Analytics Blog

To Coupon or Not To Coupon: What Does The Data Say?

In February of 2012, JCPenney CEO Ron Johnson decided consumers didn’t want coupons anymore. With this in mind, he decided to rebrand the retailer. Soon, a new logo was introduced along with “fair and square” prices and boutique departments. Long known for their coupon offerings, JCPenney seemed doomed from the start: their 2012 first quarter earnings were down 20 percent and the retailer experienced 10 percent less floor traffic. After it was clear how much damage was being done, attention shifted back to coupons. Was the coupon’s demise partly responsible for the company’s problems? It turned out the answer was a resounding “yes.”

Image Source: Coupon.org

Image Source: Coupon.org

For retailers today, it is important to understand how consumers really feel about coupons. Before a corporate decision is made to offer coupons, retailers need data, and lots of it. They also need an easy way to analyze data from various sources and perspectives in order to reveal whether the idea will truly make an impact – or just become a statistical failure. Predicting Consumer Behavior Coupon.org recently ran a story called “The Comeback Coupon,” revealing that higher-income families ($100,000+) clip coupons more than families bringing home under $35,000 per year. College graduates also clipped significantly more coupons than those who completed high school only. Surprisingly, despite all the Internet coupon websites and in-store QR coupons we see, a whopping 89 percent of coupons are still printed in newspapers and magazines. However, it is a near certainty that online and mobile coupons will rise in popularity in the coming years.The rise of online coupons is perhaps most obvious on YouTube, where a large number of viewers are subscribing to coupon experts such as The Krazy Coupon Lady. (Her channel alone boasts over 3 million video views to date.) Stores are also engaging consumers with “popular purchase” discounts, where consumers can make a list of favorite items and, at checkout, learn if the product comes with a discount offer.

But before initiating a coupon campaign, it is prudent for retailers to employ some data science. Analysis of the relevant data sets can be done effectively using Hadoop. Hadoop has an ability to cross-reference large data sets, and when combined with analytics software (such as Alpine), it can reveal consumer groups or cohorts based on coupon usage patterns. This is known as descriptive analytics and can be done through unsupervised learning algorithms like Decision Trees and K-Means clustering. Price points are another area where data science can be useful. Prices are typically well thought out by manufacturers in order to compete with the prices of similar products. While they do have the flexibility to fluctuate if a product’s popularity rises, their relationship to coupons can be a complicated one, and relying on a suggested retail price as a starting point for a coupon campaign can be a bad idea. Smart retailers undertake very detailed analysis that attempts to measure not just the revenue derived from product sales but the marketing acquisition cost of the customer, return risks, customer service costs, etc-which can become Big Data very quickly. The Power of Big Data There are many variables to consider in the coupon decision. Beyond the demographics of your consumers, there are: needs vs. wants (you need food, you want designer shoes); regional price points; and types of products to coupon, among other factors. Data science can help retailers make better decisions when designing coupon campaigns. Information coming from Hadoop and Enterprise Data Warehouses can now be merged with consumer buying patterns provided by 3rd parties. Incorporating other factors like consumer retention and loyalty can also determine the likelihood of coupon acceptance. Further, it can answer questions like “Does the coupon cheapen the retailer?” and “Does the coupon engage the customer?” Applying big data analysis to consumer behaviors in order to make sound retail decisions is simply good practice.

Sound Decisions By October 2012, JCPenney had fallen to its knees and offered a coupon. However, according to Reuters, it was just a $10 “loyalty” coupon and did not move the needle. With data science, big retailers can now make sound decisions and set realistic goals. Alpine Data Labs, one of the leaders in the field of data science with Hadoop, offers innovative software that takes away the time consuming process of building and deploying statistical models for retail decisions.