IBP or time for AI-BP?Advancing S&OP with AI: The Path to Integrated Business Planning 2.0
S&OP may be a 40 year old concept, but it is developing at an increasing pace, enabled by the possibilities offered by the latest generation of Advanced Planning Systems.
Laggards and early adopters: mind the gap!
Looking at our customer portfolio, the S&OP maturity gap is striking. On the laggard side, we find engineering companies like machine builders, warehouse automation suppliers and other assemble and engineer-to-order companies. Many of these are taking the first cautious steps into the scary world of S&OP, fighting the classical battle for buy-in with the commercial functions.
On the innovators side we find the FMCG multi-nationals, that left this battle behind ages ago. These companies are fully engaged in exploring the world of Machine Learning, AI and data lakes, utilising internal and external data, combined with the limitless power of cloud computing. This is where it gets interesting, so let’s see what this group is up to!
IBP 1.0: one set of numbers
The main idea behind Integrated Business Planning is to drive the business based on “one set of numbers”. The financial plans in this case are based on the bottom-up volume plans that are managed by supply chain and translated into Euros. IBP shifts the focus from demand-supply balancing to managing the gap between target and financial outlook and replaces the traditional (financial) Monthly Management Reviews.
Volume-value conversion is done at topline as a minimum, but often also goes down to gross margin or sometimes even EBIT. Nothing new so far, although calculating the toplines can be challenging due to the complex and country-specific trade conditions that are typical in FMCG.
AI forecasting offers new possibilities for creating more accurate forecasts with less effort, so is highly relevant for IBP. Driver based forecasting exploiting internal and external sales drivers, automated decomposition and segmentation and many other developments will help to create more data driven forecasting with less human bias and higher accuracy. This topic deserves more attention than a single paragraph, so we will come back to this in a separate blog.
IBP 2.0: decision support by financialised e2e scenarios
IBP can be much more than managing the gap. An area that is often overlooked is the use of financialised scenarios, ideally using end-to-end metrics, that can be created with a few mouse clicks. Two main areas where the quality of decision making will benefit:
- Tactical decisions, taken in the IBP process.
- Operational decisions, taken in S&OE.
The paradigm in FMCG, specifically A-brand manufacturers, is to deliver the market at any cost. Demand is leading, Supply Chain and Operations should follow and accommodate by pushing the boundaries whenever necessary.
This will (and probably should) not change, but end-to-end financial scenario analysis can still offer big benefits, as the trade-offs that today may be made implicitly will become explicit. Does it really make sense to push sales volumes via short-term promotions, to reach the quarter end revenue target? Or should we accept a revenue drop and activate option B, that has a higher EBIT as it avoids the additional costs? In S&OE, do we plan weekend shifts, pre-produce or hire temps to support the higher order intake of our latest innovation? Not by looking at standard costs, but by evaluating the impact on variable costs that we have simulated in our APS.
So what’s different?
In theory, most of the scenarios could be built in Excel, following ad hoc data collection, manual analysis and all sorts of heuristics. In practice, this will only be done for the big tickets, which means that the vast majority of trade-offs is made based on gut feel and in functional silos. Pity, as this way of working is unlikely to give optimal financial outcomes.
AI/ML can already have a big impact on the efficiency and quality of forecasting and inventory management. The improvements in IBP decision support are driven more still by the combination of concurrent planning (combined demand and supply databases), digital twins including detailed financial models and the limitless power of cloud computing. This combination enables the push button and end-to-end scenario creation and analysis that is so useful for IBP. Conceptually probably not even so very different from your manual efforts, but a totally different way of working in practice!
Why bother?
Many things can be automated, but making non-routine cross-functional trade-offs should not. The quality of decision making can be improved significantly though by offering automated decision support based on volume and financial data and metrics. This will enable a shift from manually finding feasible solutions to automatically creating options that optimise financial KPIs like margin and EBIT. Since we want data informed and not data driven decisions, the final cross-functional trade-offs should always be made in the S&OE and IBP processes, by people that really understand the business and all its complexities!
