Running a large supply chain without advanced analytics is "increasingly unrealistic," says Gartner Research

If you want to deliver your product to your customer's doorstep faster than ever before, you don't need your trucks to travel at supersonic speeds. You just need to make your decisions faster. That's one reason Noha Tohamy argues in a recent Gartner Research report that it is "increasingly unrealistic" for supply chain organizations at big companies to think they can operate without advanced analytic solutions. Only with these types of solutions will companies be able to examine large sets of structured or unstructured data to acquire deep insights, make predictions, or generate recommendations, she writes.

At first, companies may be able to rely on the business intelligence or analytics tools that are embedded in existing supply chain technology or enterprise software, such as some transportation management systems or demand planning systems. In her February report, "Deploy Supply Chain Analytical Platforms to Build Flexible Solutions," Tohamy writes that these types of analytic solutions work well for more traditional supply chain operations that focus on process efficiency and continuous improvement.

But eventually companies will reach a point where these solutions will not be flexible enough or versatile enough to handle the large-scale, dynamic problems they're trying to solve or to help with more innovative or experimental modes of supply chain management. At this point, companies should look to analytical platforms, which are "general purpose software that can perform statistical modeling, predictive analytics, and optimization modeling," according to Tohamy.

These platforms allow users either to build their own predictive or prescriptive analytics model without having to use an open-source programming language such as R or Python or to choose an existing one from a "model library." In this way, advanced analytics help companies take advantage of today's wealth of "big data" without needing to hire hard-to-find data scientists. Instead, analytical platforms offer more of a "self-service" style of analytics that allow line-of-business users (or what Tohamy calls "citizen data scientists") to create queries, reports, and models without having a deep background in statistical analysis or data science.

These analytical platforms can be applied to a wide variety of problems. For example, Tohamy cites one consumer products company that has used its predictive and prescriptive analytics model to a monthly sales and operations planning (S&OP) solution, a regional annual planning solution, and a long-term planning app to determine where to build factories in the future.

To use these platforms successfully, however, companies will need to have already achieved a certain level of maturity, including having skilled citizen data scientists, good data quality, and dedicated resources, Tohamy suggests.