Demand planners: time to widen your horizon! The challenge was declared by ToolsGroup, who would like to change the world a bit. Successful Demand Planning is not achieved by evaluating last year’s forecast to predict this year’s demand. We have stunning and new ways to evaluate demand. We’re not talking about just forecast accuracy and predictability, but the ability to create and stimulate more demand.
Shot-in-the-dark methods like advertising — do they work? Can they work better? How about your social network site — what are people saying about your company, your products? Can you improve your reputation? Can you understand your audience and their sentiments better and craft a better relationship, better products, better offerings, and more sales? What about the mobile channel and the opportunity to engage consumers through promotions and mobile sales? Your website and store are not going away. In fact, they’re more important than ever.
And we can’t overlook one of the most important questions: Where, in my world-wide network is inventory to fulfill these orders? Demand is not complete unless I can locate the inventory and sell it!
And while we are asking, who can analyze all of this complexity?
Play by the Numbers
With the combination of Money Ball, The Social Network, and Money Never Sleeps, supply chain geeks, such as me, were treated to discussions about analytics, advertising yield, promotions and big data. Up at bat first was Andy Andres, PhD, who is the sabermetrics coach for the Boston Red Sox. His title is Head Coach and Master Instructor1 and we discussed What Supply Chain Planning Can Learn from Money Ball.2
Money Ball fans know that sabermetrics has changed the way baseball works. Bill James coined the term and is widely accepted as the father of sabermetrics; however, the foundation was laid down in the 1960s and ‘70s by less-well-known statisticians. The result is that feel-good advocates have been replaced by number crunchers who look at data. (Sound like supply chain people.) Someone will write a book about this topic (supply chain and baseball) but here’s something to think about: applying analytics makes a huge difference, bringing midlevel players to the top.
Joe Shamir, CEO of ToolsGroup (see below) introduced a real curve ball in his presentation about automating supply chain planning. This topic is very confronting to the way we work, our assumptions, and how we draw conclusions and make decisions. ToolsGroup is a very analytical culture (as they should be) and they like handing out books for their customers and audience to read and think about.3 Joe is into Thinking, Fast and Slow, by Daniel Kahneman who is Nobel Prize winner in Economics. Kahneman discusses his work with Amos Tversky, a pioneer in Cognitive Science who is both a mathematician and psychologist.(If you are into demand planning, these are your kind of guys.) Kahneman and Tversky spent more than ten years exploring and researching how people make decisions, and how they may erroneously draw on their intuitive nature when making them. Groups share assumptions that reinforce decisions.
So, we have these two parallel paths — Nobel prize winners in economics and Bill James talking about the same issue: people making erroneous assumptions. So, what has all this to do with supply chain? Supply chains are run by people who make decisions every day. Joe Shamir talked about taking the decades of work done on manufacturing monitoring and control systems, which have become deeply automated and often light on people, and applying that automated approach to demand planning.
The Supply Chain Control Room promises to be a very controversial topic in years to come. Mr. Shamir had many arguments in favor of an automation approach, including direct experience with one of the largest consumer products companies in the world. Over a several year period, more analytics and machine learning technology were applied in the customer’s model. It got to the point where the systems had almost no people running it. Seems to be working for them, too.
Kahneman’s perspective after decades of research is, “The planning fallacy is only one of the manifestations of a pervasive optimistic bias. Most of us view the world as more benign than it really is, our own attributes as more favorable than they truly are, and the goals we adopt as more achievable than they are likely to be. We tend to exaggerate our ability to forecast the future — ” 4
In Money Ball, the book, by Michael Lewis, Lewis discussed scouts and their assumptions about what made good players. And statistically they were wrong. So in baseball and demand planning, decision-makers may not be as good at it as they think they are. It’s not that people are not smart, he says. Part of the issue, Kahneman points out, is the level of complexity with which people are confronted. When faced with complexity — too many variables — people resort to intuition. (Kahneman and Tversky postulate that intuition works very well in some settings, but not for math and financial-based decisions.) So software, if I can weigh in with all these brilliant minds, can handle layers of complexity and large volumes of data with varying values and levels of importance. Great math can filter through and elevate the issues — the exceptions — and automate the response to the rest.
So, does the future have a lights out supply chain planning department? The question is really, can we create an accurate enough planning model to take people out of the loop? Machine learning technologies have evolved in the last few decades and if applied to a specific situation they become very good, at understanding the5 variables: what is important and what is not; then identifying patterns and applying specific solutions.
Of course, people will still be there. But rather than punching in data, they will set goals by evaluating it. Instead of poring over spreadsheets, they can evaluate the processes, data, and business tactics to see if the desired results are being achieved. So people will be there — but their role will be different. ToolsGroup is bravely launching a very important and controversial topic, since planners tend to pride themselves on being part of the smart set in supply chain.
Demand Variability — the Customer
Baseball, like consumer sales, has the people variable. How people — players or customers — feel on any given day can affect performance, be it scores or sales.
So in addition to using better analytics, using more ways to engage and motivate consumers will help to create demand, not just predict it. Thus, the potential value of this whole new world of social and mobile. This is a new analytics playing field which ToolsGroup has weighed in on. It’s a big topic, so I will just highlight a few points:
- Ad placement and page layout impact sales: how you design print and web-based catalogues and advertising can affect sales.Analytics can help you find the right approach.
- Social sentiment — this new field of analytics has many problems (as would be expected in any new field). Understanding how the analytics work and learning to use them can yield better marketing strategies, provide insights into product acceptance and ultimately, create demand.
- Location-based advertising and promotions also can have a big impact on demand. This topic is front and center for CPG and retailers now, as the whole society has become mobile.
There are multiple opportunities to engage the consumer and stimulate demand. Yet most supply chain applications providers have not gotten with the program here! I was heartened to see Joe Shamir walking through a very practical application of demand leveraging this Omni channel world.
Conclusion – doption of New Methods and Technologies
The art and science of supply chain continues to grow. Art gets replaced by science, gets adopted and becomes standard practice. Standard practice then becomes table stakes, leading companies to seek new and yet better ways to improve. “Companies are always evolving. So those that don’t invest, face a continuing decline of competition position.” 6 Thus, as sabermetrics became standard practice in baseball — most teams have analytics professionals now — the Oakland A’s lost the one brief shining moment when they were the innovators. Teams like the Boston Red Sox adopted these methods and went on to finally break the ‘Curse of the Bambino’7 and win the pennant and World Series in 2004.8
As Dr. Andres said in his concluding remarks at the conference, “Innovators take advantage of market inefficiencies and they get a leg up, but that window — that advantage — gets closed as something becomes standard practice. Then the search begins for something new.”
Yogi Berra, with his legendary way of stating the obvious, said, “You can observe a lot by watching.” So what are you seeing?
Amazon site for Thinking, Fast and Slow
For more information about baseball and Cognitive Science — Michael McBeath, Arizona State University
More content on sabermetrics here
1 Andy Andres Boston University; MIT Science of Baseball Program Andy.Andres@GMail.com — Return to article text above
2 Based on this year’s performance, I think the Sox should start listening more to Andy. — Return to article text above
3 Last year they gave away Money Ball, the book, and tickets to the movie. — Return to article text above
4 Thinking, Fast and Slow, paraphrased from page 255. — Return to article text above
5 ToolsGroup announced their demand sensing and machine learning products last year that is a patented approach beyond Neural Networks for Machine Learning. — Return to article text above
6 Thomas O’Guinn. — Return to article text above
7 Curse of the Bambino refers to the Red Sox’s not so analytical decision to sell Babe Ruth to the Yankees, turning the well-performing Sox into losers and the Yankees into winners. — Return to article text above
8 Alas, they could use some new motivators now. — Return to article text above
To view other articles from this issue of the brief, click here.