The customer (we'll call them "JKL", since they do not allow their name to be used in vendor materials), is a top Fortune 500 company with billions in spend and tens of thousands of vendors.
JKL generates procurement data from AP, PO, and Pcard systems, along with line item detail for temporary labor, marketing, telecom, print, travel, and other verticals. An aggressive M&A strategy means that acquired companies with incompatible systems are often in the process of being integrated into JKL
Before Spendata's involvement, JKL's business analysts were supported by an IT-initiated program to consolidate data in a "data warehouse" so that it could be accessed and analyzed. Tableau and SQL were the primary tools used to access the warehouse. These tools proved to be inadequate, creating an over-reliance on IT personnel, delays in updating the taxonomy mapping, and results of limited value. Data from acquisitions was not incorporated until they were transitioned to the main AP system, so the data in the data warehouse was incomplete. Users did not do analysis using the tools provided by the data warehouse, instead just filtering on the desired vendors and extracting the data into Excel for analysis.
Using Spendata, JKL created an AP spend cube combining the AP data with PCard data. Spendata's RER (Reverse Engineering of Rules) function was used to translate existing SQL-based mapping rules into Spendata rules. This replicated the data in the data warehouse while incorporating additional mapping and an improved and faster update process. The percent of spend mapped with stakeholder feedback was increased from a much lower figure to 99+% The data warehouse was reconfigured to pull curated and accurate information from Spendata rather than raw data from source systems, making previously-unreliable Tableau reports useful. A PO cube and category cubes were then created, inheriting automatically from the enterprise AP spend cube and providing category-specific data to the associated stakeholders.
The main AP cube was then expanded to incorporate data from the additional, non-integrated AP systems. This now gives the organization full visibility to the total spend of the organization, speeding up the benefits of post merger integration. These newly acquired organizations now immediately receive the benefits of the pricing from the entire organization.
Additional analysis cubes were created to integrate contract headers (for on- and off-contract analyses), risk status, and Scope 3 estimation. Individual analysts now routinely build their own analyses using custom cubes that are derived from enterprise or category-specific cubes. Core analysis refreshes are performed by a central team, with the results automatically inherited through to category cubes and to individual analyst cubes. Multiple category cubes are currently maintained by JKL, all of them automatically refreshed as the core spend cube is refreshed.
All Spendata work is performed by company employees, behind company firewalls. IT involvement, or involvement of the central data team, is no longer required by business analysts.
The monthly refresh process is automated using Spendata's API, then manually reviewed by the core team, a process taking no more than a few hours. The API allows any of Spendata's functions to be accessed, including cube creation and mapping.