When building Supplier InfoNet, SAP's stealth development initiative that launched the application did not take advantage of much off-the-shelf SAP technology (outside of BI) such as the classification engine in Spend Performance Management. Rather, they built their own solutions from scratch -- as an example, in the case of classifications, InfoNet leverages machine learning technology and adaptive matching as part of a content factory that processes the inbound customer files to be loaded into the application. This matching system learns preferences much like other classification tools, but allows greater flexibility than many solutions on the spend analysis side. For example, a company might not want to treat "Marriot" as a single entity, but rather might care to look at individual franchises. Another likely example is Caterpillar, which could be represented as a group of individual dealers. In addition, the system can come up with suggested matches and linkages based on patterns it recognizes and learns.
With InfoNet, SAP has to address a major challenge in how to best normalize KPIs and systems data across companies. David Charpie, VP and Global Business Incubator at SAP shared with Spend Matters that their philosophy is "not to solve matching or KPI definitions at a corporate basis." Rather, he observes, "we believe the right answer is to step back in looking at the problem by letting each individual (or organization) define what they want a KPI to be." For example, different users and companies should be able to define a KPI for on-time delivery as an item that arrives within a one-hour window or a one-day window (or longer). Currently, InfoNet addresses this issue at the organization level by working with customers during setup to understand their definition of the KPI in order to accurately portray their corporate view of the suppliers into InfoNet. By leveraging SAP's in-memory technology, called HANA, Supplier InfoNet intends to fully realize this vision in a future release, letting individuals pick and choose their own KPI definitions (and modify them as desired) to allow them to personalize the performance results and benchmarks contained in the system to their own needs and requirements. This is critical because even within the same company, different facilities, business units and managers may hold suppliers to different standards based on specific contractual arrangements and programs (e.g., JIT vs. non-JIT).
A few curious asides are worth mentioning regarding InfoNet. First, it raises some interesting legal questions around the sharing of information across entities without a supplier's consent. To make sure they do not run afoul of any privacy laws, the first thing SAP does is strip out information on individuals (e.g., sole proprietors) from the system. Secondly, SAP's machine learning analyses assess the value and integrity of each of its data sources to determine their ability to improve the accuracy of the predictions. The idea is to provide greater weight to those sources that are accurate and consistent indicators of future performance. This is then factored into how it rates, ranks and provides alerts around suppliers. The additional information this could potentially provide is staggering -- wouldn't you want to know who among your suppliers are consistently dishonest/ignorant? This could also have significant impact on a company's training allocation since it should be easier to find out who actually needs additional training. That is food for later though: back to supply chain.
Where does InfoNet change how SAP positions itself as a company? Perhaps with InfoNet, the best run businesses do not just run SAP (which is a catchy but questionable claim, regardless). They will now rely on SAP's networked business information to power their operational business decision-making and strategy setting. If this sounds like SAP is becoming a company that will have an operating unit (and value proposition) that looks closer to D&B and Bloomberg combined, that's exactly the direction InfoNet is heading. However, SAP shared with Spend Matter that they are not thinking of themselves as a "content business." The team wants to see the InfoNet functionality embedded in regular SAP solutions as a core value proposition in the future (i.e., sign up for SAP ERP in a manufacturing environment and you can opt into the SAP network for aggregated supplier, and maybe even customer, performance).
Stay tuned as we dive into how InfoNet works in practice, walking through what we observed during a live demonstration of the solution.
- Jason Busch and Thomas Kase