Marketers have heard plenty about the tremendous promise of data-driven marketing, but when it’s time to make decisions about campaigns, budget, and offers, where do you start?

Trends and statistics akin to a crystal ball reside within that massive pile of existing customer information that can predict the future, improve response rates of campaigns, and increase retention and revenue from current customers. Here are five places to start leveraging customer data, alongside your own instincts, to improve campaigns.

1. Decide When To Consider A Customer ‘Lost’
One way to know exactly when to identify a customer as lost or defected is to look at all the transactions for customers who have made more than one purchase. Find the time after which 90 percent of those customers have made their next purchase, and call that the activity period. In other words, the active file includes customers who made their last purchases within the time span that it takes 90 percent of customers to make their next purchase. Some customers will purchase outside of this time span, and others will be marked as inactive and then put back in the active file when they return to purchasing. Setting the activity period this way ensures almost all customers who repurchase will stay active.

2.  Decide How To Divide Your Marketing Budget Between Acquisition And Retention
Often times, e-commerce companies find themselves addicted to acquisition–spending too much time attracting first-time buyers for their sales.  If the bulk of your revenue comes from existing customers, then it’s folly to undermarket to them while allocating most of your budget to acquisition. Acquisition needs a disproportionate share, but not so much that will restrict the lifeblood of sales from your current customers.

Here’s one formula for adjusting your revenue distribution with the assumption that revenue from a group of contacts is roughly proportional to the budget allocated to them: Existing customers should get R/(R+P) of the marketing budget. P represents your penalty–how much more it costs to acquire a new customer compared to making a sale from an existing customer. If you don’t know this number, pick one in the commonly accepted 5x to 8x range. R represents the ratio of revenue that comes from new customers compared to existing customers. For example, if R equals three, this means that three times as much revenue comes from existing customers, a healthy ratio. The result of this formula is the percentage of your budget that should go toward marketing to existing customers. For example, for R=3 and P=5, 3/(3+5) = .375, or 37.5% of your budget should go to existing customers for them to contribute 75 percent to your total revenue.

3.  Which Customers Are Most Worth Keeping
Usually, marketers use total revenue or an RFM score to decide which contacts to include in reactivation campaigns, rewards programs, and direct market campaigns with expensive collateral. Often, every customer in the active file is considered to be valuable. Customer lifetime value is used often on a segment basis, not per individual customer.

To understand your most valuable customers, you should develop and use a value metric that considers variables other than total revenue and recency.  An important and common omission is breadth of purchasing. Other useful components of a value metric are the future lifetime value of a customer and changes in the time between orders.  A good value metric will correlate predicted purchases in the next quarter with number of items purchased and inversely, with recency. While revenue is one component, it is not the only component. Revenue often contributes less than 40 percent to a true value metric.

4. Which Customers To Call When You Absolutely Need A Sale
When needing to reach short-term a revenue goal, knowing which customer to call and expect a positive response is critical. Instead of targeting the customers who spend most, use other metrics that can identify a likely near-term buyer. For example, who represents the greatest up-sell opportunity? Focus on customers with high up-sell probabilities, pitching the right products to those with a high likelihood to make another purchase.

Another approach is to look at customer buying patterns, such as a purchase delay metric measuring how many purchases a customer has missed based on their buying patterns. Each customer has his own unique pattern. A purchase delay of two means that a customer has missed two purchases off his usual cycle and may not be a good target for a near-term sale. Customers with purchase delays less than one are accelerating their buying rates and could be excellent prospects for that sale you need.

5. When To Campaign To A Customer
Wouldn’t it be great if every customer were ready to buy every time you were ready to send out a campaign? Most companies make a calendar-based decision as to when to send campaigns. Typically companies e-mail everyone in their active files from once a month to every day, send postcards monthly, and send catalogs seasonally or more.

Your data can help make a more informed decision about when to press “send.”  Calculate a likely buyer score — the probability that a customer will make a purchase in some predetermined time period covered by your communication, typically one month. Then send your more costly pieces only to those customers whose likely buyer score is greater than 1 percent or 2 percent.

A marketer’s instincts are powerful when it comes to some of these decisions, and relying on your gut is sometimes the way to go. Still, at least look at what the data says. When your instincts match the data, your confidence level is boosted. When they don’t, at least you understand the alternatives. Give your data the chance to help.

Piece originally featured in CMO.com.

Tags :