Why your data strategy should be strictly agile

You may think the title is contradictory, it is, but it’s true.

It’s vitally important for all businesses to have a cohesive and comprehensive business data strategy. Your strategy should detail the data lifecycle, ie how the data flows through a system

The business data lifecycle

Simple example data lifecycle diagram

Lifecycle processes control the flow of data from left to right, these processes may include validation, correction, data enrichment etc, the key is that the processes tightly control the Data Lifecycle. The Analyse lifecycle stage is where data engineering and science activities occur, with tight control being placed as data re-enters the Data Lifecycle.

A data Taxonomy should be developed that meets your business, it defines the possible classification and categorisation for your data, it should contain an appropriate number of categories and sub-categories to meet your business needs. It’s important to consider both current and possible future business directions when defining. 

Managing the data through its lifecycle and classification of said data is only part of the story, Metadata is the key to making greater use of the data you store. Consideration as to the metadata you collect and store and who, how and to what degree metadata is viewed needs consideration and planning.

Business data lifecycle enforcment

It’s important that once the business data strategy is in place the rules are enforced. Without controlled enforcement the processes that control the data flow can’t control that flow, data is not classified correctly the metadata has no home and therefore is of little use.  Your systems need to know how to process data through its lifecycle, it needs to be able to classify your data and the metadata held needs to be available and add value.

Business data lifecycle agility

There will always be arguments regarding why one data source should be the exception to the rule and why some teams will often believe that they are somehow different and special. As long your business data strategy foundations are sound and have considered all the business needs more often than not, you’ll be able to explain how and where their data source and process flow requirements fit into the existing strategy

Occasionally in a dynamic and ever-changing business world, these arguments will be difficult to win. You may be able to put a case as to how the business data strategy supports this data source but when doing so consideration needs to be taken into how much it effects the processing of the data and more importantly how it impacts on the usefulness of that data source. New and unforeseen circumstances will arise that either don’t fit within your strategy or would require unacceptable compromise in the processing or usefulness of that data source.

So though “It’s important that once the business data strategy is in place the rules are enforced” ultimately you also need some flexibility. Where the arguments no longer stack up and data no longer fits your processes, flows may need to be re-considered, your taxonomy up-dated or your metadata rules amended to meet the latest business objectives.

When making changes to any element it’s important not to rush, each impact needs to be considered in the round. What will be the impact on exiting processes and data, what is the general business direction and are there other elements that should also be considered.

Conclusion

While being strict about enforcement of your business data strategy rules those rules, you also need to be ready to adapt those rules when business changes, in the long term an out of data or unworkable strategy is not much better than no strategy.

About the author