In-Depth

TDWI Conference Kernels: a Data Strategy Survival Guide

If your company lacks a data strategy, can you really reap the benefits of business intelligence?

Bellwethers tell us the direction of wind; in business, whither the market goes. The record-breaking attendance at TDWI’s Spring conference would thus seem to indicate that business intelligence (BI) continues its rapid growth into the mainstream. Despite that expansion, roadblocks still abound, one of which boils down to a lack of data strategy in many corporations. This issue took center stage at the Monday keynote in Chicago, the proverbial City of Big Shoulders.

In her speech, Larissa T. Moss, president of Method Focus Inc., covered a spectrum of issues related to data strategy:

  • why organizations need a data strategy
  • how an archaic mindset adversely affects BI initiatives
  • key components of a successful data strategy
  • roles and responsibilities involved
  • the impact of a data strategy on BI

To open her talk, Moss detailed two hypothetical scenarios: one in which a CEO asks the CFO for a complete inventory of their company’s financial assets, to which the CFO sheepishly responds that they have no such handle on their finances; and a second in which the same CEO asks the CIO for a similar report on the company’s information assets, to which the CIO sheepishly responds that they have no such inventory.

“Nobody should be laughing at this point,” admonished Moss, “because this is true: we don’t treat our data as a corporate asset.” She continued: “After 40 years of data processing, we’re still concentrating on the processing more than the data.”

Moss then offered a possible explanation for why decades of work in data processing still have yet to result in organizations understanding and appreciating the importance of treating data as a corporate asset. First, dating back to the early days of computing, two camps have existed in nearly all organizations: the business side and the information technology (IT) side. This occurred because, way back when at least, computers were so new, and required such specialty for operation, that engineers were needed to handle the IT side of the house. This dichotomy set the stage for decades of internal struggles.

Second, an old worldview dating back even decades further has still maintained its grip on the modern world of work. “When we automated our operational processes, we typically had to automate all of it. That mindset is still with us,” she said. “The driver is an industrial age mental model.”

Moss explained that in client engagements, she asks companies to prioritize five categories for any given BI initiative: time, scope, budget, people and quality. She said that while most executives want to rank all five at the highest level, she insists that they choose one level for each, in order to see where corporate priorities truly reside. Inevitably, time jumps to the top, scope takes second, followed by budget and people. Of course, that leaves quality last!

“Quality usually falls of the boat before the boat leaves the dock,” exclaimed Moss. She then walked the audience through a process of elimination, resulting in a conclusion that scope must always land at the bottom. “We have to bite the bullet at some point,” she explained. Thus, in the information-age mental model, quality comes first, time second, people third, and budget fourth.

What are the components of a comprehensive data strategy? According to Moss:

  • data standardization and integration
  • data quality
  • metadata management
  • data modeling
  • security and privacy
  • performance
  • DBMS selection and management
  • unstructured data
  • business intelligence
  • data stewardship.

On the issue of data quality, which Moss described as paramount, she offered a graduated scale to describe the different levels of maturity. From nascent to mature, the stages are: discovery by accident (uncertainty), limited data analysis (awakening), addressing root causes (enlightenment), proactive prevention (wisdom), and optimization (certainty).

On the topic of metadata management, she noted: “Business metadata is the stepchild of data; because that’s the hardest one to compile; because it’s in people’s heads, in people’s minds.” Of course, it’s also critically important metadata, as it provides the glue that holds integration efforts together.

As for data modeling, Moss extolled the value of the oft-elusive business requirements modeling. Sure, many organizations take time to build their data models carefully; but such models that do not map to complementary business requirements models will almost assuredly fall short of effectively facilitating the insights that comprise the pot of gold at the end of the BI rainbow.

Covering a range of issues, Moss then offered insights on: performance, stressing the importance of benchmarking, capacity planning, designing optimal schemas, coding efficient SQL calls, monitoring/measuring usage, and tuning the environment; DBMS selection, which should be driven by the size and maturity of a BI environment; unstructured data, which she called “the black hole of IT,” for which a unified content strategy is required; and the ultimate goal: business intelligence, which should provide decision-makers with that coveted 360-degree view of their business. “I don’t know how you’ll get BI without a data strategy,” she stated.

Moss surmised that data ownership and data stewardship should take precedence for any company that hopes to achieve the fruits of effective BI. In order to get there, she recommended not one or two, but a whole team of data strategy professionals, including:

  • a data strategist who understands the strategic business goals, DBMS platforms, internal application databases, future data demands and data volumes—this person creates and maintains the data strategy


  • a strategic architect who develops the overall architecture for both operational and BI environments to include software, utilities, tools and interfaces—this person determines if the environment will be one-tier or multi-tier; and participates in architecting database and data flows


  • a database administrator/architect, who understands user requirements, knows different database design techniques and when to apply them, is responsible for the physical aspects of application databases; and maintains the application databases themselves


  • a data administrator who knows the industry and business processes, understands the corresponding data and business rules, has expertise in modeling and knows normalization rules—this person standardizes and integrates the data through the information architecture, creates and enforces data naming standards, and collects/maintains the business metadata


  • a metadata administrator who knows industry metadata standards, understands databases and ETL architectures—this member of the team builds and maintains a metadata repository or administers a purchased MDR product; selects and installs metadata integration and access tools; integrates and loads metadata from various BI and developer tools


  • a data quality analyst who knows the internal application databases and how to extract data from them; is familiar with data profiling and data cleansing tools; understands the user requirements, business processes and rules—this person audits operational source data, participates in writing data cleansing specs, identifies root causes for dirty data, and facilitates negotiations between data originators and information consumers about data quality improvements


  • and, last but certainly not least, a security officer who knows the applicable governmental security and privacy regulations, understands the business requirements for securing the data, as well as the security features and capabilities of the application components; and ensures that appropriate security settings are placed on databases, BI tools, developer tools and Web portals alike.

With such a team of professionals guiding an organization’s data strategy, Moss contended that the ultimate goal of BI can indeed be achieved, resulting in better and faster decisions, increased analyst productivity, employee empowerment, cost containment, cash-flow acceleration, revenue enhancement, fraud reduction, demand chain management, better customer service, lower customer attrition, better relationships with suppliers and customers, and improved public relations and reputation?

Moss demonstrated so much confidence in the value of such a data strategy team, that she offered a closing rhetorical question, wondering aloud if this prescription might someday upend the traditional BI model: “Could we deliver BI without building data warehouses?”We’ll see about that!

Must Read Articles