Understanding customer behavior is essential for organizations to know where they're doing well and where they need to make improvements. Especially in today's super-connected world, customers are leaving behind several nuggets of information that will help companies get a better grasp of what their customers are doing.
For Skinit, an online retailer of custom "skins" and cases for electronic devices like phones and laptops, getting insight into its customers' activities online was essential to make the necessary improvements to its online property. As an online-based company, the brand had a lot of insights from its customers' buying journeys, but needed an efficient and in-depth way to analyze the data.
One pressing need was to identify any errors on the website and address them before these became a problem for customers. Kate Bartkiewicz, the company's former manager for business analytics, notes that the web analytics team was spending a lot of time recreating these errors to try and determine how customers got to an error page and fix the underlying cause. "We don't want to spend time recreating the scene but instead finding a solution," Bartkiewicz explains.
Skinit wanted to find a solution to this reoccurring pathway problem that was wasting its three-strong web analytics team's precious time. While the company had ample customer behavior data in hand, the data was an aggregate that didn't reveal issues that individual customers were experiencing. "We needed the ability to replay customer sessions."
To address the situation, last year Skinit implemented an analytics platform by Cloudmeter, part of Splunk, which gives the web analytics team not only raw aggregate data, but also the ability to replay individual users' sessions. Bartkiewicz notes that instead of depending on a random sample, the analytics team can set the parameters to search for sessions that have specific criteria, for example instances where a customer abandoned the cart on the payments page.
A recent example where the new system solved an issue in real time was when Skinit noticed that a small group of customers was attempting to order a product that was no longer in production. While the company was able to fulfill the order, avoiding a negative impact on the customer experience, Bartkiewicz wanted to find the cause of the problem. The replay sessions allowed the team to determine that some customers were accessing an older landing page that had a link to this product. "It only took a couple of minutes to solve this problem," she notes. Similarly, every time Skinit detects anything unusual in the data, for example an increase in certain orders, the company can take a more granular look and address the reason for the spike.
An added bonus of the new analytics platform is the ability to segment customers. "We want to know who's buying and what they're buying rather than just what items are selling," Bartkiewicz notes. Today, the team can look at each customer's journey on Skinit's website and determine what other products they looked at. This insight is allowing Skinit to identify customers' preferences very quickly to make personalized recommendations, for example sports-related cases related to the city where the customer lives. So a customer in Pittsburgh may see views of Steelers-related skins when they land on the site.
The time saved on getting to the bottom of problems has allowed Skinit to spend more time optimizing the website. "It has allowed us to move from being backward looking and instead solve problems." Bartkiewicz notes that the no-code based solution led to a very quick implementation and her team was able to make the necessary changes. "The value of this cannot be understated because getting into a sprint is not an easy process," she notes. Further, improvements to the site have led to a decrease in cart abandonments since Skinit is addressing issues that would have otherwise stopped customers from completing their purchase.