Data - volumes of information in mass scale that can be gathered on nearly all areas. Data that can miraculously offer you insights and deeper understanding of our own market and product…
This is the rosy-hued mirage that big data offers companies.
Not surprisingly, organizations started implementing a mix of several business intelligence tools and systems. Added to this, enterprises started using complex IT infrastructure leading to data discrepancies, miscommunication, organizational challenges and a bulk of unmanageable data.
Why did this happen? Organizations are in a constant state of flux. Which means aggregating and assimilating information in a sustainable manner is difficult. The existing BI governance adopted by companies is either business driven and highly decentralized or IT driven and highly centralized. While one is agile but costly, the other is economical but inconsistent. Ironically the weakness of one model is the strength of the other. The need for a proper governance structure that involves all parties has never been greater. The answer lies in implementing a federated BI governance model.The federated model removes the unmanageable complexity between functional business and non-functional IT requirements.
Dawn of the Federated Analytical Culture
Here, a governed analytical culture is built stage-by-stage where organizations can use the right tools and analytics to enable rapid decisions. Traditional BI is is complemented by agile BI. For reporting, columnar databases are used and a hierarchy for MAD (Monitor, Analyze or Detail) view is created.
The idea behind this model is to create a standard, self-reliant, federated BI architecture with minimal opportunity costs. It is both business and IT driven, agile and managed by a federated BI team. Simply put, federated governance is that structured balance between a tightly controlled centralized BI and a very abstruse, decentralized BI. Federation takes the middle ground between centralized and decentralized BI approaches. Ideally, it provides the best of both worlds while minimizing the downsides.
The Three Stages
Building a governed analytical culture is done in three stages. The first stage follows the Kaizen principle - an approach to work that systematically seeks to achieve small wins through which business efficiency is improved. A candidate business group that needs dashboards to solve a business problem is identified. A team including a Subject Matter Expert, a data steward and a Dashboard Artisan is formed. They are empowered to use rapid visual analytics. Dashboard consistency is achieved using the MAD methodology which we discussed earlier. This process is reviewed for usability and the feedback is used to refine the dashboards.
Platform-centric Approach to Governance
The second stage is strategic expansion - here organizations can develop a steering committee, working group and adopt evangelism combined with best practices, content promotions and user experience. Under the strategic expansion stage, taking a platform-centric approach to governance can help enterprises to seamlessly facilitate automations, authentications, and monitoring. It is becomes easy to create audit and usage reports and fine tune the performance.
The Governance Pendulum
The final stage is the enterprise rollout. Here, the traditional sandwich approach (combining top-down and bottom-up approaches) is established and augmented by the federated, agile governance model. Once the enterprise rollout stage is reached, you can limit the IT headcount to the minimum while maintaining the executive steering and working committees who will be in charge of risk management, program direction and policy implementation. The center of operations will manage information assets, supports operations and users.
At this stage, governance becomes a balance between centralized BI and decentralized BI. You have:
- A federated BI team
- A standard architecture
- A model that is business & IT driven
- Cost efficiency
- A self-reliant and agile data culture