The analysis of data is littered with potholes.
Here is a list of 7 deadly sins which affect the use of business intelligence:
- Poor data quality. Data can suffer from a number of ailments – incorrect values, missing values, default values, and so on. Don’t underestimate the effect that poor data quality can have on your reports and charts.
- Jumping to conclusions – relatively small data sets contain random behaviour that can easily be interpreted as significant. Interpreting random noise as a trend is very common.
- Inappropriate durations – analysing data which span a time period that is too long or too short will inevitably produce meaningless reports and graphs. It requires good domain expertise to estimate over what period data should be analysed.
- Analytical overload – too many reports and charts confuse the picture more than too few. Clarity and significance should be the guiding principle.
- Data silos. This should be a thing of the past, but it isn’t. Functional silos still exist with various departments, jealously guarding their data. This inevitably means incomplete and misleading analysis.
- Analysis paralysis – the effect of having too much choice in analytical techniques and tools. Keep it simple.
- Misrepresentation – using area or volume based charts (bubble charts for example) without understanding the magnification of effect that takes place.