Business Intelligence
Business Intelligence (BI) is best described as business insights by means of professionally prepared data:
- Environment: BI data may or may not be embedded in a complex DWH landscape. Complex ETL processes may possibly be required.
- Volume: The data volume can be very large ('Big Data'). The
- Focus is the analysis of internal business data (be it sales, production, behavioral, or other data). Competitive Intelligence usually focusses more on analysing business external data. The
- Methods are from classical, hypothesis-driven statistics, as well as the more exploratory Data Mining or Data Science, and the corresponding standards , e.g. CRISP-DM 1.0, Catalyst, DFG, DeGEval / SEVAL, etc. With
- Results in the shape of professionally processed data, as insight in the shape of key performance indicators ('KPIs'), texts (reports, executive summaries), lists and tables, visualizations, or as effective dashboards or cockpits (Visual Analysis) in
- Real-time or at the push of a button you can see the state of an enterprise, a process or certain KPIs.
- Objectives are:
- Professionalism and transparency of methods and procedures
- Modelling/Elimination of uncertainty
- Generation of (better) results (Added Value)
BI does not look only into the past and answers the question: "What happened?" or "Why did it happen?". Predictive Analytics (see a selection at Advanced Analytics) and Data Mining try to answer "What will happen?" BI is not objective, however. The non-objectivity already begins with the selection of the fields, methods, and may well also end in the interpretation. Therefore it needs the input from the business side. The topics of an enterprise often make the difference:
- Banking and Insurance
- Marketing, Markets and Tourisms
- Customers, Services and Competition
- FMCG/SMCG
- Engineering
Business-specific approaches may be:
- Data/Business Process Modelling and Re-engineering
- Analysis of Gap, Downtime, or Performance
- SPC Statistical Quality Control
- Development of especially business-specific customised KPIs
- Genetic Algorithms (z.B. Travelling Salesman Problem)
- Solution of special planning problems (i.a. Operations Research)
- Data Quality (e.g. COPQ, Audit Trails, and criteria like completeness, missings, consistency, duplicates etc. BI is only as good as the quality of the data analysed.
- Project related Management and Analysis Methods, e.g. SWOT Analysis, clarification of requirements and support (Stakeholder Mgmt), internal and external communication and documentation, and much more.