• by Mark Klein , December 18, 2014

If you knew more than a hundred new people would show up at your door to buy today’s special, would you plan to have that special in stock? You’d be foolish not to do so.

Now, if you knew which customers were likely to defect, or to buy certain products not previously purchased, what would you do? What would your emails say? Would you act differently than you do currently?

Most companies are still sending identical emails to their mailing list.  Some are using segmentation and sending up to, perhaps, five or 10 different versions. But despite considerable lip service to the proven notion that relevance raises response rates, few companies are taking advantage of advances in customer analytics to drive uplift by sending truly individualized emails.

Here’s what state of the art predictive analytics can tell you today, with high accuracy:

• The probability for each customer not to make another purchase before going inactive (“risk probability”).
• The probability for each customer to make another purchase within the next 30 (or 60 or 365) days (likely buyers”).
• The probability for each customer to buy a product not previously purchased, for each of your products (“cross-sell”).
• The probability for each customer to buy a product they have previously purchased, for each of your products (“upsell”).

In other words, companies today can now know what each of their customers is likely to do. The question then becomes, how can they use this new knowledge to raise revenue? It’s not as simple as some may think.  We consider three cases where your email campaign should adapt a different methodology.

Case 1: High cross-sell scores. When we see a customer with a high cross-sell score for a particular product or product category, say 8 or greater on a scale of 0 to 10, we’re convinced that proper marketing can close that sale of that product to that customer. Cross-sell is admittedly very hard to do, with response rates usually well below 1%.  But a cross-sell sale is in some ways like acquiring a new customer, and is often the best way to build revenue.

Spray-and-pray or even segmented emails do not work for cross-sell because they don’t call the recipient’s attention to the specific product you're promoting. To capture this opportunity, the cross-sell email campaign must be individualized, with each email offering the specific products with high cross-sell scores for those product/customer combinations.

Case 2: High risk probability. We’ve all heard the cliché about how it’s easier to sell to an existing customer than to acquire a new one.  It’s true. Since risk probabilities are among the most accurate of customer analytics predictions, when your forecast says a customer is heading toward defection, you need to pay attention. Sending business-as-usual emails won’t prevent the attrition.

Again, individualized emails are the recommended tactic. To get the customer back into buying mode, offer the product(s) they are most likely to purchase, probably an up-sell, and do it with a bigger-than-usual discount if this is a valuable customer.

Case 3: A very loyal customer. There is no need to give away the store here.  Customers like this are regular buyers and don’t necessarily need heavy discounting to continue their loyal ways.  You leave a lot of money on the table when you send them your usual promotional campaign with big incentives.

But their continued loyalty to your business does deserve recognition. One approach that has been successful with this type of customer is to offer a high margin cross-sell product coupled with a decent discount. The good discount is accompanied by a big thank-you.  By tying it to a cross-sell, you expand the market basket of this customer. By offering the discount on a high margin product, you give yourself room to still make a profit on the sale.

Relevance raises response rates, a lot

These cases are just a few of the many examples of doing your email campaigns differently because you have information about buying behaviors of your customers. The river is rising and there are customers knocking on your doors. Sending relevant, individualized campaigns is a great way to use this new information.