If you’re over 50 or a rock-and-roll enthusiast, you’ve heard of Boz Scaggs, a 1970s top 40 solo artist and one-time lead vocalist for the Steve Miller Band. Boz Scaggs was not a business intelligence (BI) professional, nor did he ever sing about data or analytics. But his last name is a fitting acronym to describe the primary responsibilities of a successful BI and analytic leader.

So what do BI leaders do? How should they prioritize their duties to have the greatest impact on their BI and analytic programs and organizations? Boz Scaggs—or at least his last name—leads the way:

• Sell BI and analytics
• Coordinate analysts
• Agree on best practices
• Govern reports
• Gather requirements
• Standardize tools and data

S—Sell BI and Analytics

Successful BI leaders are consummate technologists, but even better salespeople. They know how to sell the value of BI and data up, down, and across the organization. They are tireless proselytizers who never miss a chance to explain the benefits of BI and analytics and promote the BI program’s services.

While BI and analyitc leaders present executives with thorough business plans aligned with strategic objectives, they go a step further and create a buzz around BI and analytics. They create catchy logos and posters to draw attention to their services; they publish interesting facts gleaned from corporate data via email and Web; and they host in-house summits that tout internal success stories and spur outside-the-box thinking.

“[Savvy BI leaders] create a buzz around BI and analytics.”

Most importantly, successful BI leaders speak the language of business. They spend time learning the business and always couch technology solutions in business terms. The most successful serve as consultants to business executives and managers, proactively developing technical strategies to address critical issues. Consequently, business managers view them as strategic partners, key players who can help them devise and execute plans to achieve goals for which they’re held accountable.

C—Coordinate Analysts

Second, BI leaders build an extended team that spans the entire organization. Most BI managers think the BI team consists of their direct reports in the corporate office. It does not. A BI team consists of all the business people who in any way, shape, or form manage data, build reports or conduct analyses. In a data-driven organization, this adds up to dozens, if not hundreds, of individuals!

Savvy BI leaders identify these people, especially the report developers and analysts in every business unit and department who use non-standard tools and data to create departmental reports and spreadmarts. Although these individuals often intentionally operate under the radar of the corporate BI program, a successful BI leader seeks their input while providing direct support, turning potential adversaries into allies for improving the delivery of data to the entire organization.

“A successful BI leader turns potential adversaries into allies….”

This extended team of data-centric individuals forms the basis of a BI Center of Excellence (CoE). The BI leader uses the corporate BI team support report creation in the business units, working to standardize definitions, tools, and methods to improve efficiency and sustainability. In turn, analytic managers in these departments allocate portions of their time serving on governance teams to help the BI director create an enterprise vision for BI and analytics and execute it. This quid pro quo is what turns the wheels of a BI CoE.

A—Agree on Best Practices

With an extended BI team in place, the BI leader facilitates an enterprise-wide discussion on best practices for doing BI. The idea is not to impose standards from on high, but to surface the techniques and tools that work in practice and share them broadly. This conversation—which can occur on governance committees, user forums, and quarterly or annual in-house summits—also documents gaps and requirements that the BI team can formalize into business plans and projects, guided by the BI CoE.

Best practices can cover a wide swath of tasks and activities, ranging from project management approaches best suited to BI and analytics development to development methods for designing reports, dashboards, and analytic models. It can also be used to identify roles and relationships, standardize terms, nomenclature, and data processes, and establish benchmarks for availability, response times, and fixes, among other things.

Finally, the CoE—which is managed by a Working Committee of business unit analyst managers—defines marketing and training programs to promote the BI resource and improve the ability of business users to access and analyze data quickly. Budgets and prioritization are left to the Executive Committee, comprised of sponsors and top decision makers.

G—Govern Reports

I have yet to consult for an organization that doesn’t have report sprawl. This is often the aftermath of a poorly thought-out self-service BI strategy. Unfettered freedom without reasonable governance yields chaos that reduces the value of a desired resource.

“Unfettered freedom without reasonable governance yields chaos that reduces the value of a desired resource.”

Successful BI leaders gain control of report sprawl in a variety of ways. If there is a standard corporate BI tool, they spend considerable time identifying the roles and skills of its business users, especially business analysts who are most likely to create spreadmarts. They then carefully assign permissions to ensure only approved people have the ability to create new reports or share them. BI leaders also create highly tailored BI portals, making it easier for users to access pertinent reports.

In non-standard BI environments—those renegade shops that often generate the majority of BI reports in an organization—BI leaders work hard to point BI tools to standard data sets (i.e., the data warehouse) and train departments and individuals about the value of standardizing data and reports. Then, they apply watermarks to certified reports generated through a standard process with standard data by approved individuals. This creates a visible distinction between standard and non-standard BI reports.

“BI leaders apply watermarks to certified reports….to create a visible distinction between standard and non-standard reports.“

Finally, BI leaders create safe zones or sandboxes within the corporate BI environment where business analysts can conduct analyses and mash up corporate and local data. Besides providing an outlet for analysts to create spreadmarts, sandboxes enable analysts to reuse each other’s work, minimizing duplication and inconsistency. More importantly, it enables the BI team to watch and observe what analysts are doing and build commonly used data, metrics, and comparisons in the standard BI environment.

G—Gather Requirements

Although business units hate to admit it, there are some reports and dashboards that the corporate BI is best suited to build. This is certainly true for cross-functional reports and dashboards. But it’s also true for certain types of departmental dashboards. Every department needs a dashboard platform that delivers tailored views to every individual in the department, displaying key performance metrics aligned with core responsibilities. These types of dashboards are not easy to build, but can pay huge dividends: they often replace dozens of reports and give users one place to go to get real-time information about the status of activities and programs under their responsibility.

“Although business units hate to admit it, there are some reports and dashboards that the corporate BI is best suited to build.”

I’ve seen many departments build fragile reporting environments that are notoriously unreliable and vulnerable to outages. Often built by a single individual using Microsoft Office tools and Visual Basic scripts, the environments work great until the developer leaves the company or the organization upgrades its Microsoft platform, breaking critical reporting functionality. Most business unit managers understand the vulnerability of these environments, but would rather swallow that medicine than recruit corporate BI. That has to change. The corporate BI team has to build a reputation for delivering rich functionality quickly.

S—Standardize Data and Tools

Most business units have grown suspicious about corporate BI (or more broadly, corporate IT) because of their heavy-handedness in creating tool standards. A BI CoE replaces heavy-handedness with a grassroots cooperative venture in which individuals work together across departmental boundaries to improve data delivery for everyone.

Although this notion may be idealistic and difficult to achieve in practice, it’s clear that IT can no longer dictate standards, except in the most top-down governed organizations. And, even there, heavy-handedness usually backfires.

BI leaders need to understand the corporate culture in which they operate and adapt their BI strategy accordingly. Where possible, they create consensus around a standard corporate BI tool and data warehouse and support business units who elect to implement it. In most organizations, departments often feel strongly about using a particular BI tool, in which case, the BI team needs to communicate that it will only support the data, not the tool or application that processes it.

Conclusion. Running the BI program like Boz Scaggs takes a lot of talent and effort, and it’s not for the feint of heart. Many have tried, only some succeed. Savvy BI leaders prioritize their efforts, double down in areas they can have the most impact, and leverage each success to take the next step in the journey. (For inspiring success stories, read my latest book, “Secrets of Analytic Leaders: Insights from Information Insiders”).