Behind the Curtain of Social Listening

The following blogpost has been contributed by our team in Germany. Many thanks guys.


SocialMediaActivitybyDaybyHourThe volume of user-generated online content has reached the potential for consumer brands to conduct extensive market research for quite some time now.


We tap this potential to help companies to better understand their customers and the perception of their brands. The groundwork is already laid out with the available data: The massive volume of consumer voices and unperturbed data collection ensure result validity. Consumers are not motivated through incentives and express their views intrinsically.


This allows long-term projects that monitor and analyze a brand’s environment and perception as well as short term analyses that provides a snapshot of these factors – a great advantage, compared to traditional market research.



Where Social Listening can provide answers

While the potential for Social Media Listening is huge, there are limitations on monitoring and analysis. The main questions where answers can be provided are these:

  • Who is talking about a brand? This includes influencer identification, stakeholder clustering and gaining insights into demographics
  • What is being said? Topics and subtopics can be analyzed and categorized as well as the context and opinions of users.
  • When are conversations taking place? Occasions, times of day and frequency of updates are important insights.
  • Where are conversations taking place? Different conversations, brands and issues are being discussed on a variety of channels and platforms.
  • Why are users engaging in specific conversations? Their motivation or the objective they want to achieve through the discussion as well as the way they do it need to be understood.


Example 1: Frequency and Occasions – Social Media Trendline


Social-media-trendlinesReplise has conducted a smart TV market analysis in time for the IFA 2013 consumer electronics exposition in Berlin. 19,541 mentions were collected in the defined time period from January 2013 to August 2013 with the Replise system on Facebook, Twitter, video platforms, forums and blogs.



The above graph shows when users posted content. If a set of particularly high peaks can be identified, the Replise analysts read a sample of the content to identify the issues that triggered the increased output. The two highest peaks in the example above were triggered by the Consumer Electronics Show (CES) in Las Vegas and a promotion by Samsung. Pulling the data from the Replise system and providing the first insights only takes a few hours.


Example 2: Topic clusters and Context – Social Media Conversations

Social-Media-Topic-ClustersReplise took a closer look at Samsung, LG, Philips and Sony for the smart TV research. To be able to analyze the context each brand is being discussed in, a sample of 1,661 posts was randomly chosen from the 19,541 original posts.



The sample is then read manually and the posts are categorized by trained analysts. The graphic above shows the shares of different topics Samsung is being discussed in and is only one example for the research results. Conducting this deeper content analysis takes about two days.


What is the right question for you?

Replise has standardized a variety of reports and research methods to streamline the research process. To find out more about our products and how you can leverage the data that is available about your brand, check out our Social Media analytics and listening products page. If you’re interested in our methods and would like to know more about our research design, don’t hesitate to get in touch with our Research Directors. If you’re a research professional who is interested in the potential of Social Intelligence, we’d probably like to hire you – check out our current job offers.


Lefty 2014. May 8.