With more tools emerging — like predictive analytics, big data and probabilistic modeling — the future looks bright for next-generation sales and operations planning (S&OP).
The dilemma in S&OP is this: With Gartner’s four-stage maturity model, over the last 20 years, almost 70 percent of all companies have been operating in the first two stages — Reacting and Anticipating. Why? It’s not easy to start an S&OP process and even more difficult to sustain it. The first two stages tend to be geared around getting a plan together and maintaining a regular meeting to balance supply with demand for the good of the enterprise.
These two stages are totally focused on inward processes and tend to take at least four years to solidify. Then, the big jump to Stage 3: Collaborating. This means expanding the process to your suppliers and customers. The focus changes to profitability, and in Stage 4: Orchestrating, it's driven by demand sensing, shaping, scenario planning, “what-if’s” and enterprise trade-offs, including risk/reward analyses.
The classic deterministic models we know and love in supply chain do not handle uncertainty. However, there are such new tools as predictive analytics, big data and probabilistic modeling. Others, including Python, R, Hadoop, Spark and more, are emerging and embrace uncertainty and can effectively read and make sense of new unstructured data coming from the internet. Ten years ago, 80 percent of all our data inside our supply chains was structured. Now, it’s just the opposite.
To be successful in stages 3 and 4, we feel S&OP professionals will be better served utilizing these new tools to digitize their supply networks, run effective “what-if” scenarios to quantify how their supply chains will react to stimuli from both within the organization and outside, while evaluating unstructured data from the internet to develop statistically strong patterns associated with suppliers and customer buying habits and sentiments.
As more and more supply chain and S&OP professionals leverage these new tools, we feel confident that many more companies will successfully make the jump to “outside-in” Stages 3 and 4. We expect companies will move up the predictive analytics maturity model from “descriptive” (what happened), through “predictive” (what might happen), into “prescriptive” (what should I do about it) and finally to “cognitive” (the system learns), meaning, the system begins to assist in the complex decision-making process.
With that said, we feel the complexion of the S&OP process will change dramatically from a very structured, linear process, to a more ad-hoc, event-driven environment. This may involve more high-frequency/high-impact decision-making. We’ve witnessed these new tools accelerating the learnings across the enterprise and ultimately enhancing the precision and effectiveness of decisions in complex supply chains.
IBM is a prime example of an exemplar company leveraging these new tools. Watson, their cognitive computing engine, evaluates IBM’s global supply chain network, 24/7. It identifies potential disruptions, assesses and provides risk mitigation options to IBM professionals’ laptops, iPads and phones with extensive assumptions, corporate risk appetite rules, risk/reward trade-offs and more. These alerts, with options to maximize performance and mitigate risks, await IBM pros as they sip their first cup of coffee.
Gregory Schlegel is founder of The Supply Chain Risk Management Consortium and executive-in-residence at Lehigh University's Center for Supply Chain Research.
Timely, incisive articles delivered directly to your inbox.