Autonomous! Insights from ToolsGroup's TG18 User Conference
on Apr 12, 2018
Future supply chain jobs will be different than jobs of the past. That we know. Technology, in the past, had increased the role of planners, expanding their analytical capabilities and scope.
Today, technology is moving ahead, forging a new path, one that we cannot fully glean; and it may reduce the presence of planners in the office landscape. However, planners can gain greater scope and authority with the use of this increased analytical power. But they must learn to understand and control the technology to control their future.
What one software company is doing to help build that future.
Full Article Below -
A worldwide supply chain audience spent an energizing, yet disquieting, two days in Boston, hosted by ToolsGroup. Why disquieting? The supply chain profession is going through the biggest change since its foundation, catalyzed by a generational shift in the workforce, as well as epochal changes in technology. And, at the same time, another factor is the reflective disruption of the traditional business model whether for manufacturers, retailers or healthcare providers.
In fact, Pat Smith, Managing Director of ToolsGroup USA, kicked off the conference by asking the audience, a representation of retailers, brand companies, service, and manufacturing companies, “Who is going through disruptive change?” It seems the entire audience raised their hands. So it was fitting that Pat’s keynote jumped right into the major technology shifts, their importance, and their ability to help companies gain a grip on this disruption.
Topics like machine learning, leveraging all that big fuzzy data from the web (see side box for expanded discussion on big data), IoT and Blockchain were all easy to pop into the conversation. But these are just enablers of the bigger topic: the changing world of business competition due to the millennials—our employees and customers. That change is disrupting the dynamics of how we trade with partners and serve customers. But to the point, the very job of supply chain personnel is changing. (This was a very timely topic, especially for us, since ChainLink is currently conducting research on these changes. Read Thinking Machines in this issue.)
Today, as we all know, there is a staggering amount of content and information out there. Just as we are getting a grip on understanding our own data, we realize that there is a tremendous amount of external data about the momentous things that affect our business—customer trends and demand, quality of supply and suppliers, andour competitors. Thus, we must learn how to harness that information, a huge task, I noticed, that a very few have embarked upon. The role of the technology provider, today, besides supplying very smart algorithms, is to, in Pat’s words, “liberate humans from data overload.”
Beyond the eloquent words, what exactly is that liberation? Of course, many tech companies are touting AI as a way of sifting, sorting, and solving problems. And that is good. And really, machine learning software that has had some years of on the job training can assume some tasks for the users, reducing their workload and providing sharper clarity and accuracy of plans. But that only deals with one element of the planner’s job, with the job being fundamentally the same. The problem is that we have a scarcity of planners. And the new planners coming into the workforce have very different expectations of how they want to work.
So liberation has become, quite literally, liberating planners from planning.
To that end, ToolsGroup now provides managed services1—specifically, planning as a service. This concept of service will begin to take on a greater role in the supply chain tech market as we continue to experience the turnover from boomers to millennials.
Liberating users from the day-to-day is essential, since there is a lot of work not getting done. If we are to understand and deal with the huge shifts in our markets, planners will have to focus on new things that will help meet the changing needs of the customer, and/or optimize the business for increased profitability and cash for new investments in the future. These are the kinds of analyses they have little time for today.
Many of the speakers voiced concerns about being “far behind.” Most, though, learned through attending conferences and talking to peers that although they might be behind in some things, they were on par or ahead in others. That, however, did not give them peace. In fact, what it says is that as a profession or function we are underinvested. For example, Pat asked the audience how many actually had an integrated supply chain—not just ASNs. That qualifier quickly excluded just about everyone in the audience. Very few hands went up. So although we talk about digital supply chains and trading partner connectivity, we have very little of it.
Pat also asked how many had automated their pricing and promotions. Again, only a smattering of hands went up. In dynamic markets like retail and consumer products where promotion and price changes are a daily activity, and where dynamic pricing on the web can change hourly, it would seem that companies would be all over that. No. However, it is on the to-do list. Our to-do lists keep growing but we are not checking off the completes!
We just don’t have the time. In fact, as one manager of planners told me, “We are so underutilized for our talent and knowledge by our companies.” Too much time is spent on menial and day-to-day tasks and not enough on the big picture. But if 60% of your tasks were taken away, you would have that time. And here is where the managed services become so important today.
The Journey Forward, the BIC Story
And, so, that is exactly what BIC has embarked upon. Bill Chiarelli, Senior Director, Supply Chain at BIC Consumer Products, talked about BIC’s journey and the challenges they face. Not only are their traditional markets being challenged as consumer habits change, but their workforce is changing. Planners are hard to hire or retain. Bill said they don’t find the work satisfying. He also said that they tracked people after they left BIC and often they actually leave the profession. In addition, it seems nobody wants to own the demand planning team. Neither sales nor supply chain wants the responsibility.
Think about it—a major, ubiquitous brand company with good paying jobs is having trouble finding and retaining professionals! And, equally important, consider their urgency in understanding the market2 so they can have a winning, growing game plan for the future. Yet there they were. It’s not a criticism, because they are not alone, it seems, in this dilemma.3
This talent issue is a problem across the supply chain as well as other job functions. And as new environments and technologies allow for the envisioning of new tasks or roles, as a society, we run a risk. The talent gap may be exacerbated by the very technology that is designed to lighten employees’ load and make them more effective. The lack of talent for emerging roles is creating a vicious cycle that is leading to increased automation and could increase that lack of fulfillment in their jobs if we don’t plan ahead and design jobs that leverage technology and the expanded access to big data.
BIC is currently working with ToolsGroup to implement demand planning as a service. CPG companies have many product forecasts that can be put on “autopilot” by using really smart software that can adjust or elevate important exceptions. Most organizations that have been doing forecasting with advanced demand planning packages are pretty good at knowing which products really don’t require a lot of attention—low-touch forecasts—and which products need that human touch. Thus, BIC is initially putting their low-touch products (about 60%) on autopilot. That is a couple of billion dollars worth of product.
Companies have made efforts in the past to outsource planning, but most companies did not talk about these initiatives very openly as they were often contentious amongst their existing employees. However, that obstacle is obviously going or has gone away.
The growth of managed process services in the analytical and planning role is taking bolder steps forward today.
Autonomous Supply Chain Planning
ToolsGroup has been pursuing the path to autonomous planning for several years.4 As a result, they have built up a deep understanding of what can be done and the path users may take to adopt more automated planning environments.
Besides BIC, ToolsGroup has many customers who have had several years of experience using machine learning and the expansion of the data. Tools Group has significant history in machine learning/neural networks and social analytics, and continues to incorporate more types of data in their demand analytics software. There were several practical presentations on incorporating machine learning and the improvements it can provide in forecast accuracy.
ToolsGroup also has some customers who will talk “on the record” (and some who will not) about their migration to Autonomous SCP (Supply Chain Planning). Costa Coffee is a case in point. Today, Costa can manage 8,000 very dynamic demand points with one planner. These demand points are not like that of a big box retailer with weekly demand, but demand that can change by the hour!
Figure 1: Costa Coffee’s Implementation of ToolsGroup solutions (source: ToolsGroup)
Here are other examples of ToolsGroup customers who have implemented some version of autonomous planning and inventory management:
Absolut Vodka has a single planner for the entire global brand across 350 different Vodka products across 150 countries.
Cipla Medpro pharmaceuticals automated their statistical forecasting and say it is 20% more accurate than their own market intelligence. In fact, their top supply chain executive director said, “We are now at the point where we can confidently switch off our manual override and put complete trust in the system.”
Lennox Residential achieved 99.7% no touch, computer controlled automation of their planning and replenishment. At Lenox, 997 out of 1000 planning decisions have been automated to the point where there is no manual intervention.
Service companies are another example of the predictive power of machine learning as service providers not only improve their inventory availability for repairs, but also avoid downtime by doing pre-emptive predictions on maintenance requirements.
So what can Autonomous SCP look like for demand planning? Figure 2 is a simple way to consider the role of the machine vs. the planner.
Figure 2: Planner vs Machine Roles (source: ToolsGroup)
Migration to these types of environments does not happen overnight, ToolsGroup needed to point out, since everyone listening to these presentations was envisioning—hoping for—the machine-learning silver bullet. Several years worth of planning data and history are required so that the machine/software (and the people) can learn.
If I am willing to take my “organizational hands off the wheel” a bit, how far can this go? The supply chain function has proven to be dynamic and has grown and stretched more than many other functions since the mid-nineties. There will be new models of how these departments will look in the future and what the tasks and goals will be.5 And these changes will be embraced by the broad-minded supply chain profession.
As Tools Group continues their innovation in autonomous planning, they will continue to pursue an increased understanding of the nuances and challenges of truly big and fuzzy data on the web and in other data streams. This unique mission gives them a bold and innovative position in the market.