Historically, the pendulum of the business intelligence (BI) market has moved between centralized governance of data and self-service and agility. Today, the industry is swaying back and forth, as CIOs struggle to find the right balance of control, transparency, and truth.
Recently, the pendulum has swung too far in the direction of data discovery. While data discovery tools provide speedy data discovery and manipulation, these tools can create analytical silos that hinder the ability of users to make decisions with confidence.
These self-service capabilities create challenges for CIOs, according to Gartner, Inc. Without proper processes and governance in place, self-service tools can introduce multiple versions of the truth, increase errors in analysis and result in inconsistent information.
Is ‘imperfect but fast’ a fair trade-off?
That said, CIOs have come to accept data inconsistency as the price to pay – a tradeoff to achieve speed – to give business users the ability to analyze data without depending on a central BI team. Both parties seem to have adopted the maxim “imperfect but fast is better than perfect but slow.” But is this price too much? The answer is worth delving into.
Backlashes include siloed and inconsistent views of key metrics and data across groups. For instance, lead-to-cash analysis requires data from three different departments (Marketing, Sales, and Finance) and three separate systems (marketing automation, CRM, ERP). A consistent and reliable view of the information between departments and systems – one that provides a common definition of “Lead” or “Revenue” – is necessary to avoid confusion and conflicting decisions.
Finding consistency with data-discovery tools requires the daunting task of manually delivering a truly governed layer of data without a comprehensive understanding of core business logic. This means having the ability to build and test integrated data models, tools for performing extraction, transformation and loading (ETL) routines across corporate systems, channels for proliferating enterprise-wide metadata, and a demand for governance-centric business procedures. All of which are a burden to CIOs.
The end goal: Transparency and speed
Trusted data does not have to be synonymous with restrictive access and long wait times. By implementing transparent governance, CIOs can enable local (decentralized) execution with global (centralized) consistency, reconciling speed with trust at enterprise scale.
But to deliver the agility of data discovery with enterprise governance, CIOs must work to do the following:
- Adopt a data-driven culture. CIOs need to create a team approach to BI that balances the use of skilled resources and the development of more localized business skills to deliver ongoing success.
- Enable data access to all business users. Creating protocols to access new datasets ensures that all business users can identify opportunities that add value. Multiple layers of security during discovery and consumption are crucial to uphold security. With simple and secure access, users can easily identify opportunities to derive insights.
- Create a consistent understanding and interpretation of data. Certify and manage key input datasets and governing information outputs to help align organizational accountability for data discovery. A single view of governed measures and dimensions, for users in both decentralized and centralized use cases, ensures consistency.
Following the aforementioned points to deliver trust and transparency results in big gains. According to Dresner Advisory Services, organizations that view data as a single truth with common rules are nearly 10 times more likely to achieve BI success than organizations with multiple inconsistent sources.
There is a powerful and direct correlation between business success and having a trusted view of enterprise data. Companies evaluating BI solutions must look for modern architectures that support transparent governance at business speed and deliver a unified view of data without sacrificing end-user autonomy. The companies that do will continue to win.
The preceding article was originally published in Gigaom on November 6, 2015.