The information technology analyst firm Forrester Research recently published a report that described the gap between data governance and business process management (titled Avoid Process Data Headaches: Align Business Process and Data Governance Initiatives, it's available here; see also this InformationWeek article). Forrester sees a wide disconnect between data quality and process improvement initiatives, exposing both efforts to the risk of potential failure.
This stimulated me to think more deeply about the data quality issue for enterprise performance management.
In my March 18 blog “Cut Through the Confusion to Unleash the Full Power of Enterprise Performance Management” I described enterprise performance management as an umbrella concept that broadly integrates operational and financial information into a single decision-support and planning framework. Its components include strategy mapping, a strategic balanced scorecard, operational dashboards, costing (including activity-based cost management), budgeting, forecasting, and resource capacity requirements planning.
I also noted that these methodologies fuel other core solutions such as customer relationship management (CRM), supply chain management (SCM), risk management, business process management, and human capital management systems, as well as lean management and Six Sigma quality initiatives.
The Forrester Research report focused my attention on the front end of these integrated systems. Often referred to as the “pipe,” this is the point where raw transactional data (e.g., from ERP and CRM systems) and master data management (MDM) records become inputs to the steering and controlling enterprise performance management systems that guide an organization's economic growth.
Most departments and functions in an organization are blind to the initial input data from which insights, analysis and decisions are derived. And that raises some nagging questions: Who owns this foundational data? Is it the CIO and the information technology department? Or is it the business users?
I'm unsure how to stop the finger-pointing that's so common between these two groups, but I am sure that collaboration between these two typically siloed functions will be essential to solving the problem that Forrester describes. One can make a case that IT contributes to the problem with poor data models and inadequate application design and development practices that are further complicated by inadequate testing. But business users are not free of responsibility and accountability for the quality of EPM's input data.
I have to confess that I myself have turned a blind eye to this problem. In my writings I presume that the input data is of high quality and discuss how and why organizations should convert data into usable information and deploy it for performance improvement.
The phrase “garbage-in garbage-out” overstates the problem. But to the degree that input data is impure, surely it will impede the dream goal — enterprise optimization.