Select a page
Strategic Investor Relations for Technology Companies

Do You Really Need a Big Data Strategy?

big-data-puzzle-blue

With increasing frequency, CIOs are being asked by their senior management, “What’s our big data strategy?”  But do you really need a big data strategy?

In our view, companies should instead focus on data governance and data management.  If data is a company’s most important asset then a formal data governance program and data management best practices are among the most strategic investments it can make.  The absence of either is why companies are challenged to harness their existing data resources – no less the torrent of big data that will continue to inundate the enterprise.

Data governance and data management together demonstrate how a firm understands and uses its data assets as well as how those assets are managed over time.  Data governance and data management become more strategic as firms evolve from static, database-centric systems of record that report on historical results toward dynamic, real-time systems of engagement that generate insights faster and inform better decisions on fresher, more accurate data.

Tech-Tonics believes that investments in newer big data technologies should be a puzzle piece of data governance and data management strategies – and budgets.  Get these strategies right and it will be easier to integrate big data into analytics workflows to improve decision outcomes.  Modern systems allow IT to develop the framework that business users can access without having to rely on knowledge of what may be needed.

A formal data governance program coupled with effective data management represents a better path to improving return on data assets (RDA).  Companies with higher RDAs are more competitive.  They outperform their peer groups in achieving corporate return on investment (ROI) and risk management objectives.  In turn, these factors drive higher market valuations.

Raising RDA is hard, time-consuming and political.  It requires a commitment by all stakeholders – especially senior management.  Successful execution is nothing short of cultural change.  But the payback on becoming a truly data-driven enterprise is compelling because it can yield tangible gains in productivity, end-user satisfaction (both employee and customer) and sustainable competitive advantage.

Tech-Tonics recommends firms take a holistic approach to data governance and data management.  A well-executed data governance program with unified data management provides a structure that supports flexible access to consistently accurate, high-quality data and analytics.

Data Governance for Risk/Return and Compliance

The purpose of data governance is to ensure that information accessed by users is consistently valid and accurate to improve performance and reduce risk exposure.  The more users trust the quality of the data they are working with the more reliable and predictable their models and decision outcomes will be.

There is a direct relationship between data governance and operational performance.  As more companies expand the use of associative search and data visualization tools to a wider user community the value they derive from analytics is proportional to the quality of their data.  Data quality supports data governance by making sure that data assets are reflected correctly within data stores and throughout business processes.

As analytics and decision-making bifurcate throughout the organization, assessing and managing the risks associated with data lurking within the enterprise – and coming from external sources – becomes more complex.  Data governance also provides consistency to data that is strewn across departmental and organization silos.  As users’ appetite for data sources continues to grow, a framework that emphasizes data quality facilitates integration of existing data sources with external ones, including newer big data formats.

Figure 1. Data Governance for GRC Requirements

compliance-regulations -binders

Source: Dreamstime

While many companies have made substantial infrastructure investments to speed data delivery, bad data at the speed of light is still bad data.  Firms can achieve higher RDA and mitigate decision risks with a long-term data governance strategy that supports deployment of modern analytics technologies.

In financial markets for example, advanced analytics can serve as an early warning system for market operators and participants to prevent disruptions caused by trading errors, improper systems oversight or other compliance violations.  These solutions can provide better visibility into operational data and reference past experiences to help predict system-impacting occurrences in advance.

Advanced analytics can also help monitor and prevent abnormal client behavior and to detect risk exposures through internal or external fraudulent activities.  Firms can mitigate risk, reduce service outages and better adhere to GRC (governance, regulatory, compliance) reporting requirements.

As part of data governance, firms should create standards by defining a set of best-practices or principles that will ensure the organization creates and maintains good quality data.  Strong data quality informs a deeper understanding of the key performance variables (KPVs) that drive decision-making and the business.

Figure 2. The Data Governance Cycle

Governance Graphic

Source: Tech-Tonics Advisors

The key factor for success of a data governance program is an environment that promotes communication, collaboration and trust among and between business users and IT.  A formal data governance program includes safeguards on data handling and usage, with clearly defined rules, roles and responsibilities.  Data ownership becomes the provenance of the data governance program with all constituents held responsible and accountable to adherence by a data governance council that oversees the program.

Working together teams can identify the existing information infrastructure, applications and data sources that drive workflows.  Understanding business and technical requirements to identify the value data provides is the first step to developing a data governance cycle.  It also helps pinpoint bottlenecks and inconsistencies in data elements that inhibit optimal data usage and oversight.

Once data sources have been identified, they must have uniform definitions – such as information pertaining to a customer, security or counterparty.  The program establishes policies and procedures for data handling, and ensures that users of that data are clearly identified and authenticated.  The role of data stewards includes ongoing monitoring of data sources and data quality to ensure adherence with best practices.  Diligent governance helps maximize RDA while limiting risk exposures, including non-compliance.

Metrics that establish adherence to data governance policies map back to data handling rules that are used to create baseline key performance indicators (KPIs).  These KPIs can then be used to gauge how effective the data governance program is as reflected in operational performance and risk management objectives.

Without data governance, it is impossible to know whether the information presented is accurate, how and by whom it has been manipulated.  And if so, with what method, and whether it can be audited validated or replicated.  As departments maintain their own data – often in spreadsheets – and increasingly rely on outside data sources, a verifiable audit trail is compromised, exposing the firm to compliance violations.

Unified Data Management is Critical

Too often, decision makers only have access to pieces of information – or conflicting information – and make decisions based on an incomplete view of what is happening.  After data sources have been identified and harmonized, homing in on KPVs represents the next step in data acquisition.  Data needs to be integrated to provide connections between datasets in order to perform analytics.

However, integrating data from a wide variety of formats and data structures without considering data quality is not only costly but can also have a catastrophic impact on enterprise systems that rely on that data.  From a recent poll, Gartner estimated the average cost of poor data quality to an organization at $8.2 million per year.  Moreover, 22% of respondents estimated their annual costs at $20 million or higher.

As a result, Tech-Tonics believes firms should adopt centralized data management.  With all users accessing the same data stored in a unified platform, or by having users access federated data stores that have been validated, trust in the data rises and results in more reliable models that drive decisions.

A strong data management platform ensures users can access data assets as they need it, share information when required, and have the tools to “see” analytics results without the pre-definitions of restricted data sets inherent in legacy business intelligence platforms.  Information remains consistent so that even when users within different business units require individual insights, the underlying data will be monitored to ensure that it is accessed within a governance framework.

Although different business units may define entities differently, it is still important to make sure that all users access the same data.  Firms need to manage diversity of terminology and definitions by maintaining strong metadata while providing users with the flexibility to analyze data using modern tools.

Figure 3. A Data Lake Centralizes Data Management

Data-Lake

Source: Dreamstime

Effective data management also eliminates the need for a separate big data strategy.  Big data becomes part of the firm’s data management strategy.  This is especially true as 90% or more of data that drives business processes and decisions are from traditional sources.  As more unstructured data is incorporated into workflows, technologies can be introduced to aggregate, analyze and manage these newer data formats in an evolutionary manner.

It simply doesn’t make sense for most companies to either rip-and-replace their existing data management infrastructure and analytics tools or to risk creating yet another data silo for the relatively small component of big data that drives their analytics and decision processes.  Positive results are more likely with incremental steps to analytics projects and business transformation.  Success with smaller user groups and data sets begets confidence from all constituents and makes it easier to get funding for expanding the project to the next phase.

As for the question, “What’s our big data strategy?” the best answer may be that “Big data is part of our data governance program and data management strategy”.

This article first appeared in iCrunchData News