by institutes as, e.g. the Leibniz-Institut für Wirtschaftsforschung and the Gesellschaft für Konsumforschung. German citizens provide information about their consumer behaviour via questionnaires or over the phone.
However, here is the problem: Not all German citizens are involved in these surveys.
German teenagers, who make up 10% of the total German population, are not represented in economic forecasts.
That is why our team KonjunkTool established the Real-Time Social Media Sentiment Indicator (RSMS).
Against the backdrop of digitalisation, we decided to use social media to reach a large number of teenagers in Germany. The most popular social media platform among 13 to 18 year-olds is Instagram with 71%.
Our solution is based on polls promoted on Instagram by which we will be able to get the teenagers participating in our research.
A business Instagram account enables us to launch advertisements, which reach up to 200.000 teenagers all over Germany. These teenagers are being asked just one short question:
“Do you plan to spend more or less money over the next 7 days than you have spent over the past 7 days?“
The results of this survey show us in real time whether the private consumption of the surveyed is more likely to rise or to decrease about specific developments in lifestyle and economy. Those statistics can be used by economic institutes such as the DIW.
By introducing this new indicator in macroeconomic forecasts, not only do we represent another 10% of the total German population but simultaneously improve the accuracy of economic forecasts with just a small effort: answering one question while being online!
The development of our idea was done without external sources but mainly through the reliable expert information by Dr Fichtner.
- slides_YES_Fichtner.pdf (22.02.2018) (presentation from the Kick-Off Meeting uploaded in the Cloud)
- Döhrn, Roland 2014: Konjunkturdiagnose und –prognose. Berlin und Heidelberg.
- Rosenblueth, Arturo und Norbert Wiener 1945: The Role of Models in Science, Philosophy of Science, Vol. 12, Nr. 4, S. 316–321.
- Taleb, Nassim Nicholas 2013: Der schwarze Schwan. München.
Contact us: firstname.lastname@example.org