Adapt or die: Analytics should drive your enterprise evolution

As humans have evolved throughout our existence, so has the way in which we work.  What has allowed our species to survive for millions of years?  It comes down to learning.  From learning the best methods for building a fire to learning the proper way to gather food, learning from mistakes and successes meant life over death. Our ability to learn processes, adopt them over time and then teach them to others has been key to our successful existence.

In contrast, business software, which operates in an environment that’s changing much faster than the African savannahs, remains surprisingly static. Not only is software particularly hard to change (making ‘change request’ such a dreaded phrase), but those changes are bound to be manual. And now, Big Data not only pours into large organizations at unprecendented volume and velocity, but with an extreme variety in forms. At the same time business processes, amidst an Internet of Things, execute at the speed of light. To survive like humans did, companies needs to adapt. And for that, their software needs to learn. 

The evolution of business

Not surprisingly, as people have changed, the way that businesses interact with customers and respond to their needs must change in parallel. There is no cookie cutter example of the perfect customer or the perfect customer interaction. Whether it’s via Twitter, Facebook or the age-old method of picking up the phone, customers want businesses to anticipate their needs before they can anticipate them.

Similar to how we’ve adapted as humans, computers can be taught that same sort of behavior. In business, we see this when we predict that based on a customer’s profile they’ll want a certain type of product.  When they choose something completely different, we’re thrown off by their behavior. By adapting to changes in behavior, adaptive analytics can actually anticipate a customer’s needs in a way that goes beyond the cookie cutter mold.

Businesses face a number of challenges, with losses among the primary concerns. Whether it’s losing a customer or money, both are equally detrimental to a business.  We’re becoming more and more accustomed to the growing rate of change that we don’t think there are alternatives.

This is extremely hard on a business when we think about churn or attrition. The average cost of acquiring a customer, for instance, is upwards of $300 for an average telecommunications company. Multiplied by one million customers, the costs could be exorbitant. Retaining customers, at an individual cost commensurate with (future) value, is much cheaper. This requires predictive models that are bound to get quickly outdated with competitive offers, demographic changes, new regulations, or new available products. Is there any way for a modern business to keep up with the change?

Real-time business

The importance of adaptive models in business is most present when it comes to changes in real-time. In the business world, real-time means as fast as a customer demand can happen – which is really fast, as they’re not communicating by hand-delivered mail anymore, and often not even directly, but instead by a Tweet or post on social media. In the event of a sudden change in customer sentiment, businesses don’t have the time to meet a customer’s new needs. Different from predictive models that need to be re-focused following a change of behavior in the customer base, adaptive models will react automatically.

Like a child that touches a hot surface for the first time and quickly pulls away, an adaptive model effectively learns the difference between a positive or negative response.  To do this, adaptive systems look at data in a fluid form. The basic attributes of a customer are collected, such as age or gender, and many other attributes depending on the context. The customer response will then be related to the customer attributes. An example of how adaptive analytics work in real-time is when an 80-year-old woman calls her cable provider and the customer service representative recommends a particular package or channel for her to purchase. When the woman does not accept that particular offer and requests something completely different, the model will instantly readjust to avoid making the same error for not just this customer but for customers with similar attributes. If enough elderly ladies reject offer A and ask for B the business will automatically adapt to this change in demographics. In reality, of course, the models may look at hundreds or thousands of attributes not just, as in this example, gender and age.

Learning to adapt

As a special breed of predictive analytics, adaptive analytics is a very influential technology for businesses. Today, getting actively recalibrating intelligence out of your data, rather than depending on pre-scripted responses, is what gives businesses the competitive edge they need to survive.  While big data and automation are just the beginning, we need to continue thinking ahead and learn from the past and present to continuously evolve models that provide the maximum benefit possible for the adaptive enterprise. This is the only way that businesses will be able to keep up with the changing demands of their customers and meet their needs moving forward. Survival of the fittest has always been a measurement for how organizations have been successful over long periods of time.  Adaptive analytics help businesses today to stay fit and agile with an eye on adapting to future needs.  Is your business fit enough to survive?

To express your thoughts on Computerworld content, visit Computerworld's Facebook page, LinkedIn page and Twitter stream.
Windows 10 annoyances and solutions
Shop Tech Products at Amazon
Notice to our Readers
We're now using social media to take your comments and feedback. Learn more about this here.