Social CRM in the practice: Databased decisions

The everyday presence of social media in our lives is indisputable. In many companies, traditional call centres are changing their communication channels to direct news via Twitter, Facebook or WhatsApp. In these channels, consumers began to express their opinions and experiences in terms of products and services, which led to the well-known phenomenon of digital word of mouth.

The analysis of such data has become essential. However, this vast amount of User-Generated Content (UGC) makes manual analysis impossible and prevents the extraction of knowledge that gives companies a competitive advantage. In this sense, automatic data analysis by social media has gained its place in customer relationship management.

Foster Provost and Tom Fawcett[1] in a leading article rightly stresses the relationship between Big Data, Data-Driven decision making and Data Science, with particular emphasis on the impact on the economy. Let's look at the concepts closer to understand this relationship. The following factors characterise big data:

i)Volume - refers to the quantity of data to be analysed. For example, on Facebook, over 10 billion news, 5 billion interactions and 350 million photos are shared daily;

(ii) Speed - The factor refers to the rate at which these data are generated - as illustrated by the previous example. Besides, this speed prevents efficient real-time data analysis;

(iii) Variety - In the past, most data available for analysis were structured in relational databases, i.e. in table form. Currently, more than 90% of the available data are unstructured (and quite different) in nature, such as texts, videos, photos, locations and audio. Such data types require new analytical tools;

(iv) Truth - This factor is crucial for companies since information must be accurate. In times of fake news (Fake News) and indiscriminate bot use, information distributed in social media should be evaluated and filtered to ensure the reliability of results;

v) Value - In the end, these data must have a potential business value that provides insight for more accurate decision-making.  

Data-Driven decision making refers to the practice, decisions based on data analysis results and not to make decisions based on impressions or intuitions. Finally, Data Science can be defined as a multidisciplinary field that uses scientific methods, processes, and algorithms (triad) to extract knowledge and insights from data, whether structured or unstructured.

In this triad (scientific methods, processes and algorithms), the analysts aim to develop efficient methods for analyzing large amounts of data, revealing new and potentially useful insights for decision-making in specific industries.

With the data analysis expertise of the Laboratorio de Inteligência Computacional de Universidade Federal do Pará, Brazil and the Laboratory of Computação Aplicada da UFOPA, Brazil, researchers have linked the understanding of the requirements of the social CRM market with those of companies working with the Social CRM Research Center. At the University of Leipzig, we have created a mapping with these three elements: 

  • Methods for analyzing social media data,
  • Decision-making in the context of customer relationship management,
  • To extract knowledge and learned lessons.

The first part of the mapping was carried out by research into existing literature and consultation of professionals in the field of CRM, which is multidisciplinary. For this reason, we involved experts from other areas such as marketing, administration, computer science and design.

As mentioned earlier, all areas can benefit from the analysis of User-Generated Content. Understanding customers' aspirations and ambitions enable them to develop new products (through collaboration and innovation), improve the user experience, help plan marketing campaigns, increase sales, and support after-sales communication through social media.

In a second step, the researchers involved in this project analyzed the social media data types, such as user profile, text contributions, site sharing, etc. Selecting the data allowed the analysis capabilities to be determined, as shown in the following figure.

Through confrontation between the sectors involved in the decision - making process and possible analyses with the data available, the possibilities for use of the results in decision - making processes have been mapped. A summary of the outcome can be seen in the following table.

Table 1: Potential analysis and services in the context of social CRM, adapted from (Lobato et al., 2017).

  CRM-related industries

  Portfolio of possible services in Social CRM

  Distribution

  Product recommendation

  Purchase Forecast

  Identification of leads (potential consumers)

  Marketing

  Market analysis

  Social-Media Campaign

  Adaptive brand management

  Impact assessment of advertising campaigns

 Service & Support

  Social Media FAQs

  Automatic assignment of jobs to the relevant industries

  Community-Support Forums

  Innovation

  Social corporate networks

  Identification of the wishes and needs of consumers

  Identification of trends

  Collaboration

  Identification of digital influencers

  Recruitment of employees

  Experience of consumers

  Recruitment of brand ambassadors

  Identification and promotion of influencer communities

 

It is interesting to note that several tools already exist, which provide some of the analysis presented. Others require detailed and individual studies[2]. Our research at the Social CRM Research Center focuses on solutions to deepen and refine these analyses with the support of a team of social media researchers and analysts capable of gathering, analyzing and interpreting knowledge to advance companies.

 


[1] Data Science and its Relationship to Big Data and Data-Driven Decision Making Foster Provost and Tom Fawcett Big Data 2013 1:1, 51-59

[2] A continuing discussion of analysis potentials in social CRM can be found in the article social CRM: Biggest Challenges to Work in the Real World, submitted in the Workshop on Intelligent Data Analysis in Integrated Social CRM, published in19th International Conference on Business Information Systems.

 

Translated into English by Frank Amankwah.

Guest author

Fábio Lobato
Fábio Lobato is a professor at the Institute of Engineering and Geosciences of the Federal University of Western Pará. Since 2015, he has been associated with the Social CRM Research Center, where he works in data analytics to support decision-making.
His research spans all stages of the business data analytics process, from understanding the needs of the business to selecting data sources, extracting knowledge, validating results, and incorporating knowledge into decision-making processes. The focus is on process optimization and the development of new analytical methods aimed at giving companies a competitive advantage. Ostensibly, small and medium-sized enterprises are considered, taking into account their resource constraints.