Analytics: getting better all the time

Posted by Steve Cormier

Find me on:

8/3/11 4:27 PM

The pace of competition in business is increasing, fed in part by advances in information dissemination and computing power.  It doesn't matter what type of business, constant improvement is the focus.

The obvious sector is retail, where margins are often thin, competition can be fierce, and immediate product feedback allows businesses to adjust their advertising, promotions, or product placement in time frames inconceivable even a decade ago.    

A customer on the Internet can experience a new set of offers based on aggregate information from thousands of their fellow shoppers, drawn from their reaction to an offer experienced within the previous hours or even minutes.

Retail isn't the only sector subject to demands for constant improvement.  Healthcare, with pressure from government, constant innovations in treatment, and an emerging sense of consumerism on the part of patients, is seeking to identify improved practices and achieve better patient communication.

Utilities, faced with constrained resources and ever increasing need, are looking to smart feedback systems to help them tune their supply and demand balances, based on how people react to various factors in their consumption of services.

The solution to all of these demands for constant improvement is found in analytics.

What is analytics, really?  

Analytics is about getting feedback from existing systems to determine how to alter them for the better.

To do this, data (often large amounts) is drawn from existing systems, and compared against desired outcomes.  Where the result doesn't match the expectation, further analysis can be done to see where things are going wrong.  The system is altered, and the cycle starts again.  Doing this over and over again can lead to deeper customer satisfaction and better rewards to the provider.

Predictive analytics takes this one step further.  By applying algorithms based on extending existing data patterns, expected results of possible changes to the system can be modeled.

Predictive analytics, properly applied, can yield substantial competitive advantage.

Analytics at work

One problem in constructing an analytic system is that, in many environments, it isn't so easy to get the data needed in an accurate form.  So, a big part of analytics is governance, which is about how we make sure that the data being provided to our terrific analytical engines accurately represents the state our of existing environment.

It also provides to users, in an easily accessible format, the exact definitions and derivations of the data they're working with to ensure data integrity and accuracy.

Getting better all the time is a now a fact of business life.

Consumer choice is wide because information is plentiful, and the consumer can quickly discover a better offer.

Analytics is how you can assure that you will be the one making that offer.

Topics: Optimization, Analytics