It shouldn’t be a big shock to learn that procurement analytics is a big deal right now. After procurement organizations have built some basic spend cubes (or “spent cubes”) and dashboards, they’re looking for deeper predictive insights into spend, contracts, suppliers, costs, process improvements, supply risk and other areas. In fact, analytics was by far the most cited technology area expected to have a business impact within the next two years by CPOs surveyed in the recent 2019 Deloitte Global CPO Survey.
The biggest area of interest within analytics have been:
* Self-service analytics/visualization for business stakeholders and procurement staff
* Predictive analytics for power users (e.g., for price/cost/volume forecasting)
* Performance analytics and dashboards (e.g., supplier scorecarding, category dashboards, etc.)
* Support for digital initiatives such as AI/machine learning (which is usually about focused predictive analytics problems), RPA (that either requires some analysis within a process or conversely is about helping to automate the analytic workflows), or big data analytics (e.g., using IoT sensor data from the supply chain)
The Predicaments
However, while analytics are hot, the implementation barriers can be stone cold killers:
* Poor data quality. 40% of CPOs cited the inability to generate insights and analytics because an even greater number (60%) cited poor master data quality, standardization, and governance.
* The master data quality problem is very familiar to practitioners who run any type of analytics that have to do with suppliers, items and contracts — i.e., most of them!
* Some ERP suites and procurement suites have fragmented master data within their product lines, and nearly all these solutions don’t have master data that can be used as part of an MDM-type solution (e.g., having a supplier master that can serve a true SIM solution from an MDM standpoint rather than just creating another vendor master file to add to the heap).
* Generating forward-looking insights based on external data and intelligence rather than just simple spend forensics — especially category-specific insights that are typically built from scratch.
* The struggle to create analytics that go beyond off-the-shelf operational reports from the various modules/tools in the market.
* Dashboards that are attractive, but can be visually overwhelming and not help you prioritize where the key opportunities are.
* IT organizations that may be pushing legacy data warehouses and BI tools that don’t allow more democratized analytics to be developed with an increasingly digitally savvy generation of business users and tools (that might also need to get adopted by an older generation of procurement practitioners). Data visualization and predictive analytics were the top two digital skills prioritized for procurement technology training over the next year.
In the rest of this Spend Matters PRO brief, we’ll dive into the current and future state of the procurement analytics area, and make some predictions about what we expect to see in 2020 from a market standpoint, but also a more detailed technical standpoint.
