Defining Business Analytics & Its Role in BI

Hi Paul,

To your first question, I believe most people would define those terms differently. That said, it's one of those questions that sparks a discussion--different people you talk to will distinguish them differently. The rest of your questions are thoughtful as well, but are outside the scope of a short comment...perhaps I will have to tackle these soon in another article!

Best,

Diann

Hey Diann,

So I'm right to say business analytics is simply 'Data mining'?? Do you know of any emerging trends that are now shaping business analytics? Considering the lengthy time taken to perform one overall cycle of business analytic and the unrealistic nature of the metrics been set, are u seeing emerging trends that are reducing this cycle time and setting up clearly defined metrics?

Also, since most business analytic tools are designed for people with expertise knowledge, how do  results from these analytics be translated to the broader base of business users in a very simplistic and clear manner?

Hi everyone,

Here's a police BI example you might be interested in.

http://www.itbusiness.ca/it/client/en/home/News.asp?id=54673

As for the retail example, Tom Davenport has studied many examples of retailers successfully using analytics. And keep in mind too, the next trend is using information from outside the company (blogs, twitter, etc) to help that company predict where it should go (in addition to what internal).

But beyond that I don't think there's a definitive answer. By that I mean that there are companies that have indeed built their success on being able to utilize the data overload and predict what their customers want--for example, Harrah's or Netflix. But as with anything else, each company has its particular set of challenges and only it can know what the perfect mix of tech tools, gut instinct and so on is for that particular company.

Hagerty highlights this with the questions he brings up and looking at successes in your particular field. Another helpful guide would be Tom Davenport's Competing on Analytics.

But as Mary pointed out earlier: This is complicated. And while there may not be a magic formula, utilizing data to help understand what's happening and make decisions has been used by some companies to outshine and outsell their competitors--this is especially true for customer-driven companies.

Best to all,

Diann

I definitely think that "what to measure" and variations of that is certainly a big issue. From the Accenture study:

"The reasons executives cited most often as to why 40 percent of major decisions are based on judgment rather than business analytics were: because good data is not available (61 percent); there is no past data for the decisions and innovation they are addressing (61 percent); and their decisions rely on qualitative and subjective factors (55 percent)."

Mary,

Not to be harping on you or this topic, just want to say I have been in favor of computers for police use forever. A cheerleader if you will. But the diversity of database vendors and construct tables makes their particular info very unsharable. That was the primary reason the FBI, CIA and NSA weren't sharing terrorist info before 9/11/2001.

Recently the FBI proposed using the 50 states driver's license databases to search for wanted felons using facial recognition software. The ACLU is up in arms over this supposed intrusion on privacy. Many people still call criminal profiling a psuedo science. Can you imagine the hue and cry when the first person gets arrested by a computer? Or more accurately, arrested because a computer accused them?

The FBI and many state police forces employ full time human profilers. Are they a dying breed? like the marketing experts? or will they be relegated to testifying in court to support the computers choice of defendant?

Respectfully,

Kurt

Well, maybe the retail example isn't the greatest. There are other instaces where monitors in a supply chain network -- in pharmaceuticals or in police work, for instance -- combined with analytics can actually produce information on where the likelihood of malfeasance, fraud, theft, etc., is highest. I like that capability.

Another example: We're seeing more criminals nabbed at airports because they fit a profile delivered by software.

Now, I'm not saying software alone will do the trick. Analytics is really only as good as the human input.

It's the same principle at work in the computer -- a boost and streamliner for human thought.

Mary,

I repect your opinions. But geographic diversity predates the micro-computer by decades. American car corps have been buying pre-assembled parts from Mexico and Canada. Even the American legend Harley Davidson has been using parts from overseas since the 70s. I remember a time when a "Made in Japan" sticker was a kiss of death for a product. Assembly lines in China will still be responsive t the orders of a Corp head in Berlin no matter what method of prognostication is employed.

I believe Joe was being kind when he called it "Smoke and mirrors." What algorithm is going to tell the marketing department that "LUCKY" Jeans will sell more straight leg jeans and not to make bell bottoms because the hippie revival is onlyb strong in eastern europe this year? Or did this post just swing a software into making some poor schmuck to start making bell bottom jeans?

Respectfully,

Kurt

I'm not sure I completely agree here, Kurt and J. First of all, analytics is required in a market that's enormously based online. In the bad old days, there was no Web for retailers to use and to compete against.(And retail is just one industry I use as an example.)

Further, there's the issue of scale. Today's market is much larger than retail markets of the past.

Third, retailers themselves aren't just operating locally with a selected export schedule; They frequently use offshore manufacturing and supply chains, and their exports can run internationally in complex patterns.

If anything, analytics seems to be a must given the changes we're seeing worldwide in various markets.

I'm with Kurt on this one.

BI/BA is nothing more than new-age market speak for the old adage that went something like "Know your Product, Know Your Customer"...

It's Business 101 or at least it used to be, maybe it isn't any more and that's a problem...

How did companies do BI/BA in the past, with good old Customer Surveys, I'm pretty sure the 'Net could be used in the same way.

It's simple, if you want to know what your Customers think, you gotta ask them!!

All this trying to analyze site hits, social network pulse, etc, etc. is just a bunch of smoke and mirrors the BI/BA Vendor/Consultants are trying to peddle to create a viable business niche for themselves.

I still contend that any business that doesn't have the internal knowledge to know their market and customers, isn't a business that's destined for success...

Joe

Diann,

This is a well writen, sentient, thought provoking article. But it  egs two questions. First, How did anyone  make money before computers?  I mean Macy's, Saks, Sears, Tiffany's all made tons of dough before anyone imagined IBM. How? Sure some companies like Montgomery Wars dropped off the landscape. But it wasn't due to failing to use Business Analytics. It was more like spreading themselves too thin (Department stores, auto clubs, car repair).

Second. I can't help but think that Business Anlytics, sounds like a new scheme by a marketing expert trying to sell a new use for data that's been collected and paid for once already. A genie in a software bottle, so to speak. The most famous analetics softwre I can recall was a bot system that was set to watch the internet for cues to market trends so that investors could avoid a loss before it happened. The bot ended up predicting that the world ends in December 2012.

Human falability is built into the software that runs the analysis algorithms. Computers do what people tell them to do. Nothing more. Nothing less. Who's computer program could have predicted the PET ROCK or Tickle Me Elmo or Furbies? Nike sneakers foun out by asking an expert why their shoesc weren't selling? He told them they weren't charging enough money for their products. When they quadruppled their price, the sales blew the roof off. What comparitive software will recommend an action similar to that?

Respectfully,

Kurt