Thursday, January 15, 2009

Loyalty programs: Mining for gold in a mountain of data

In the 1950s, stores gave customers redeemable Green Stamps based on how much they

bought. It was a crude precursor to today's loyalty programs. In addition to the old stalwart

frequent flier programs, now it seems as if every retailer wants customers to sign up for

their loyalty programs. If your keychain bristles like a porcupine with dozens of plastic fobs,

you're not alone. In 2000, Americans held 973 million loyalty program memberships. Today,

according to Colloquy's 2007 Loyalty Census that number is 1.3 billion -- or about a dozen

for every household.

To customers, there's not much to loyalty programs; on the surface they appear to be a

simple piece of plastic and a "Here's how much you saved" line at the bottom of a receipt.

But experts at the W. P. Carey School of Business say that for companies, the programs can

be phenomenally more complex and important than taking a trifling percentage off a

customer's bill.

Building loyalty with technology

The name -- loyalty programs -- belies the larger purpose behind them. While it's true that

companies want their customers to be loyal -- to always shop in their store or fly on their

airline -- there is now much more riding on these implementations. Companies don't just

want you to come back and shop again (although they hope you do); instead they also want

to know what you buy, when you buy it, why you buy it and what else you want to buy.

Companies want you to come back tomorrow, but they want to upsell you on the artisan

baguette versus the plain white bread, and while you're at it, they'd also like you to pick up

the olive tapenade and the imported brie. Not only will you like them, but their margins are

quite good for the store.
 

Your purchase is stored in a database which records everything you buy and have previously

bought. With the resultant mountains of data that accrue over time, a company can bring

data mining and analytics to bear to isolate trends and patterns. That data may be applied

at a macro level where a store will, for example, see that people buying upscale bread also

like fine cheeses and may, thus, conveniently place the two next to each other. On a micro

level, stores send coupons to individuals for specific products based on their own personal

shopping history.

In addition, tracking sales per individual lets a company see who's profitable and who isn't.

Casino loyalty programs look for high-spending customers who are good losers; not

surprisingly, more effort is put into attracting and retaining them versus gamblers who know

when to hold and when to fold.

On top of all of this in-store data, companies can now easily overlay outside information,

says Goul. This may be personal data that correlates your shopping with your credit report

or it may be broader data. Companies may correlate your soup-buying or ice-cream-eating

habits with the weather. In essence, companies now have a ton of data at their disposal.

Tx:wtn

No comments:

Post a Comment