Online customer behavior lessons for retailers
Online customer behavior isn't that different from the offline kind. If you've ever walked into a store, looked around and left without buying something, then you've sent a message to the retailer. The same is true when you navigate to an online retailer, click around, then leave. The difference is that while you might wander into a store to kill some time at the mall, that's not as likely with online shoppers. People who visit online stores are usually looking to buy an item or doing research for a future purchase. By analyzing their actions and other online customer behavior data, you can get a good sense of why they're not buying and what you can do about it.
Analyzing bounce rates
Jeff Oxford, SEO director of consultancy 180Marketing, says a key metric to watch is bounce rates. As he explained in an interview with me, a retailer's search engine ranking often skews results for click-through rates (CTRs). For instance, the webpage in the No. 1 position will always have significantly higher CTRs than a webpage in the No. 8 position.
Instead of looking at CTRs, Oxford said bounce rates, or the percentage of visitors who leave a site without checking out any other pages, is a better indicator of a site's viability.
"A product page with a high bounce rate could mean that the price is too high, the product reviews are negative, there isn't enough information or the product just doesn't solve their problem," he said. "Focusing on a webpage's bounce rate is a great way to improve usability and conversion rates, but also positively impact search engine rankings."
Using site testing
Trying to get to the bottom of what's not working is a complex undertaking. Bob Phibbs, CEO of The Retail Doctor, explained that A/B testing, where all elements of a page are kept the same except one variable, is one way to get closer to the truth.
"The more you can nail down why time on a specific page is low or trending downward, you can alter one element and test or ideally do an A/B test just one element at a time," Phibbs said in a recent interview. He noted that while analytics can synthesize the data created by these tests, retailers will still need someone to examine the data and draw conclusions on what needs to be changed.
Oxford agreed: "Testing price points, adding high-quality product photos or videos, including unique selling points in product descriptions and even improving the load time of the page can decrease bounce rates and keep more visitors on the website."
Leveraging heat maps
Another way to get a sense of what's catching consumers' attention is by using visual analytics like heat maps. This technology tracks and displays customers' movements on a page by noting mouse movement, clicks and scrolling. While this visual representation of data may be useful in adjusting site design to encourage purchases or tailoring marketing efforts based on customer interests, Phibbs noted that retailers need to put this information in context. For instance, a red dress might catch a shopper's eye because it is a bold color, but if she does not click on or buy it, that may just be an indication that she doesn't like the style. In this situation, retargeting ads won't help move the customer closer to a purchase and would waste valuable resources.
As Phibbs notes, this process is not yet fully automated, and retailers still need someone to interpret the data. However, analyzing results from heat maps and A/B testing can help rule out some variables. For instance, if the problem is that prices are too high, testing lower prices that yield higher CTRs will demonstrate that price was likely the main barrier to sales. In other cases, like the red dress example, it may be a case of personal taste, price, the consumer's budget or number of other factors. Either way, retailers should be paying close attention to see what patterns emerge.
Untangle structured and unstructured data to optimize your retail business and improve the consumer experience. Visit IBM's Retail Solutions Industry Page to learn how.