Modern business intelligence has not proven the value of data as much as the value of people when they’re empowered by data. Your employees are smart, curious, hard-working, and know the questions they need to ask and answer to drive the business forward. With the right insights, you can witness an immensely valuable transformation.
Everyone seems to grasp the role of modern BI in capitalizing on the value of data, but many organizations struggle to get data-driven decision-making to spread enterprise-wide. Could it be because they’ve forgotten that this transformation is not really about data or technology, but their people?
Why does IT need agile for modern analytics deployment?
To be successful, modern analytics and business intelligence can’t be considered a destination with new capabilities and processes as checkpoints to success; instead, it must be seen as an ongoing journey, centered around people. For this reason, to deploy, support, and scale an IT-enabled, people-driven modern analytics solution, we must abandon not only traditional technologies, but traditional methods of integrating them into the fabric of the business. While many organizations recreate their existing reporting workflows in a modern container, this leaves a great deal of opportunity beyond the technology where agile principles can be applied to realize the transformational benefits of modern BI.
Moving from top-down, IT-led business intelligence toward this modern self-service analytics model feels like night and day to IT. Everything changes — from business requirements and technology platforms to IT’s role in the analytics pipeline — and this new world brings a lot of fear. As guardians of the organization’s data integrity, IT is wary of an ungoverned, wild, wild west, wherein unfettered data access and rogue content risk creating confusion or contradictions.
Ultimately, this fear comes from lack of understanding and delegating responsibilities. While the risks and resistance of change management can be mitigated with education, planning, and responsiveness, it’s critical to understand how to apply these processes in an agile approach to modern analytics. This means placing people over process, throwing away all-or-nothing milestones, and embracing incremental improvements to deployment plans and governance.
It also means establishing new criteria for success and business outcomes that allow IT and the business to move from “traditional” to “modern” in iterations as its mutually beneficial. Then IT can be more responsive to the needs of the business and more proactive in supporting and scaling the necessary infrastructure.
How do you get from vision to ubiquity with agile analytics?
First, it’s important to have a clearly defined vision. This doesn’t mean a formalized plan for achieving long-term goals, but rather a sound understanding of the immediate business case, the measurements for success and cadence to evaluate them, and the roles and responsibilities involved in the modern analytics workflow.
There’s nothing wrong with starting small. Successful deployments often start with a single department or a use case within multiple departments. Key data sources can help you estimate the relevant audience size for your deployment because user engagement will drive server scalability and sizing decisions. Server scalability and sizing, in turn, inform hardware and licensing requirements, which align with budget planning and procurement requests.
IT retains the setup tasks that enable the business, including software installation, user provisioning, access rights, governance oversight, and some development tasks related to content and data sources. Business users may fall into different roles depending on their degrees of skill and interactivity with the platform. Some may need to perform data preparation, analytical exploration, and content creation; others may be suited for simpler interactions with visualizations and only needs to consume curated dashboards and reports.
IT should delegate access and responsibilities over time to data stewards who are familiar with the data, governance processes, and business needs to be trusted to connect to new data or publish and certify metadata models for other business users. Authoring capabilities based on existing certified data sources can be delegated to business users to create new content or answer ad-hoc questions. This may also apply to how users onboard, train, receive support, and foster a data-driven community across the organization.
Your planning for physical (or virtual) infrastructure should be just as flexible and iterative. Given that analytics are often mission-critical and modern BI solutions often see fast growth, you should consider reassessing server utilization and user needs more frequently than with other technology solutions. You may need to change your topology to scale more frequently than other enterprise platforms you’ve managed.
Proactive planning and monitoring helps you better prepare, support, and scale. A “set it and forget it” deployment can be met with inadequate resources that fail to support the workload of highly-engaged users. Similarly, you shouldn’t wait for a spike of performance issues or support tickets to address possible expansions or implementing new technology.
Deployed with agile methodologies, modern business intelligence grants as little or as much change as the organization decides it’s ready for. With a defined vision, careful planning, monitoring, and measurement, agile analytics deployment helps an organization navigate change management and see more stable growth toward data-driven enterprise transformation.
Josh Parenteau leads the market and competitive intelligence team at Tableau with over 20 years of experience in the BI and analytics industry, including previous work at Gartner.