Recent research shows that approximately 90% of organisations rely on data to drive predictive decision making, yet these same organisations overwhelmingly agree that a backlog of data debt negatively affects their ability to deliver new advanced analytics capabilities. In fact, only 26% believe that the quality and integrity of their data is high enough to support advanced analytics and AI.
By Gary Allemann, MD at Master Data Management
Delivering predictive analytics and AI capabilities requires a strategic plan. Yet, while 95% of businesses have a data strategy in place, the same research finds that only 29% of these businesses have a strategy that is clear, understood, and actionable.
The gap between business and technology must be bridged
A fundamental issue is that, too often, the data strategy is a cookie cutter exercise focused on the latest buzzwords. Or it may be too tactical – focusing on IT or the data analytics team’s immediate technical challenges without alignment to the business.
Actionable data strategies clearly show the link between proposed data management capabilities and the achievement of business goals and objectives. Communicating the link between data and business outcomes is also key to shifting the culture of your business to being data-driven rather than buzzword-driven.
Business goals can effectively be broken down into three broad categories:
* Strategic objectives that are intended to change the business – for example, breaking into new markets or launching a new category of product. In many cases, these objectives will be driven by senior management and align with the vision of the business.
* Mid-level managers are more likely to drive operational objectives – for example, reducing the cost of servicing customers or increasing the effectiveness of marketing campaigns.
* At a tactical level, managers need to maintain or enhance existing IT systems, deal with new compliance and regulatory requirements, and create IT capabilities to keep the business running.
Similarly, data sets can effectively be classified according to how they support these three broad categories: data to minimise risk, data to deliver enhanced insights, and data for the running of the business.
Setting your data strategy up for success
An actionable data strategy should begin with an understanding of the key goals and objectives at a strategic, operational, and tactical level. One can then identify the data sets and capabilities that are critical to delivering on each goal.
As one gathers more information, clear relationships and dependencies between goals will typically emerge, as will a clearer picture of the required data sets and data capabilities. For example, marketing’s objective to enhance customer care may be both driven and enabled by IT’s goal to deliver a true 360-degree customer view. However, the need to protect customer data privacy and comply with PoPIA is a prerequisite for both.
There is an underlying dependency on customer data to deliver each of these goals. The strategy can prioritise customer data knowing that the needs of multiple key stakeholders will be met.
Defining an actionable roadmap
Once one has a clear understanding of where to focus, one can then complete a gap and risk assessment. Our Enterprise Information Framework is a useful tool to understand the information management prerequisites that must be in place to support various initiatives.
For example, to deliver a Customer360 degree view (master data management) we may identify gaps in our data privacy and data quality capabilities that must be plugged to achieve success. Again, filling these capability gaps can be prioritised based on their importance to meet multiple business objectives, or business expectations may need to be managed based on current capabilities.
Obviously, a roadmap may include investments in technology. But we also need to recognise that data management is a people and process problem. The plan should also consider training requirements – for example to drive basic data literacy – and process tweaks to change unwanted behaviour.
Each step can be prioritised as follows:
* Start within three months
* Start within six months
* Postpone to next year.
Inherent in this approach is the assumption that management will revisit the strategy at least once a year, and possibly more frequently, to track progress and to realign to shifting business priorities. Some longer-term activities may become more urgent, or they may be dropped altogether.
Keeping your data strategy relevant is key to paying off your data debt and shifting your data from a liability to an asset.