Knowing Your Customers with Data & AI

By | 2017-09-06T08:31:47+00:00 September 6th, 2017|Application Lifecycle Management (ALM)|0 Comments

All businesses have customers. Do all businesses know their customers? How do businesses retain current customers and find new ones? Can businesses afford to chase customers with limited lifetime value to the company? The answers to these questions come easier with modern customer data analysis and predictive AI.

Understanding your customers seemed simpler in the past because information was limited. We relied on our intuition and experience to supplement the information we had.  Today’s modern business has data streaming into data repositories from multiple channels, in quantities that, while seemingly mundane now, would have staggered the imagination a generation ago. Analysis of the compiled data leads to accurate assessments of customer characteristics, which when combined with profitability data, can point to a business’s strongest customers. Using the results of analysis to make data-driven decisions about customers removes ambiguity and personal bias from the process, creating repeatability and potentially reducing random, human-introduced error.

Using data analysis to review your current customers is a good start, but using AI to predict future customer trends allows proactive behavior. Marketing is an inherently difficult endeavor.  Successful campaigns return 1-2% positive responses and increasing this by even a single percent would earn someone a hefty bonus. By simply using AI to filter through the vast majority of disinterested people that would ignore the information presented to them, you can very quickly have higher-quality data for targeting customers. Gold miners don’t pick random spots to stake a claim, they search for locations likely to have higher yields and then mine the profitable dirt.  Why should any other business be any different?  AI will empower you to data mine for good customers in the profitable market segments for your business.

Let’s look at an example of this. New, connected building air conditioning systems return data to a manufacturer’s engineering teams.  A decade ago, this data was used to monitor for equipment failure and, upon failure, notify the owner.  This is a valuable service, but the troves of operating data extend partners ability to provide value.  The data can be fed into AI algorithms to optimize equipment design based on actual usage patterns.  Cooling costs in datacenters is 35-40% of the total operating cost; reducing this cost by even a few percentage points results in municipality scale energy savings.  As we zoom out, multiple sites can coordinate with energy suppliers to balance loads, saving the costs of increasingly expensive infrastructure projects.  Not a bad return for collecting simple machine data.

Collecting more complex, not so well-ordered data presents challenges, but can have increased payoff.  Imagine a company that could predict which customers were likely to churn vs. those that wouldn’t.  The focus for retention and advertising would be entirely different.  My local bank was offering $400-$1000 to sign up. The lifetime value of a customer to the bank must be considerably higher than this amount to justify and upfront cash outlay.  Data analysis drives decisions to outlay the cash and the various tier of customers.

Analysis of data using standard or AI algorithms is the new normal. Failing to keep up on the new technology while your competitors take action will erode your customer base and increase retention costs.

About the Author:

Leave A Comment