Big data can be used to make every stage of the strategic communication management process more effective and efficient - from situation analysis, strategy development and implementation to evaluation.
INTERPRETATION/EVALUATION: In the next step analyzed patterns, connections and forecasts have to be interpreted (prescriptive analysis) to derive recommendations for action.
Before showing the report to decision makers, the big data application process have to be evaluated. The evaluation shall show weather the right analysis is made with suitable data. The goals from the first step are the main reference point.
REPORTING: The last step is to prepare and convey the results for decision makers, f.e. in the form of network graphics, diagrams or with the support of real-time-dashboards. Fundamental data analysts and communicators should arrange, in which form and how often the reports should be produced.
It is important to implement the appropriate key performance indicators (KPI) for the communication goals. Automated processes are possible, f.e. alarm messages.
GOAL DEFINITION: Big data can support the separate phases of the communication management process, if there are goals and questions for the data analysis. There are separate goals and analysis for every phase of the process.
DATA COLLECTION/SELECTION: After defining the goals, a data pool has to be build. The experts examine what kind of data is available and what kind of data source can be used.
In accordance with variety, volume and velocity it hast to be specified: What kind of data has to exist to reach the defined goals? The generation of data is not banal because there are differente formats which need various possibilities to generate and save data.
DATA ANALYSIS (DATA MINING): In this step with statistical methods computer-based knowledge is gained (data mining). The analysis starts with defining goals for the data mining: What exactly should be analyzed? What method is suitable? Accordingly the suitable algorithm has to bo choosed to look for patterns, correlations or trends.
Subsequently there is a descriptive analysis. Here it is about frequency, the number of followers, likes or page views.
In the next step reasons, connections and patterns are examined with diagnostic analysis.
Also predicative analysis can be part of the data analysis.
DATA TRANSFORMATION: The step of data transformation describes the conversion of the raw data in other formats. The data has to be transformed so that they can be edited with the particular tool.
Another main goal of the phase is to reduce data by condensing single variables to characteristics.How to do the transformation in detail is related to the question in step one.
DATA CLEANING: After data collection the data have to be cleaned up and prepared for further processing. In other words, spam and irrelevant data have to be cleaned up, f.e. from bot-generated content. Mostly this can be made by algorithms.
In the following the texts are read randomly, proved manually, classed and eventually corrected. For experts this step is expensive but significant. Only if the data were collected correctly, they can be used for analysis.