Digital transformation initiatives are starting to show some real business results.
Take this oil and gas case study for example, where smart connected operations enabled the company to rapidly drive insights from massive datasets and automate or simplify many processes and tasks in exploration and production activities. The retail industry is already seeing changes in the product development lifecycle in order to survive in a digital world. When it comes to initiatives around AI, machine learning or IoT, the potential is enormous.
But there is one common theme that spans across industries: for all digital efforts, data is the underlying raw material. And it’s time to get your data under control.
Justifying a Strategic Approach to Data Management
According to John Mancini, a former president for the Association for Intelligent Information Management, one of the reasons that both digital transformation and big data initiatives fail is that data is simply out of control. Organizations must take steps to tame their data to succeed. The data involved – and its metadata – must be clear, auditable, trustable and easy to find throughout its entire lifecycle.
To do this, enterprise information management (EIM) and digital strategies must be carefully aligned. An ad-hoc approach to EIM – instead of constantly measuring effectiveness against a framework – will stall momentum of the EIM program and everything else that depends upon it. Even taking the first steps towards governing and setting up a certification process for data sets or reports will increase the probability that your digital initiative will succeed.
A recent report from Forrester found that this is part of the reason why digital transformation initiatives are a bigger challenge than just finding a vendor and buying a technology or using some open source libraries. It is more of a cultural and organizational problem. This assessment might sound familiar, since this is a recurring theme in the EIM space. According to an EIM Readiness study from 2014, implementing an enterprise-wide EIM strategy is non-trivial and most organizations are struggling to achieve buy-in from departments and to enforce enterprise-wide policies and standards. Anne Smith, from the Enterprise Information Management Institute cites cultural barriers as the first reason why data governance programs fail.
Depending on what you are trying to achieve, your EIM strategy will vary. For an enterprise struggling to speak the same vocabulary and arguing over semantics, it’s a priority to build a business glossary. For those that still do not have self-service analytics capabilities because the data lineage is still waiting to be untangled, the priority is a metadata repository of physical data assets. This will clarify where data comes from and the hops it makes along the way.
A common approach to kick-start data governance is to identify KDEs – also known as CDEs for key or critical data elements – in order to prioritize the data elements that need to be crystal clear for the business to boost revenue, improve product quality or ensure compliance to internal/external regulations.
The Bottom Line
Just as it would be foolish to expect something you cooked with suspicious-looking ingredients to taste amazing, a digital initiative built with suspicious data is risky at its best. As tempting as it may be to dive into a new IoT or analytics initiative, consider EIM improvements in parallel to ensure visibility into data quality and governance metrics.