Marking the Mystery Customer with Entity Analytics

Today, I had to drop into my local post office to send a parcel.  Waiting in line I couldn’t help but notice the array of seemingly random items in front of me.  Footballs, mini hammers, DVDs, children’s books – not exactly mail-related items.

It got me thinking – we talk about retailers using analytics to target customers with campaigns and offers they are most likely to accept.  But what if you don’t know who your customer is?

Many organisations deal with walk-in walk-out customers on a daily basis.  This makes it hard to identify segments and characteristics of their clientele which could be used to drive more effective marketing campaigns.  Sure, they could offer membership programs and customer rewards to try to capture information about the individual and their spending habits, but with the amount of reward programs on the market it’s getting harder and harder to convince people to sign up for yet another round of email spam.

Fortunately, companies such as IBM are tackling this issue.  Originally, entity analytics was designed to identify criminals – searching for common features and behaviours that would signify that “Bob J” was also known as “Angry Bird” on the streets.  This gave law enforcement agencies the ability to better understand criminal behaviours and patterns, and increase the charges bought against guilty parties.

Now, that same technology is available in the IBM SPSS Modeler product and is being applied to the world of Retail.  Comparing common details such as address, name similarities, buying patterns, credit card details (or a hash thereof), we can put a digital face to the person standing at the counter.  This uncovers a wealth of opportunity to understand both individual customers as well as segments and buying behaviour.

Entity analytics is not something that is restricted to the back office.  Using IBM Analytical Decision Management, we can combine our entity analytic models with defined business rules in real time.  When “Katrina Read” pays for a parcel using her credit card, the customer service agent is informed by the system that it’s the same “Katrina R” who has previously bought mini Essendon footballs at the counter and is likely to take up an offer for discounted football merchandise.

Delivering entity analytics coupled with next-best action at the point of contact increases the chance of offer uptake and making that additional sale.

Don’t know who your customer is?  Maybe it’s time to find out.