IT executives define big data architecture in terms of where they process and place their data, whether data is in-house, in data marts, in public clouds, etc. But there’s a second meaning to big data architecture that needs considerable attention: How are you architecting your data for the business?

Business data architecture is not synonymous with the technical architecture for big data processing and storage. However, it has everything to do with whether you’re building the best information assets for your company to optimize its big data and analytics.

Few companies succeed in this big data quest for relevance without identifying the foundational pieces for their business information. The foundational underpinnings for your big data are the primary vectors or entry points into the data that give the business the best capability to navigate through data and derive meaning.

Here are several examples of foundational vectors into data that are common in business today.

SEE: IT leader’s guide to big data security (Tech Pro Research)


Almost every company uses locational information (and the big data aggregated to it) to evaluate where vehicles are located if they are in logistics, where at-risk populations for heart disease are if they are in healthcare, or where the most likely buyers are for their products if they are in retail. Having a geospatial path into big data is critical.


Manufacturers that want to improve their products want to see data organized around products. This information can come in the form of engineering specs, commentary from manufacturing about production issues, or input from sales or customer service. Product is the organizing construct for this information–and a foundational building block for product companies.


A healthcare company trying to produce diagnostic prescriptions for stroke victims in hard-to-reach areas wants to perfect a detection and diagnostic engine that can help rural physicians treat and respond to strokes. To perfect the tool, the company runs many algorithms multiple times to refine the diagnostics and to ensure that a broad enough spectrum of data is analyzed. That data becomes a foundational construct.

In each of these use cases, the foundational pieces for information–how you access and aggregate data–meet immediate business needs. They also provide building blocks on which the company can create tools for more applications. For instance, if you perfect your diagnostics engine for stroke, you might want to build out this concept for a new product that diagnoses and prescribes treatments for cancer. In this way, you continue to leverage your data investments into more business opportunities.

SEE: 60 ways to get the most value from your big data initiatives (free TechRepublic PDF)

Here are three best practices that assist in business information architecting:

1. Architect your business information structure at the same time you architect your big data processing and storage infrastructure. This ensures that the two are synchronized and that they don’t work at cross purposes.

2. Cross-check your information foundation assumptions. This can be done by visiting with other executives in the business to ensure that the access and aggregation paths into the data are how they want to dissect the data–and that they feel these foundational blocks will hold up over time.

3. Continuously evaluate your business information structure.

Businesses, like technology, constantly change. In the best possible scenario, you will always be spot on with your information architecture as your business evolves. In other cases, new access and information aggregation paths may be needed for your big data that you couldn’t have imagined before. You can build these as needed.

Also read…

Your thoughts

Has your organization been successful in building a business architecture that enables you to derive optimum value from your big data? Share your experiences and advice with fellow TechRepublic members.