Let’s assume you conduct a search on your favorite retail Web site for an “umbrella.” Now assume that nine months later this retailer sends you an e-mail campaign featuring a series of umbrellas.
Maybe this marketing message is relevant to you. Often, the message isn't relevant. You had a need for an umbrella nine months ago, and you filled that need.
In this case, the “half-life” of your umbrella query is small, maybe just a few hours. Within a few hours, half of the people who conducted a similar search for umbrellas purchased an umbrella.
Recency was an important concept in an analog marketing world. “Half-life” is an important concept in a digital marketing world.
You measure half-life by identifying an activity and then calculating the time that passes before half of the individuals will act upon the activity.
Twenty years ago, the half-life of a subsequent purchase was long, maybe six months or a year. This meant that businesses could store names and addresses in a database and market to those customers for a long period of time.
Today, the half-life of a subsequent purchase varies by channel and activity:
- the consumer shopping over the telephone might have a six-month half-life;
- the consumer shopping online might have a three-month half-life;
- the consumer shopping in a store might have a two-month half-life;
- the consumer who receives an e-mail campaign might have a response-based half-life of just 12 hours; and
- the consumer who searches for an “umbrella” might have a response-based half-life of just 12 minutes.
Database marketers capture information to mine it at a later point in time. Increasingly, they have an opportunity to transform the information into something more meaningful. In the case of half-life, you can create a series of 1/0 indicators in a database that tell you if the customer is within the half-life window for a various activity.
A customer who last purchased over the phone six months ago might have a half-life indicator of “1,” while a customer purchasing in a store three months ago might have a half-life indicator of “0.” Execute marketing programs against the half-life indicators that have a value of “1.”
Finally, half-life indicators interact with each other. When a customer clicks through an e-mail campaign, online responsiveness temporarily increases, often significantly, before reverting back to a normal state. Thus, quantify the interactions, adjusting half-life indicators appropriately.
The modern database marketer creates data marts that store transactions across channels. Once data is in the data mart, another data mart is populated with actionable indicators designed to facilitate marketing programs. Half-life indicators represent a great way to identify points in time when a customer is likely to change behavior.
Kevin Hillstrom is president of MineThatData, a database marketing consultancy. He can be reached at kevinh@minethatdata.com.
- Companies:
- MineThatData
- People:
- Kevin Hillstrom