Header Background Shapes Blue

Sizing Up the Data Hub for Your Enterprise: Is it Worth the Cost?

The first blog in this series explored the vision of the data hub as a converged platform that enables business managers and BI analysts to become part-time data managers and data governors. They can use the data hub to integrate, refine and explore data, and contribute to data governance, master data management (MDM), and data quality.

An ambitious vision! Because vision is not reality, this blog asks the hard questions to help enterprises determine whether the data hub would really benefit their environments. Data teams must reconcile the inherent tradeoffs between change and the status quo. On balance, the pain of data silos makes data hubs a viable bet for many enterprises. But each IT organization and data team should explore the following tradeoffs to come to their own conclusion.

TradeoffEvaluation Criteria
Ease of new tool
vs.
Learning & adoption
Does the data hub reduce the overall administrative work, or just redistribute it?
What are the training requirements to learn the new data hub?
Can the data hub enable us to retire another tool for simplicity
Benefits of reinvention
vs.
Cost of redistribution
Can we implement the data hub without undermining our governance and compliance initiatives?
Can we align with strategic initiatives – e.g., compliance, data modernization, or cloud?
Would it make more sense to reconfigure or consolidate existing tools?
Efficiency of one framework
vs.
Need to specialize
Will the data hub leave critical aspects of data management unaddressed?
Can we maintain an open architecture (data formats, APIs, etc.) to add or change third-party components?

Ease of new tool vs. learning and adoption. The data hub seeks to improve productivity by empowering business domain experts – i.e., managers and BI analysts – to integrate, refine, curate, and govern more of the data that drives operations and analytics. Armed with the right graphical tools, these domain experts likely can perform those tasks more efficiently than domain-agnostic data engineers and data stewards. But does that justify the opportunity cost? Time spent learning and managing these new tasks takes away from the core job, which means the productivity benefits need to be significant for the larger organization. The data hub needs to justify the training time and reduce overall administrative work.

Other factors influence ease of use as well. The data hub should further reduce administrative work and operating costs by replacing one or more existing tools. Business managers should have fewer interfaces to learn, fewer passwords to manage, and fewer support throats to choke. The enterprise should have fewer subscription fees to pay. If not, the value of the data hub declines.

Benefits of reinvention vs. costs of disruption. “Do no harm” rightly remains a top priority for any data team and IT organization. They need to ask whether they can implement a data hub without undermining, even briefly, their governance processes and ability to comply with regulatory requirements. The answer to this question depends on data hub functionality – e.g., its methods of handling Personally Identifiable Information (PII), alignment with new legislation such as the California Consumer Privacy Act (CCPA), and reporting capabilities. The answer also depends on how smoothly the IT organization could roll out the solution and adapt existing people and processes to manage it.

Many enterprises can grease the skids of adoption by merging a data hub rollout with existing strategic initiatives related to compliance, data modernization or cloud migration. They should verify and scope these possibilities before committing funds to a data hub. Perhaps most importantly, they should avoid chasing shiny objects and also consider the alternative of just reconfiguring tools they have today. Perhaps their existing MDM vendor just rolled out enhancements that greatly accelerate their bottlenecked match and merge process.

The efficiency of one framework vs. the need to specialize. Data hubs offer a suite approach, which raises the usual questions about best-of-breed alternatives. This suite should not impede any mission-critical aspects of data management. Certain business units and teams inevitably need to specialize. For example, data scientists might need to assess customer sentiment with natural language processing scripts. The data hub needs to accommodate such specialization, for example by integrating custom script plug-ins. It needs to support an open architecture with open data formats, open APIs, and easy integration with third-party tools.

Like death and taxes, you can count on the convergence of technology tools. Equally certain, however, is the fact that data environments keep getting more complex, and data silos keep multiplying. So enterprises need to take a hard look at the promises of converged tools such as a data hub to truly simplify their environments. The data hub likely passes the test for many enterprises, but be sure to do your diligence on evaluation and selection.