TCS Seeks to Enable a Digital Integrated Business Planning Process

By Steve Banker

Category:
ARCView

Summary

On October 25th, 2017 Tata Consultancy Services’ Head of Supply Chain - sbtcs1.JPGAnandh Rajappa - introduced ARC Advisory Group to a new TCS solution, Digital Integrated Business Planning. This intriguing solution leverages both existing and emerging technologies with demand-driven thinking to create more adaptive and responsive supply chain capabilities. This ARC View summarizes the solution, what is different about it, and our view of the key capabilities needed to implement this solution.

From IBP to Digital IBP

Integrated business planning (IBP) is the core supply chain process. This seeks to intelligently match supply to demand. IBP is a step forward from the sales & operations planning process (S&OP) that companies traditionally used in that it has tighter integration with financial planning. But Mr. Anandh Rajappa, Head Supply Chain Consulting at Tata Consultancy Services (TCS), believes that IBP can become even better by embracing a demand-driven approach supported with new tools and technologies.

TCS has a comprehensive tool box in place to help companies improve their IBP processes. This includes a maturity model that maps a company’s existing process, compares it to a best-in-class IBP process, and highlights the gaps between the existing and best-in-class processes.

However, technology has advanced and what constitutes a best-in-class IBP process is also changing. A more modern approach to IBP is what TCS calls the Digital Integrated Business Planning Process. According to Mr. Rajappa, the digital aspects of this solution add increased real-time visibility into materials, logistics, capacity, demand, quality, and supplier financial risks.

Just as TCS has tools in place to jumpstart improvements to many processes – including the IBP process – it has developed consulting tools and frameworks to jumpstart a company’s digital journey.

TCS’s philosophy for improving IBP is to speed up the demand-supply matching process. IBP, as practiced by most companies, is often too slow. Many companies create monthly demand forecasts and then create a fixed manufacturing plan that also goes out once a month. In most cases, that manufacturing schedule is adhered to even though monthly forecasts are far from perfect. In some industries, for example, getting to a forecast that is just 70 percent accurate represents best-in-class performance.

TCS favors a demand-driven approach to balance supply and demand. If lead times are short enough, manufacturers should produce to order. While this often is not feasible, today’s planning engines can forecast what customers will consume on a weekly, or even daily basis.

Today’s demand forecasting engines consume much more downstream data, as well as new feeds like weather, economic, and social sentiment data. As new data streams have been added, forecasts have become incrementally more accurate. If newer, short-term market demand forecasts start to show considerable differences over the next week or two than initially forecast, companies should adjust their manufacturing schedules accordingly. To a certain extent, changes to demand can be handled based on intelligently buffering raw materials and work-in-process in optimal locations and volumes.

Meanwhile, executives only need to be notified if the new supply plan will lead to a significant divergence from the initial financial plan for the month. If the divergence is large enough, executives can call an ad hoc, “all hands-on board” meeting to explore options for getting back on track. But to support a customer-centric supply chain, better and quicker demand supply matching is imperative.

A Connected, Collaborative, and Cognitive Supply Chain

Enhanced visibility enables a “connected, collaborative, and cognitive” approach to IBP. Let’s take these in turn.

First, when it comes to a connected supply chain, EDI is inadequate. As Mr. Rajappa pointed out, “20 to 30 percent of (most manufacturer’s) suppliers are not EDI enabled.” Further, EDI is not fast enough. For example, one key EDI message is an advanced ship notice (ASN). But all too often, an EDI ASN really represents an “already shipped notice.” For many supply chain message types, EDI messages arrive too late and with too many data quality issues. Supply chain networks, system-to-system integration, and (as a last choice) portals, all represent mechanisms for better connectivity.

In supply chain management, practitioners understand the need for collaboration. We understand that forecasts could be improved with inputs from key customers, supply planning could be improved through a better understanding of supplier constraints, and replenishment could be improved through vendor-managed inventory and similar programs. But better connectivity to more real-time data greatly improves collaboration.

Finally, the cognitive supply chain is becomingly increasingly real. Huge investments have been made in machine learning. However, too much hype still surrounds this topic. Using machine learning should not be a long and costly science experiment. Practitioners should also understand that there are many different approaches to artificial intelligence and machine learning. Certainly, different approaches are good for solving different types of problems, but applying machine learning is not a panacea.

Many supply chain solution providers have been using machine learning techniques for a long time. Demand planning solutions, for example, have used machine learning to adjust demand models for twenty years. More recently, planning software companies have invested heavily in building out their machine learning capabilities and run successful beta projects to improve supply chain planning.

TCS’ vision of the cognitive supply chain is what differentiates the company. Supply chain planning relies on many models of the supply chain. Some work better solving in problems associated with strategic time frames, some in tactical planning horizons, and some are very good at detecting supply chain risks and enabling an organization to react to large disruptions faster than ever.

TCS’ vision for a supply chain model appears to be more granular and comprehensive than that of other supply chain solution providers. When it comes to supply chain risk management, for example, some solution providers have created a multi-echelon supply chain model for their clients: these are where your Tier 1 supplier factories are located, here are the locations of Tier 2 suppliers, and so forth. These risk models also show how goods flow to final assembly, the origin points, transit lanes, and choke points (like ports). Then, a combination of machine learning agents and people monitor social media and real-time news feeds on an ongoing basis for terms like “bankruptcy” or “port fire” to detect disruptions and then diagnose the impact of that disruption on a manufacturer’s supply chain.

TCS is extending that vision. It is not only major disruptions that can cause a supply chain to falter. Supplier quality issues, carrier delivery problems, equipment maintenance issues, over promising, and a host of other issues can create irritating problems that cause a supply chain to operate far less efficiently than it could.

More Granular Digital Supply Chain Modeling Possible sbtcs2.JPG

 

TCS envisions using machine learning to understand what type of data need to be collected on an ongoing basis. This data will differ from one supplier and supply chain partner to the next. For Supplier A, we need to understand their bill of materials (BOM) to understand that supplier’s sourcing risks. For Supplier B, we need real time visibility to the goods they produce that are allocated to us. For Supplier C, we need to understand how engineering changes demanded by us will affect delivery times, and so forth. Further, the data a company collects from its suppliers will change over time as that supplier’s capabilities change. This more manual visualization layer is supported by an analysis and simulation layer. Over time, as the machine is presented with the same problem and resolutions, business rules can automate many of these issues.

Conclusion

TCS’s Digital IBP concept represents something that is now feasible. The company is taking the concept of a granular supply chain model and extending it beyond what others in the market are envisioning. But a few things need to be understood. First, machine learning is a big data analysis technique. The more data, the better the analysis. It is likely that what TCS is proposing will not just require big data, but “supervised learning.” In “supervised learning,” humans examine the machine’s inferences and grade the inferences to improve the ability of the machine to learn. Finally, the vision will require integrating to new data streams and, in some cases, supplier systems. Cultural as well as technological issues will need to be overcome.

Putting these concepts into effect will require working with a consulting organization with the right supply chain software partners, the ability to integrate to a variety of systems and data streams, data scientists, master data management capabilities, and an understanding of change management. It appears that TCS has worked to hone their abilities in all these areas.

ARC Advisory Group clients can view the complete report at ARC Client Portal on New Client Portal or Office 365 or Box.com on this website.

If you would like to buy this report or obtain information about how to become a client, please Contact Us

Keywords: Integrated Business Planning, Digital, Cognitive Supply Chain, ARC Advisory Group.

Engage with ARC Advisory Group