Monthly Archives: September 2012

Part of the excitement is around analytics has to do with analyzing Big Data.  But, Big Data needs a good definition as well.  Businesses have always kept data.  What makes this data “Big?”

As an aside, I was talking to someone last week who does a lot of work with the supply chains of different Fortune 500 companies  We joked that it could be a minor miracle to get just a spreadsheet’s worth of good data about the supply chain– much less, anything that looked “Big.”

Obviously, there is a wide range in data that is available.

In 2010, the Economist published a special report, The Data Deluge, arguing that firms are collecting a new and vast stream of data.  This data is coming from sensors, RFID tags, smart phones, on-line transactions, social media and so on.

And, they argue, businesses and organizations are just starting to tap into this data.   Indeed, the rise of the field of analytics is to help firms get value from this data.

The report from the Economist is worth reading.  Click here for the introduction article and here for the link to the survey (be sure to click on the drop down at the top of the page to get to all nine of the articles that are part of the survey).


BusinessWeek recently highlighted Peter Kooman of Optimizely in its Innovator column.

This article talks about a statistical technique, A/B testing, that is becoming a more mainstream.

A/B testing is a way to test two versions of the same website and see which one works better.  That is, a random group of visitors to a website are shown page A, and the other group is shown page B.  Then, you track what each visitor does.  An A/B test is a statistical test that you apply to show which website did better.  “Better” could be measured by more views, visitors staying longer, visitors deciding to buy, or visitors deciding to buy a certain item.

Right away, you then start using the better of the two websites.

Mr. Kooman is quoted as saying that they were able to increase donations on the Bush Clinton Haiti Relief website by 10% by placing a picture of the disaster next to the “Donate Now” button.  They have also used this technique to determine which headlines will get more clicks for a news agency.

You can think about A/B testing like hypothesis testing from your statistics classes.

High-traffic websites are a great place to use a massive number statistical tests to test ideas.  It is very easy to create a test– just create two versions of the same page.  So, instead of laboring over coming up with the best design, best layout, best price, and so on, you can simply test different ideas and let your customers determine which one is better.

And, with a website, you can implement the better idea right away.

The Sept 21 WSJ ran an article on Harley Davidson’s efforts to improve their manufacturing plants.

It was interesting that the article was not just about reducing costs in the plants.  Instead, it focused a lot on their efforts to be much more flexible.  That is, the plants needed to be competitive when demand was low and competitive when demand increases.

When you look at the revenue graph they presented in the article, you can see that revenues decreased by about $1.5B from the peak, but are now increasing.   In such a market, you see why they want to emphasize flexibility.

The article ends with a nice quote about the new mix of automation as well as jobs done with a human touch:

Robots now do most of the welding and metal slicing. They slide flat sheets of steel into an 80-ton press that molds metal into fenders. A computer takes snapshots of each frame or gas tank moving along the line, relaying that information to the painting equipment so it can prepare needed hues.

Automation has its limits. People, some wearing biker garb such as muscle shirts or U.S. flag head scarfs, still do quality-control and assembly work. To check for leaks, workers plunge gas tanks into water basins and watch for bubbles. It’s the same method used for a century at Harley. Mr. Magee shrugged. “It works,” he said.

Fortune magazine ran a short excerpt from Rick Smolan’s new project, “The Human Face of Big Data.”  (Rick is the same photographer who created the “Day in the Life” series where he gave out thousands of cameras for a day in a certain area).  The on-line version also has a short interview with the photographer.

The article is interesting because it points out the various areas where we are now collecting vast amounts of data that needs to be analyzed.  This ranges from education to medicine to crime to farming (pictured) and more.  In the farming example, they point out that a cornfield can be a billion points of information.

It is also interesting to note that the term “Big Data” has entered the mainstream thinking.  People realize that Big Data is not just a small technical niche nor is it strictly something for businesses.  But, it is possible to capture large amounts of data for a wide range of activities.

Of course, we will need people and tools that can turn this data into valuable information.  This is where Analytics comes in.

Northwestern’s McCormick School of Engineering has a new Master of Science in Analytics program.  The first year of the program kicks off in a couple of weeks.

Here is a description from their homepage:

This full-time, 15-month professional master’s degree immerses students in a comprehensive and applied curriculum exploring the underlying data science, information technology and business of analytics. Supplemented by an internship placement and industry supplied projects, graduates will be exceptionally well equipped to harness and communicate the full value of data to the organizations they serve.

This is an exciting program.  It will be great for the students, but also a great opportunity for industry to interact directly with Northwestern.

The Harvard Business Review article, Competing on Analytics, kicked off the movement by suggesting that firms that dedicate all their efforts to “industrial strength” analytics can dominate their industry.

An article in Analytics Magazine that we discussed in an earlier post stated that: “In many ways, business analytics is the next competitive breakthrough following business automation but with the goal of making better business decisions, rather than simply automating standardized processes.”

And, you can find many other similar claims.  I think something important is happening with analytics and it is not a fad.

I have the opportunity to teach at Northwestern in the Kellogg School of Management’s MMM program.  One night, we were discussing the article Competing on Analytics.  It can be easy to read from the article that a strategy should be to turn your organization into one that competes using analytics.

Ben Reizenstein, from Kellogg’s MMM program brought up an insightful question:  Is analytics a strategy?  Or, is it just a tool to help you do what you do, but better.  He was leaning towards the latter.

With all the hype around analytics (much of it deserved), we might have lost track that analytics can help support your strategy.  Maybe it is not a strategy, but a great way to execute your strategy.

For example, in Competing on Analytics, Marriott is used as an example.  But, for a hotel chain, maybe the strategy is to provide the best business hotels, or the best resort hotels, or to be in every market, or to only be in large urban  business districts.  And, the use of analytics is a way to help the hotel chain really execute that strategy.  But, its commitment to analytics is not necessarily the strategy.

As the field of analytics evolves and definitions change, it will be good to keep this question in mind.


This month’s Inbound Logistics magazine published an article I wrote with IBM on different examples of analytics applied to the supply chain.

Here is a copy of the first and last paragraph:

Many companies are building analytics strategies, which use data to facilitate better decisions. To develop improved analytics strategies, consider the three different types of analytics: descriptive, predictive, and prescriptive. Each type uses data in a different way to provide a different type of value.

Which type of analytics is best? When you develop an analytics strategy for your supply chain, you will most likely need a balance of all three approaches.

In the end, understand the different categories helps you create a strategy that is good for you.  No one category of analytics is inherently better than another.  They each have their place.

In yesterday’s post, we discussed a short definition of Analytics (“the ability to collect, analyze, and act on data.”).

Of course, this is a broad definition.  To determine your appropriate analytics strategy, it is important to understand the different categories that make up the field of analytics.

A team from IBM published one of the better definitions of the three categories of analytics in the Analytics Magazine.  The article is worth a read.  Other organizations and universities, are also converging on the same three categories.  These three categories, from INFORMS, are:

Descriptive analytics

  • Prepares and analyzes historical data
  • Identifies patterns from samples for reporting of trends

Predictive analytics

  • Predicts future probabilities and trends
  • Finds relationships in data that may not be readily apparent with descriptive analysis

Prescriptive analytics

  • Evaluates and determines new ways to operate
  • Targets business objectives
  • Balances all constraints


As a side note, the article in Analytics Magazine also is careful to point out that the above three categories apply to structured data.  There is a branch of analytics that applies to unstructured data like analyzing consumer sentiment using Facebook and Twitter or analyzing written text for medical diagnosis like IBM’s Watson (from Jeopardy  fame).


It is clear that analtyics is a very hot topic now.  Companies are committed to pursuing analytics and discussing their successes.  Consulting firms talk about their services to help their clients take advantage of analytics.  Software vendors tout their analytics solution.

But, what exactly is analytics?

Quick searches do not seem to yield good definitions.

Without a definition, there is some danger that the term will just become a another buzzword.  People will simply attach the word to any project to help it get more visibility and attention.

In this post and others to follow, I will attempt to better define analytics and some of its components. This is a fast moving field, so it is likely we’ll have to revisit this.  But, analytics is not a fad and will become a standard part of business.

To give us a starting point, here are four short answers from people and organizations who have put some thought into explaining analytics concisely (note that their goal is to explain and sell analytics, not necessarily give us a good definition):

4.  From IBM

Business analytics helps your organization recognize subtle trends and patterns so you can anticipate and shape events and improve outcomes.

3.  From Accenture

High performance hinges on the ability to gain insights from data—insights that organizations need to make better decisions, faster.

2.  From SAS

Most companies have plenty of data. But gleaning insights from this data to make better decisions remains a challenge. Business analytics can help.

1.  From Thomas Davenport.  Since Tom Davenport Harvard Business Review Article in 2006, Competing on Analytics,  seemed to kick off this movement, we’ll use the first sentence of article’s summary:

Some companies have built their very business on their ability to collect, analyze, and act on data.

This Davenport definition seems simple and concise and captures the spirit of the other explanations.  So, a working definition is that analytics is the ability to collect, analyze, and act on data.

This definition, of course, is fairly general.  For example, haven’t companies always collected, analyzed, and acted on data?  They have.  So, we will dive deeper in future posts to refine our definition and understanding.

Two days ago, we mentioned a Wall Street Journal article on pricing strategies being deployed by on-line retailers of consumer products.

In this case, when retailers are selling the same product to the consumer market, the analytics and optimization problem is about a determining how to react to your competitors price.  It is no surprise that retailers do not like this game since it would seem to lead only to lower prices.

Sometimes the best way to solve a problem is to completely change it.

In this case, it is not surprising that retailers want unique items from the manufacturer.  (And, the items may be unique in very trivial ways.).  This way, it is harder to do on-line comparisons and easier to maintain margins.  Of course we can guess that there will soon be Apps (or maybe they already exist and are used) that will help the consumer determine when items are the same.  But, then expect the retailers to keep tweaking the products so they look unique.  If you are the supply chain manager, don’t expect to a reduction in the number of SKU’s.

But, in some non-consumer commodity businesses, like metals, forest products, liquids, or other industries with a high portion of transportation costs, you can use information about your competitors to help with pricing strategy.

For example, if the cost of transportation is a key factor in the final cost of the product you can determine the transportation cost of your competitors to service various markets.  The location of their facilities is commonly well-known or even public.  You can use network modeling techniques to model you and your competitors to understand the cost to serve various markets.  You can then understand where you are competitive and where your competitors have an advantage.

Note that this is not perfect since you don’t know their manufacturing, sales and marketing, or overhead costs.  So, it tends to work better in industries where transportation costs are significant.