Recently I heard the frequently used phrase again: we have to provide “the right data to the right people at the right time”. The phrase is related to business intelligence and hints that we should provide business users with the right data at the right time to enable them to make business decisions.
Have we ever really asked ourselves what does this mean?
The data warehouse should be the source for all data on which to base business decisions, so let’s start here. Do we really have all data that is needed to make business decisions in our data warehouses? The answer is generally NO. Data warehouses usually don’t encompass all data sources in a company and there is always a little something that is still missing. There are many reasons for this, such as that a data source hasn’t been loaded into the data warehouse yet because of a development backlog, or that a data source is not available in digital format, or that the company doesn’t even have the data and needs to make arrangements to get it. Business decisions that are made based on incomplete data are not very useful.
Secondly, what is the quality of our data? If we ask the information system administration team they will immediately confirm that their data is perfect. But when we dig into the data, we will most likely discover inconsistencies, missing or illogical data and so on – and these are only problems that we can discover by looking at the data. We would probably never be able to identify data that is of a poor quality because someone made a mistake and entered it incorrectly.
Yet another aspect of data based decision making is to question the version of the truth that a particular set of data represents. Many companies still use “spreadmarts”, that is different versions of data that float around various Excel files because each user organizes data the way they like. Sometimes this is redundant, but sometimes it is justified, for example, when a department needs to filter data a certain way for their use while a different department needs something else. Therefore there is no one single version of the truth in the data warehouse. Data should be considered from different perspectives.
Everything hinges on data. Management usually understands that data has value and that without it there wouldn’t be many business decisions. It is more difficult to understand that the data has to be nurtured, maintained, and governed from a corporate perspective. Has anyone ever calculated the costs associated with incorrect, missing or bad data (for example, lost opportunities)? How much does a bad decision based on bad data cost?
It isn’t enough to be aware that we need data. The data has to be of a good quality and available to be useful.