LLamasoft Supply Chain Blog

← Back to Blog

Picture This: Why Supply Chain Visualization is Key – Part 1

By Alan Pinkert  October 28, 2015

The night before January 28, 1986, the engineering team in charge of designing the rockets for Space Shuttle Challenger were faced with a problem.  The shuttle was scheduled to launch the next day, but predicted temperatures were below freezing, and concerns arose that a key part of the rocket, the rubber O-rings, might fail under such conditions.

The engineering team faxed a report to NASA recommending to delay the launch, citing that they did not have enough data to support launching in such cold temperatures.  But that team was overruled, and the next day, the Challenger broke up over the Atlantic Ocean, claiming the lives of all seven astronauts on board. The subsequent investigation included an examination of the engineers’ recommendation, and found images like the following:


Edward Tufte, a pioneer in the field of data visualization, made the argument that the way the engineers presented their case was to blame.  In his 1990 book, Visual Explanations, he writes, “As analytical graphics, the displays failed to reveal a risk that was in fact present.  As presentation graphics, the displays failed to persuade government officials that a cold-weather launch might be dangerous.”

He created a simple graph of the data instead, which clearly shows an increase in the historical level of damage (Y) as temperatures decrease.  The black bar on the bottom left was the predicted temperature on the day of the launch:


He asserted that had the engineering team used the right visualization, NASA officials would not have overruled them and launched against their recommendation.  Tufte’s claim that blame rests with the engineering team has since been debated.  But the anecdote endures as a cautionary tale about the proper usage of data visualization.  He concluded, “There are right ways and wrong ways to show data; there are displays that reveal the truth and displays that do not.”

Fortunately, supply chain decisions aren’t usually life-or-death (although we do work on some projects where they are).  But data science skills can make the difference between millions of dollars in savings and the status quo, and supply chain executives are taking notice.  As Forbes reported, respondents to SCM World’s 2014 Chief Supply Chain Officer Report ranked big data analytics as the number one disruptive technology for supply chains.

Why is data science so key for supply chain designers?  In order to champion transformative organizational change, people at all levels of the supply chain need to advocate for the same narrative.  Ultimately, the best way to turn analytics, metrics, and raw data into a universal story that everyone can get on board with is through data visualization.

The Visual Advantage

People are great at recognizing visual patterns.  Recent studies show that subjects can recognize various scenes, such as a picnic or a smiling couple, in as little as 13 milliseconds.  We are so visually adept, that computers undertaking complex visual tasks like facial recognition still haven’t beaten the human brain.  Applied to data, this skill allows us to see trends and outliers in a picture far better than in words and numbers.  Data visualizations have several advantages over data tables.

1) Statistics Only Tell Half the Story

Even with the best of intentions, statistics can be accidentally misused.  One of the most striking illustrations of that is Anscombe’s Quartet, created by statistician Francis Anscombe.  It’s simply four data sets with the following data:


Some quick calculations would show that each data set has approximately the same mean, variance, correlation, and regression.  At face value, the statistical analyses might give the illusion that the data sets are fairly similar.  But when graphed, each data set tells a much different story:


How similar do these data sets look now?  Relying on summary statistics alone can lead to bad business decisions.  Complementary visuals are critically important to gain a full understanding of data.

2) There’s Too Much Data

Supply chains often comprise thousands of physical locations, millions of SKUs, and orders of magnitude more transactions, processes, and relationships.  Any tabular representation of this quantity of data quickly becomes too big to display on a single page and difficult to derive meaning from.

By contrast, a single visualization can reveal unforeseen patterns present in truly massive data sets.  For example, Twitter created visualizations of geotagged tweets to combine billions of data points into beautiful images of human activity.  What would you expect to see?


There’s no aggregation, no prior numeric analysis, just a map created from raw data.  Cities, roads, and other hubs are clearly visible.  But a closer look reveals ghostly spokes of maritime traffic off the coast of several countries.  Without an image, these hidden and unexpected patterns can be lost to noise.  But visualized, vast data sets can become a tool for exploration.

3) And the Data Is Too Complex

With too much data, a table is daunting, and minimally helpful at best.  Add multiple data parameters, and the story is impossible to understand without visualization.

An often cited interactive is the The Wealth and Health of Nations, which incorporates five different dimensions: income per capita (X axis), life expectancy (Y axis), population (marker size), world region (color), and time (slider/animation).  The underlying numerical data is difficult to draw conclusions from, especially because it comprises different units, like years, income per capita, and total population.  But visualized, it tells an incredible story about the rise and fall of different countries and regions of the world:

This is an interactive map, so click on the link in the paragraph above to see it in action.

This is an interactive map, so click on the link in the paragraph above to see it in action.

With visualization, we can understand the impacts of a supply chain design change on cost, service, sustainability, and risk.  Using different visual aspects like color or size to represent these different concerns allows your audience to understand multiple impacts at once.

Check back tomorrow for how to apply these principles to create impactful supply chain visualizations in Part 2.

Additional Notes

  • LLamasoft founder Don Hicks created a fictionalized version of Tufte’s Challenger analysis, but for supply chain management, back in 1998. See The Value of Graphics In Communication (IIE Solutions, 1998).
  • For anyone interested in data visualization, Edward Tufte’s book The Visual Display of Quantitative Information is an incredible read.
  • IBM used the Wealth and Health of Nations interactive in a 2013 Supply Chain Visualization article to make a similar point about dimensionality, but I think the point bears repeating, and the specific interactive is an enduring example.
  • Even if you’ve seen the Wealth and Health of Nations interactive before, I’d recommend watching the BBC’s report, where the creator shares his story in person.