Five Metrics for Measuring Ecommerce Personalization Success

In today's ecommerce world, customers have a dizzying array of choices when it comes to purchasing a product. Buying ice cream, for example, is no longer a matter of making a selection of whichever brand is the cheapest from the local convenience store. Consumers now have the capability to zero in on their preferences to choose a product that best suits their needs, wants and lifestyles. They can make considerations that go beyond location, taste and budget, such as: is this the flavor I like? Is it fat free? Is it organic? Is it dairy-free? Do the products come from fair-trade providers?

With overwhelming choices, consumers not only want personalization, they expect it. According to a study by Accenture Interactive, three in five consumers are more likely to make a purchase when a retailer recommends options for them based on their past purchases or preferences, and two in three are more likely to shop at a retailer that knows their purchase history.

This growing expectation that retailers should connect consumers with the content and products that they actually want is not a new concept – personalization has long been an integral part of ecommerce. For years, Amazon has been applauded for its ability to show different home pages for different customers based on their past clickstream paths or previous purchase behaviors — something they continue to do to this day. New instances of creating more intimate commerce experiences are occurring each day. This past October, eBay implemented its smart personal shopping assistant "Shopbot" on Facebook Messenger to help shoppers find what they're looking for based on their preferences and budgetary considerations.

With a growing number of implementations, it's clear personalization isn't going away. But how do retailers know if their personalization program is achieving in the results they want? Here are five of the most valuable metrics retailers should track to ensure personalization success.

Revenue per visit
Revenue per visit is essentially the average amount each customer spent for a single checkout purchase, in-store or online, over a specific amount of time. As the average revenue for a certain customer increases or decreases, retailers can start to gauge how well their personalized communications are performing. If revenue increases, retailers can identify whether or not offers are being redeemed, as well as which offers are being used more. Alternatively, if revenue for any particular customer decreases, retailers can see whether the customer has started to forgo certain offers and discern whether they need to switch up product recommendations.

Additionally, retailers can determine the overall success of their personalization efforts by comparing the average revenue for new customers to returning or loyal customers to see if the average spend increases as they gather more intelligence on individual shoppers.

Conversion rate
When analyzing a personalization program, retailers should measure how many customers redeemed deals and offers that were specifically tailored for them, and ignore conversions for any store-wide offers or sales that were pushed en masse to their entire customer base. A conversion rate assessment across a retailer's collective consumer base will give an overall indication of its personalization success. However, retailers should also analyze conversion rates on an individual level by identifying how many personalized offers each customer used when making a purchase – whether in-store, in-app or online – and whether that percentage increased or decreased over time. An individual assessment will let retailers know whether they need to change personalization tactics for that single customer.

Shopping cart abandonment
While shopping cart abandonment rates indicate lost opportunities, this metric reveals a wealth of information for retailers, and represents an opportunity for retailers to deliver relevant and timely product recommendations before customers abandon their carts.

According to a recent Forrester report on shopping cart abandonment, high shipping cost was the top reason shoppers abandon their carts. Other shoppers said they were not yet ready to purchase, were using their online shopping cart to compare similar products from different retailers or were using their shopping cart as a wish list for later consideration.

With the right context and knowledge of an individual customer's behavior, retailers can salvage these sales by presenting shoppers with discounts for the abandoned product(s), or offer relevant recommendations for similar or alternative products based on the items that were abandoned. As shopping cart abandonment becomes a less (or more) frequent occurrence, retailers can better understand how and when to push out personalized offers to make a sale.

Customer lifetime value
Customer lifetime value sheds light on the amount of profit a customer is expected to generate over the lifetime of her relationship with a single retailer. It also lets retailers know who's going to invest time, energy and money in your products and services for the long haul, and who's not. The more a customer comes back to shop at a specific retailer, the greater her lifetime value becomes.

One of the most effective ways to increase a customer's lifetime value is to tune into the data that powers personalized communications. As retailers gain a deeper understanding of what factors will spark a shopper's interest in a purchase (a certain price break, a particular type of deal, the channel they favor, etc.), they can leverage that data to create more personalized experiences for them and, ultimately, a long-lasting relationship. As a customer's individual lifetime value waxes and wanes, retailers can use this metric as a barometer for measuring their overall personalization efforts.

Sentiment analysis
One of the biggest challenges retailers face is knowing how customers will perceive their messages. Sentiment analysis — particularly social media sentiment analysis — provides the most direct insight into customers' attitudes, emotions and opinions. While sentiment analysis reveals exactly how customers are responding to a specific brand, it also discloses their thoughts about topics that are important to them. This ability is crucial to a retailer's personalization success because it gives them solid cues for how to approach that customer and whether to  modify their personalization strategy. For example, if a customer tweets about how much she loves Adidas, but also is a huge proponent of sustainable fashion, a retailer could offer her a discount on Adidas' new biodegradable shoe.

Emotional intelligence technologies such as facial detection are also a powerful tool to detect sentiment. Similar to social media sentiment analysis, which analyzes text and speech, facial detection uses semantic analysis to interpret attitudes from photos and videos. While this is still an emerging technology, facial detection gives marketers another valuable avenue to analyze customer perceptions and tailor messages that are delivered based on the insights that are gathered.

Implementing effective personalization is no easy feat. However, recent technological advances in computing power and AI machine learning algorithms now enable previously impossible amounts of data to be ingested and understood. By investing in these now very available technologies, retailers can meet their customers' demands for better personalization and monitor the right metrics to ensure the success of these strategies.

Cosmas Wong is co-founder and CEO of Grey Jean Technologies, the AI-powered personalization company that provides accurate predictions of consumer behavior.

This ad will auto-close in 10 seconds