NextGen Supply Chain: Algorithm Council To Help Make Certain AI Pays Off

Editor’s Note:
Christopher G. Caine is President of the Center for Global Enterprise, a New York based nonprofit organization dedicated to the study of the contemporary corporation in the era of global economic integration. He is also President & CEO of Mercator XXI, LLC, a professional services firm helping clients engage the global economy.

George Bailey is managing director of the Digital Supply Chain Institute, a nonprofit research organization focused on the evolution of enterprise supply chains in the digital economy.

Companies work hard to understand, predict and manage customer demand. They know the data they have, and collect from an array of sources, is valuable. They understand that the Internet of Things (IoT) is in full swing with everything from electric toothbrushes to automobiles to credit cards contributing data to it. They are aware of the tremendous power of social media and its profound influence on society and business.

Yet, most companies are frustrated that they are unable to harness this data, clean it, analyze it and apply it to drive demand and the needed supply chain decisions to capture this demand. They feel that their ability to understand, predict and swiftly respond to changes in customer demand is…inadequate.

One of the causes for this is data that is either incomplete, or too narrow in its scope. Another reason is organizational. Companies are typically organized into silos that prevent effective data gathering, data sharing, data analytics and data-based decision-making. For example, many supply chain organizations lack access to the customer data collected by their sales department. They simply rely on the sales forecasts produced in each region. Similarly, many sales teams have limited knowledge of customer preferences for logistics and service requirements.

When a product or service becomes a big hit, the supply chain has a hard time catching up with demand. When a product or service does not sell, then supply chains tend to be slow to reduce cost.

Artificial Intelligence and Machine Learning (AI/ML) is currently being used most often in supply chain manufacturing. Our research, conducted with senior executives from 72 companies around the world, highlighted that AI/ML should also be applied to developing demand for new products and services, and in the management of the flow of goods and services to the consumer. Sales force intuition has not proven to be an accurate indicator of high value and demand for products and services.

“Supply chains used to be about coordinating operations and managing transactions, and much of this was manual. Technology and the digital economy have changed this irreversibly,” said Robert Sinclair, President, Supply Chain Solutions, Li & Fung. “Today transactions are automated, and supply chain management is about aggregating and analyzing data end-to-end to drive operational efficiencies and create intelligence about markets and players.”

The good news is that the use of AI/ML, driven by specific algorithms, can unlock the potential for companies to anticipate and shape future demand and predict the likelihood of future risks.

The secret to unlocking this new fountain of demand and supply is to create an empowered cross-functional team to ask the right questions, collect the right data and build the right algorithms to make the right data-driven adjustments to stay ahead of customer needs and desires. Effective use of AI/ML will require people to define the right problems and perhaps, most important, use their judgement. Imagine a team that had deep customer insight and includes members from Sales and Marketing, Supply Chain, Information Systems, HR, Enterprise Risk Management and Finance.

We call this team the “Algorithm Council.”

The concept of the Algorithm Council represents a fundamental management shift away from making situational supply chain decisions based on slow moving batch information, and instead pivots toward a machine-driven approach that is informed by integrated subject matter expertise and insights.

The Algorithm Council's charge is to develop and continuously improve algorithms that will enable companies to create opportunities to not only sense demand but to also more effectively manage and stimulate it closer to real time than their human equivalents. AI/ML is essential to capitalizing on the flood of data instead of drowning in it.

Where do you start? Begin with taking stock of the algorithms you are currently using within your company. How many are there? What are they being used for? Who else other than the owner of that algorithm knows its purpose? Then ask: “What do I want to know about market demand that I can't presently assess?” And, “How can I get that data and apply it to give me competitive advantage given the assets we have presently?” Next, convene the relevant high-level stakeholders to create a new cross-functional team–the Algorithm Council—to execute a management process to better stimulate and meet demand through a dynamic handling of algorithms.

Some companies have started to experiment with Algorithm Council-like committees, but few have incorporated the right skills and the high-level authority to build and refresh market-winning algorithms that drive customer success. Creating a successful Algorithm Council will require leadership to drive organizational change that breaks down silos and fosters a new era of cross-functional collaboration where success is interdependent. Companies that do this will leap ahead their competition and be positioned to capture customer demand as it arrives.

SC
MR