BI and analytics have evolved—so should governance
Modern BI has made self-service analysis possible and organizations have realized incredible benefits democratizing data and crowdsourcing new insights. Yet somehow the idea of self-service governance still seems inconceivable.
Governance is necessary in modern analytics because data and dashboards are more widely shared, but should enable data and content access rather than restrict it. These practices and processes help the right people access the right data, ensure that the data driving your users’ decisions is accurate, and maintain compliance with internal policies or external regulations.
To achieve both the peace of mind IT requires and the agility that business users demand, your approach to governance must shift to a more collaborative model. Success means navigating the roles, responsibilities, and processes that surround this shared responsibility and working to improve them with an agile approach, through iterations.
A governance steering committee and framework
While modern BI technology can help to hide the complexity of your data architecture and make certain tasks easier for the business to manage, both IT and the business need to buy in and participate in establishing and maintaining governance. A steering committee is a great way to ensure the bases are covered, bringing together key players to establish a clear vision and framework.
This group should define the “rules of the road” and set expectations and processes around user roles, privileges and permissions, training and certification, as well as new metrics to measure success. Consider that different kinds of data will require different kinds of governance, so your model must be flexible enough to accommodate any kind of data. Think through the roles and processes needed to support data and content governance as your model adapts to changing business needs.
Working toward a crowdsourced governance model with modern analytics
Modern BI environments are implemented and scaled to benefit analysts and business users and as such, these users should be responsible for its overall quality. IT’s role in the analytics pipeline has evolved from producer to enabler—and so should IT empower the business to have a greater stake in governance. A shift in governance is not asking IT to relinquish control so much as allowing the business to be more self-reliant within a trusted and centralized environment. Analysts and business users become the first line of defense in identifying data issues or irregularities within a governance model that IT and the business agree upon together.
Where the business may lead in defining analytical goals and desired outcomes (the “what”), IT is critical in establishing the processes that enable those outcomes (the “how”), including ensuring data integrity and security as analytics scale across the organization. Remember that it’s not an overnight transformation; it should be incremental. Often, this means starting with a more traditional top-down approach and moving toward a self-service model over time, delegating responsibilities to the right business users with the appropriate knowledge, training, and understanding of governance policies.
Training and triage — IT may give up some control by making governance a collaborative effort, but they should also share the responsibility to remediate analytics issues. Business participation should also include fixing issues when and where it's appropriate. Education and training is critical to mitigate the risks of governance and resistance of change management when deploying your modern analytics across your organization, and it helps get the business invested in participating in governance. Then, if anything goes sideways, IT can triage certain tasks and the proper business users are empowered to be a part of the process to a solution. So, IT should establish a foundation and train on everything related to governance—they know exactly how the model should be. Conversely, the business should establish a COE to train on analytical skills, best practices, and resources.
New data sources and content — Start with IT managing data and content creation. A successful content organizational framework will make it easy to discern trusted data from experimental content—for example, certifying data sources versus connecting to external data for ad-hoc analysis within sandbox projects. The next step in self-service may mean IT still owns the data but a small group of users can manage publishing new content, moving content from a sandbox to a production project and promoting it. With clear standards and a checklist for publishing new content or certifying data sources, you can manage a scalable model with business participation.
Administrative duties — Certain administrative duties should be delegated to business users as your modern analytics practice grows. IT should always manage security, authorizations, group policies, etc., but project leaders and site administrators on the business side can take ownership of adding new users, managing permissions, and monitoring user and content engagement. Typically, this is arranged by department. Admins on the business side can leverage an Active Directory or other authentication that IT integrated with the modern platform to make permissions management simpler and remain in compliance.
Lifecycle management — Content engagement is a sign of success for your modern analytics deployment, and often a business KPI. IT should empower the business to monitor engagement with dashboards and reports within their departments, identify content or fields that haven’t been used or accessed, and even perform an impact analysis to assess how changes may affect users. This gives the business more ownership and accountability for the analytics content they are creating and lets IT focus on the technology platform’s security, performance, and capacity.
Finding harmony with a modern approach to governance
If your organization is in “Excel hell” with tons of spreadsheets that leave your data unsecured during downstream analysis, you should take a look at your analytics and governance practices. Swinging the pendulum to the other extreme, if you’ve moved from traditional to modern BI, your organization could benefit from a closer look at your analytics audience, who is using which reports and why, and who might benefit from less stringent data access.
It may be easier to find your ideal governance framework and work toward a modern, self-service model if you think of governance as a spectrum. Modern BI affords access to any place on this spectrum, but it’s up to IT and the business to work together to move along it iteratively and incrementally as it’s mutually beneficial.
There’s a place for traditional, top-down governance and there’s certainly a place for a less restrictive, bottom-up approach. But there’s no either/or, and they must converge; a modern approach is all about finding the right balance and being open to that balance changing over time. The more responsive you can be with roles, controls, processes, and responsibilities, the more business value you can drive. Do you know where you are on the spectrum—and more importantly, where you’re headed?
Read more about Governed Self-Service Analytics at Scale with our whitepaper.
Learn more about evaluating, deploying, and governing modern BI and analytics with confidence through our webcast series, Embracing the Modern BI Evolution.