A top down approach looks at the bigger picture, and recommends building strategic solutions. A bottom up approach, on the other hand, focuses on solving the issue at hand in a cost and time effective manner. A key decision for an enterprise data strategy is to decide the approach for building data competencies.
Top Down Approach
A top down approach may sound good on paper, but is not necessarily feasible in all circumstances. It’s implementation involves huge planning efforts and political will within the organization. There is also a risk of getting stuck into a technology stack that is not catching up with time.
A top down approach, if not implemented correctly, may involve building solutions that do not necessarily address the business problems. It may not necessarily be convenient and agile for individual businesses. A top down approach may result into “one size fits all” kind of solutions.
Bottom Up Approach
Solutions with a bottom up approach may be able to deliver some good results quickly, but are not sustainable in the long run. Solutions designed with the bottom up approach provide design flexibilities, but may lack architectural consistencies, and may become unmanageable in the long run.
A bottom up approach may result in building similar repetitive solutions resulting in inconsistent results across the business, with increased costs and headache to reconcile the results on an ongoing basis. Bottom up approach can cause confusion, and reduce trust in data, thereby forcing businesses to rely on other techniques to take decisions.
Companies, almost always, prefer a hybrid approach for building their data strategies. The key principle of hybrid approach is the understand the specific situation of the organization and design a solution accordingly. There are mainly two factors that decide how a data strategy gets implemented within the organization:
Data Maturity Levels
If the company has grown significantly, to an extent that majority of the businesses rely heavily on data to take key decisions, then enterprise applications using top down approach are recommended to ensure consistency, standards, efficiency and quality.
If the applications of data include consolidation of enterprise data, then top down approaches are recommended. If the applications of data are focussed on a specific area and highly isolated from the rest of the organizational data, bottom up approach is recommended.