More than ever, supply chain functional or line of business teams have been frustrated by their increasing needs for broader and more-timely business intelligence (BI).  The reasons are many and in increasing cases, very valid. But more than ever, teams should now be turning their attention towards leveraging processes and technology anchored in prescriptive analytics.

The architectural approach of traditional BI approaches stem from tapping centralized data warehouses, where all forms of data and information from various business systems are stored. Such data, information and reports are often historical in nature and require additional work in either Excel spreadsheets or “rules based tools” to convert data into needs for more predictive information. BI extraction

The notions of product demand forecasting based on historic sales of the product, or a particular customer’s demand and revenue history were often a function of such needs for more intelligence. The need in traditional BI was to allow users the ability to contrast plans with actual results or to prepare sales and operations planning (S&OP) teams with needed information to make important decisions related to addressing product demand or supply gaps. The ability to leverage hidden intelligence was often constrained because of the resource limitations of IT, the complexity of the centralized information warehouse, or the turnaround time for information requests . When business teams finally get the insights they seek, the gap between data and optimal decisions has been lost or too difficult to overcome.

Two other important trends have since occurred. First and foremost, the clock speed of market changes and/or business events has dramatically increased requiring that supply chain and line-of-business teams anticipate such changes and be prepared with various scenarios for more informed response to such changes. Overall complexity of supply chain decisions has further increased. There can be occurrences of increased risk, quickly evolving industry and market opportunities, or needs for more impactful supply chain efficiency and cost reduction.

The second is today’s continual advances in information technology surrounding in-memory computing, streaming Big-Data analysis, more user friendly data visualization tools and cloud-based platforms. More sophisticated mathematical optimization techniques, which were previously only available through custom coded software or by the hiring of experienced data scientists, is now becoming available in packaged software.

In our previous Supply Chain Matters posting: The Journey Towards Integrated Business Planning, we expanded on our prediction that S&OP processes will morph into broader forms of integrated business planning that are more anchored in analytics support capabilities. Decisions are no longer anchored on what occurred in the past, but rather, what is expected to occur based on various forms of both internal and externally available information. Decisions are not one-dimensional but instead predicated on the most up-to-date information placed in proper context as to impacts on business metrics or required business outcomes.

Supply Chain Matters sponsor River Logic Software succinctly describes these decisions as:

Descriptive: Which products and customers are most / least profitable in this plan?

Diagnostic: Where are our marginal opportunities to improve profitability?

Predictive: What happens if the price of a key input component rises dramatically because of unplanned market dynamics?

Prescriptive: What should we do if our new product innovation doesn’t drive the forecasted demand?

As was observed in our prior commentary, many of today’s legacy enterprise systems are anchored in heuristics and data models focused solely on product demand, supply, capacity and inventory data from a historic perspective as contrasted to a what to anticipate perspective. S&OP participants desiring broader business intelligence and insights are more than ever expressing a need to make forms of descriptive, diagnostic, predictive or prescriptive decisions. The former two types of decisions often originate in tactical S&OP phases while the latter two often originate in executive level S&OP decision-making. The key is to bring together all existing forms of internal enterprise data, be it supply chain, procurement, product management, financial or customer focused.  It is further predicated on functional and business teams to have more user-friendly tools and techniques to assemble and context these levels of analytics that can support such decision-making.

Thus, supporting needs for more timely, contextual and integrative forms of decision-making should be predicated on a framework of decision support analytics that context available enterprise-level data that spans not just supply chain, but other business and functional information sources.

It is no longer a limited context warehouse approach, but rather an enterprise capability to transform information into analytics powered decision-support.

Bob Ferrari

Disclosure: River Logic is one of other sponsors of the Supply Chain Matters blog.