Build a Fence Around Your Customers

You lose customers every day. But do you know when they are lost?

A strong company knows the inflection points for customer loyalty and builds systems to turn customers around and bring them back. It takes discipline and hard work, but the payoff is large, and the expense is probably less than you expect.

Finding and understanding these inflection points is critical. There are many places where a change in behavior is reflected in metrics or other data. Before your customers reach these inflection points, or shortly after they reach them, is one of the best places to invest marketing dollars. Many mall-based stores see weekly transactions from core customers, week in and week out, regardless of advertising, so triggered messaging may have little impact outside of typical abandoned-cart-type messages. Wait too long and incremental margin doesn’t pay for the message, since response rates will drop significantly after the inflection point.

The moment a behavioral inflection point occurs, however, you need to spring into action. If your same mall customer has bought three times a month for two years and has not been in the store for four weeks, an immediate effort may be necessary. Think of it as building a fence that redirects your customer back into his prior behavior. Some will still go around the fence and leave, but the right influence can turn some customers around and bring them back.

Web analytics are a great source of these inflection points. A customer hits the “help” page or “contact us”? Pretty good sign there is a problem. Search term equals “complaint”? Same thing. Other, less subtle signs can be found with some investigation. A drop-off in Website visit frequency or e-mail click-throughs for a regular in-store customer can signal a customer on the way out long before traditional measures will catch the change.

Develop a systematic way of collecting possible inflection signals, and start analyzing their strength as predictors of churn. Find a range of likely values, and then start building marketing efforts based on reengaging those customers who fall into the range. Using the same mall example, sending an e-mail to someone who has not bought in 28 days, with a varying group of offers, should eventually result in a time/offer combination that most effectively recovers the potential lost customer and brings them back into the fold. Keep firing that combination, occasionally revisiting the offer and testing alternatives to keep it fresh.

Multiply this by 50, or 500. Suddenly your fence is very robust. Your customers start to wander away but are turned back before they go too far. Some keep going and are lost, but your overall retention will be much better than it was before your effort began.

Think about the math. If you have 1 million customers and lose 30% every year, you are losing an average of 822 customers a day. Its hard to spot them spread out among everyone else, so you need many small efforts every day to turn them back, not one large effort every couple weeks.

The math works in your favor financially as well. Of those 1 million customers, maybe 10,000 will match an inflection point, translating into 300,000 contacts in a month, less than a third of a single newsletter blast or direct mail piece.

They key here, of course, is a combination of solid reporting and analysis, systematic testing, good event sensing, and marketing automation. You’ll need a marketing platform that excels at tracking behavior and managing triggered messaging, and a top-notch Web analytics product is a must. Over time, your fences will start to connect, and eventually you’ll have built a pretty solid structure for catching customers before they leave.

Michael Greenberg, vice president of marketing for Loyalty Lab (www.loyaltylab.com), a San Francisco-based developer of customer loyalty programs for the retail industry, and pens a monthly column for CHIEF MARKETER.