ToolsGroup-Statistics-based Supply Chain Planning

ToolsGroup continues to carve out a space for itself in supply chain. Not satisfied with being the last remaining independent multi-echelon inventory optimization (MEIO) vendor, ToolsGroup is pushing forward with broader supply chain capabilities centered around a core statistics-focused demand forecasting capability. While many will argue the virtues of “top down” vs. “middle out” vs. “bottom up” forecasting (or a mix of these approaches), ToolsGroup has maintained a focus on bottom-up forecasting, particularly suited for fast, slow, erratic, and intermittent demand. (For a complete run-down on ToolsGroup, you cannot do much better than P.J. Jakovljevic’s piece from 2012, “ToolsGroup—Going Back to Its SCP Roots.”)

ToolsGroup, now with 250+ customers, continues to focus on supply chain planning for companies and industries with, usually, one of the following challenges:

  • A complex mix of fast and slow demand. Aftermarket parts, service parts. Consumer goods, where replenishment at a much more granular level means that demand variability is magnified. Consumer electronics firms, where relatively short product life cycles, and new product launch focus, and life cycle management influence the demand cycles.

  • Slow, erratic, or intermittent demand. Large product mixes and slower moving products, and/or geographically distributed production plants making higher service levels more challenging. Examples: consumer durables, consumer electronics, wholesale distribution.

  • Seasonal demand, particularly attribute-based demand. Example: apparel and footwear. Most fashion businesses combine rapidly changing seasonal collections with longer-lasting continuous products. This means managing at least two different parallel demand streams and inventory processes, each with its own business logic. Fashion businesses must also deal with different styles, colors, and sizes, and distribution networks that can scale to up to millions of stock-keeping units (SKU) locations.

  • Costly out of stocks. Example: food and beverage industry, where the volume of business makes out-of-stocks quite costly.

  • Critical “long tail” service levels. Example: healthcare and pharmaceutical companies, and specialty chemical companies, who need to deliver very high service levels, even for slower-moving products.

  • Forecast complexity from promotions. Example: retail industry, where omni-channel commerce and product proliferation conspire to complicate the retailer’s ability to accurately forecast.

ToolsGroup addresses these issues with a statistics-based approach that is at the foundation of its supply chain planning solution. Demand forecasting is at the core of what ToolsGroup does, and it proudly touts the power of its bottom-up forecast approach to be able to look at and make sense of millions of order lines. The advent of demand sensing, however, increased the complexity by an order of magnitude, and ToolsGroup developed its Demand Sensing solution to be able to perform demand analytics on a hundred million data points.

Machine Learning Engine (MLE)

It is, then, perhaps not too surprising that as ToolsGroup saw where the trend was going, it started to work on a machine-to-machine approach. ToolsGroup’s model has always assumed a more or less "hands-off" statistical approach. ToolsGroup’s Machine Learning Engine (MLE) was a logical next step. ToolsGroup will not talk about what is inside the “breakthrough technique,” which has been implemented first inside ToolsGroup’s Trade Promotion Forecasting solution, other than to say that it uses artificial intelligence to analyze promotional events, identify specific qualitative and quantitative variables that have the most impact on demand, and generate a set of rules that model the impact of promotional events on normal sales.

ToolsGroup’s Solutions

ToolsGroup’s message for its mid-to-upper midmarket target base is that its technology and solutions are “Powerfully Simple,” and its statistically driven demand forecasting capabilities have certainly been built around this message. Will ToolsGroup’s nextgen technology, MLE, live up to this promise?
comments powered by Disqus