Business by numbers

Whenever I get the chance, I like to read The Economist magazine. The September 15th issue has this fascinating article on Using Algorithms in Business. A key theme of the piece is that the speed and processing power of computers translates into executing tasks with blinding speed using vast amounts of data.
These are few of the several interesting examples of the use of algorithms from the article: Logistics planning at UPS; handling customer calls at call-center operator Convergys; routing information on the Internet; the Queuing of departing planes at an airport....

Global Incident map

I found this Mashup thanks to an entry in Vinnie Mirchandani's blog ( see link to the right). I will have more examples of cool mashups when I return after a hiatus of a few days.

Mapping people's tastes

Analytics is invading the area of soft stuff, thanks to the web, and here is the latest example: The Internet startup, Matchmine, is trying to turn user's preferences into a usable set of data. There is some smart money behind Matchmine. The Kraft Group, well known as the owners of the New England Patriots, is sinking $10 million into this venture.

In Regressions We Trust

In the introduction to his book, "Super Crunchers" (see blog entry from a few days back), Ian Ayres leads off with the story of Orley Ashenfelter. Ashenfelter crunches numbers. Using decades of weather data, he found that low levels of harvest rain and high average summer temperatures produced the greatest Bordeaux wines.

In fact, he reduced his theory to a formula:
Wine quality = 12.145 +0.00117 winter rainfall + 0.0614 average growing season temperature - .00386 harvest rainfall. Guess what? He has been very accurate, predicting that the 1989 vintage would be the best in a long time and 1990 would be even better.

Bill James did the same thing for baseball, as explained in Michael Lewis's "Moneyball".

In newspapers, where I have worked, some publishers have had success in finding causal relationships between various factors such as the weather, sports team's success and so forth to predict store and vending machine sales. And they are using these models to plan their newspaper distribution.

Ayres concludes that in field after field, "intuitivists" and traditional experts are battling Super Crunchers, and that is a big part of his book.

Did Belichick go too far with Analytics?

Even if you are not a football fan, by now you have probably heard about videogate. The NFL expressly prohibits a team from video-taping on the field. Alas, My hometown team, the three-Super Bowl powerhouse, the Patriots, got caught last Sunday against the Jets, despite the fact that they had been warned. Coach Bill Belichick and the team have been punished severely, and the coach for his part, has taken full responsibility and apologized to the team, the owners and the fans.

What does this have to do with Analytics? I mentioned in a previous entry that Tom Davenport's in his excellent book, "Competing with Analytics" cites the Patriots and the RedSox (baseball) as big disciples of using Analytics on the field and to run their business. ( See this CIO feature article by Davenport ).

Teams routinely have staff on the sidelines trying to decipher the hand signals the opposing team's coaches are using to instruct the players on the field. All part of the game to get an edge. Every team does this. But using the the video-tape with the formations on the field as the game is progressing can provide a team with more accurate data and an advantage.

Now, there is some question whether the Patriots have been successful in doing this for the game in progress, or is Belichick so data-driven that he just wanted to add more data to his vast database to analyze and use in future games. In this case, the information gleaned is the specific hand signals a particular coach is likely to use. After all, the Patriots meet the Jets at least twice each year. Last year, they also met in the playoffs.

Perhaps, more teams have been video-taping on the sidelines. But Belichick got caught.

Is Intuition losing ground to data mining?

I am still reading Tom Davenport's "Competing with Analytics". Now comes this book mentioned in last week's Newsweek: Ian Ayres' "Supercrunchers". Add this book to my reading list.

Ayres' thesis is that increasingly, expertise and intuition will be replaced by objective, data-based decision making, made possible by virtually inexhaustible supply of inexpensive information.

Do Amazon's computer's know what we'll like even before we figure it out for ourselves. A couple of examples in the Newsweek article are intriguing: - Are some auto dealers already using data to calculate just how far they can push their customers on price and loan rates? - Are airlines using an algorithm to predict which customers are most vulnerable to being lured away by a competitor and give them, not the airline's own customers, priority in rebooking?

Think of examples in your own business? Your thoughts?

Newsweek also provides a book review

More on Web 2.0

The term Web 2.0 has been with us for almost three years. The publisher, O'Reilly introduced the term at the first Web 2.0 conference in 2004. This excellent O'Reilly piece points out that by Sep. 2005, Google had more than 9.5 million citations for Web 2.0.

An excellent example of Web 2.0, which is often cited is that Britannica Online is Web 1.0 while Wikipedia is Web 2.0. Another example we are all familiar with is the personal web-site versus blogging.

I think a quote from this 2005 Wired article puts it well: " Web 1.0 was commerce; Web 2.0 is people". Among other qualities, Web 2.0 sites harness collective intelligence, customer self-service, and a light-weight user interface.

For a quick summary, this Wikipedia entry is good. Some Web 2.0 features have been present in web applications even before the term was coined: for example Amazon allowed customers to write reviews. And there are no real standards defining what is truly a Web 2.0 web-site.

Web 2.0 and BI

Of late, we are getting inundated with Web 2.0 technologies. The term has been around since 2004, the year of the first Web 2.0 conference. I will summarize my understanding of Web 2.0, and take a look at some of the more popular Web 2.0 web-sites, in another entry in this blog.

However, pertinent to this blog and what caught my eye is this article in DM Review magazine of the impact of Web 2.0 on BI.

more on this shortly ....

Business Analytics - A definition

My good friend, Rich Webb ( he is a BI practitioner who knows SAP inside out), and I were having a discussion yesterday on what exactly is Analytics. We agreed that it would be good to come up with a definiton we could agree on. I cringe when I hear someone say that it is "data mining". No... It is a term that is broadly used for many different processes that support decision support.

So, I turned to the web for the answer. This is a short definition from a popular Datawarehousing vendor "..all programming that analyzes data about an enterprise's business activities and customer information and presents it so that better and quicker business decisions can be made".

Wikipedia has an interesting entry on "Business Analytics", with several topical examples from E&J Gallo Winery and Capital One to Netflix. Many of the examples we hear about are enterprises that market and sell to consumers. However, analytics to improve team performance or getting "getting optimum performance for money spent", has also been used in Sports. One of the earliest examples was described in Michael Lewis's best-seller, "Moneyball" which described the use of analytics by the Oakland Athletics. My two hometown teams, the Boston Red Sox and the New England Patriots have found success using analytics. Tom Davenport, in his book, "Competing .." describes this application of analytics by sports teams.