Marketers might think using advanced analytics is even more challenging than extreme mountaineering with special gear. In reality, however, it may be much easier than you ever imagined. That’s good news, considering that predictive analytics is at the top of the list of priorities in 2014 for tools and techniques to analyze data.
What marketer doesn’t in effect say, “Don’t just show me tables of raw data from hundreds of campaigns. Tell me what it means to customer loyalty, churn, and conversions across marketing channels. Tell me what I can do to be more effective.”
Advanced marketing analytics does just that.
Complex Inputs, Simple Outputs
To the marketer’s creative mind, advanced analytics may look a world apart from the tasks of building brand and promoting products. But once a foundation is established, business users typically find supporting complex business decisions based on advanced analytics easier that they expected. It’s counterintuitive, but true. Let me share an example of how complex technologies are made easy at the end-user level.
Artificial intelligence technology found diverse uses after initial development in the 1980s. One such application was credit card fraud detection by the banks and credit card companies. These highly complex software applications enable banks to automatically detect “abnormal” patterns in an individual’s credit card use. For example, you normally shop for groceries and clothing in Southern California. Suddenly, your card is being used in New York at an upscale restaurant and a jewelry store. The software identifies the abnormal behavior in a highly nuanced manner and almost instantly sends an alert to the bank and credit card holder. The card is blocked. Here highly sophisticated models embedded in the software deliver a simplified – and highly valuable – experience to a non-technical bank employee and cardholder.
I don’t want to minimize the effort required in building a foundation to use data meaningfully in your marketing programs. But advanced analytics takes it a step further by leveraging sophisticated software and statistical algorithms to translate them into easy-to-use information. The underlying technology is complex, but the “output” is designed for marketing people.
Loyalty, Churn, and Purchase Lag
Most marketing organizations today track large volumes of data. Think Web page click-throughs, online/offline transactions, social engagement, or mobile usage. It’s important to track these metrics, but also know they are one-dimensional and are not likely to support marketing decisions. Advanced analytics delivers information of an entirely different value as you: (1) unify multichannel data; (2) segment and analyze; (3) correlate multiple data points to expose hidden relationships; and, (4) forecast seasonal trends – all to leverage these new insights and findings to optimize your marketing mix.
Let me offer a pragmatic example. Our hypothetical marketing team wants to use advanced analytics to understand churn and promote customer loyalty. The business sells branded sporting apparel and gear on its e-commerce site. The marketing team already knows when someone visits the site and, of course, tracks who buys items. They also know it costs more marketing dollars to get a new buyer than to drive repeat purchases. But, how can they determine what triggers customer loyalty or, alternatively, causes higher levels of churn?
In this case, our marketing team had assumed that customers who spend more money would be more loyal. Seems logical. But, the data correlating loyalty to size of purchase – whether it’s a pair of soccer shoes, field equipment, or player jerseys and T-shirts – shows no relationship whatsoever. What seemed intuitively obvious turns out to be an incorrect assumption. Instead, churn turns out to be highly correlated to number of orders; i.e. how many times a buyer purchases different types of products during a given period of time. For purposes of this hypothetical example, the marketing team learns that a buyer who purchases sporting apparel from the website three times in one year is likely to be very loyal. Another product line, let’s say field equipment, requires at least five purchases to produce similarly high levels of brand loyalty. In other words, one product category generates a low-churn/high-loyalty level with X number of purchases. Another product category may require X+2 purchases to achieve loyalty.
This is critical information for targeting the right customer segments with the right promotions at the right time in their lifecycle. The point is analyzing the significance of the factors like number of purchases, frequency of purchase, amount spent per purchase, or buyer demographics on desired outcomes such as loyalty helps the marketing team target customers effectively. Advanced analytics – in this case correlation discovery in loyalty and churn – allows you to easily test your assumptions against real consumer behavior and zero in your promotions accordingly.
Advanced analytics solutions have already started helping marketing teams to better understand consumer behavior across channels, tailor campaigns, and determine how best to optimize the marketing mix.
And, NO, it’s not as difficult as you think. We all know that technology that once looked complex and difficult quickly becomes commonplace in our personal and business lives. It will be no different with advanced analytics. It’s time to take the leap!
Emily Ma, product director of Tencent’s advertising platform products department, was a keynote speaker at ClickZ Live Shanghai where she discussed the ... read more
In today's multichannel world how can marketers use data to ensure the experience a customer receives is relevant to them?
The terms that customers type into your site search function can help you to gain an understanding of user behaviour and can be used to optimise ... read more