Thinking About ERP In An Ever-Changing World Of Data
By Anthony J. Scriffignano, SVP & Chief Data Scientist, Dun & Bradstreet
The Foundations: Structure
ERP is largely about structuring data and connecting it according to strict standards. Studies have shown that getting the data right is among the top areas of focus that CIOs and other leaders in ERP implementations would have emphasized even more if given the chance. After all, an ERP is not simply a massive database connected with related applications. It is a complex network of data and transactions, which employs internal validations, audits, alerts, and other modalities to keep everything running smoothly.
All of this reliance on structure makes it hard to ignore how the very nature of data is changing. Experts generally estimate that, while the amount of data on Earth is growing at an unprecedented (and arguably unmeasurable) rate, most of the new data is “unstructured,” which is often a misnomer for data that has not been organized into a useful ontology. This data is constantly growing and rich in semantic context (such as online review comments, user stories, narratives, and many other forms of text). Some estimates put the amount of unstructured data creation at nearly 85 percent of all new data.
At the same time, the nature of “new data” is also changing drastically, as IoT continues to grow. Devices are producing an ever-increasing mass of highly structured, but widely varied data. Much of this data is used in very limited ways, to drive specific processes or in a highly constrained domain. As devices begin to develop the ability to discover and interrogate one another, the ubiquity of IoT data will increase dramatically (as will the opportunity cost of ignoring this rich source of data for other purposes such as signal analysis, deep learning, and other very valuable possibilities).
A very powerful frame for considering change in an organization considers four factors at the outset: people, process, technology, and mindset
While this sort of data will be much easier to integrate into ERP environments due to inherent structure and the likelihood that it already will be product-or process-related, the sheer volume cannot be ignored.
Changing Environment: Risk and Opportunity
So far, the changes described relate to the basic “Vs” of Big Data (volume, variety, velocity of change, etc.), but there are also very important changes taking place in the environment. Many of these changes relate to regulation around the discovery and curation of data. For example, privacy laws (which typically focus on personal information), data sovereignty regulations (which focus on where data may be created and stored), and data governance regulation (which focuses on how information must be maintained) continue to emerge globally. Many of these laws require organizations to rethink much of their infrastructure and process. In ERP-enabled organizations, especially those with cross-border process through either single-or multiple-instance strategies, all of these regulatory changes must be carefully considered.
The only certainty about the changing face of data is that it will continue to change. How decision makers in organizations react will dictate whether they fall into a posture of defending the environment against all of the change surrounding it or a posture of taking full advantage of the new opportunities.
Leading Change: Where to Focus
A very powerful frame for considering change in an organization considers four factors at the outset: people, process, technology, and mindset.
From a people perspective, we must consider that the skills that have made us successful so far are necessary, but not sufficient. Ongoing focus on new skills, such as working with unstructured data, using non-regressive methods (e.g. AI or heuristics) to derive value from never-before-seen data is crucial. Equally important is the ability to recognize new types of information that are becoming available to the enterprise, thereby avoiding silos of information that otherwise might be used only for a single purpose.
ERP is all about process, but how must that process be re-envisioned in the context of the continuous and unprecedented data-related change? Clearly, processes must become more learning-based, and to some extent more self-correcting. For example, while closed-loop manufacturing processes have been around for decades, we now have the ability to augment these processes with signal analysis, intrusion, and anomaly detection based on far more advanced capabilities that can learn from the ongoing operation. Such approaches are being integrated into new technologies, such as autonomous self-driving vehicles and drones. Integrating the same sort of intelligence into more traditional processes, especially in light of the far richer troves of available data, could provide dramatic benefits to the ERP-enabled organization.
From a technology standpoint, while the IoT is an obvious focus for ERP-enabled organizations, it is not the only area where the pace of technology change gives rise to both risk and opportunity. Of particular focus lately are the many forms of AI, including deep learning and cognitive approaches to problem formulation. As organizations put focus on these emerging fields, we will certainly see both use and misuse of technology.
Finally, we should consider mindset. How are we reacting as leaders in an increasingly data-driven enterprise? Einstein is often quoted as saying that we cannot solve a problem with the same mindset we had when we created that problem. The shift from MRP to ERP was about connecting everything. It seems we are still on that journey, but now there is so much more to connect and amazing potential if we get it right.