Marketers clearly influence consumers by delivering relevant messages and content and making the shopping experiences more personal to them. This personalization enhances the buying experience for a prospective customer, and as a result can also accelerate sales, improve conversion rates and increase revenues. Personalization offers the buyers a great shopping experience along with surging the sales figures of the sellers. Amazon, the e-commerce retail giant, was a pioneer in these efforts, creating very effective, personalized product recommendations beginning in early 2000. Research studies also support it as a survey unearthed that the purchase behavior of 23% of the online shoppers are influenced by recommendations and reviews on social media.

What marketers think about personalized recommendations

The data analytics and marketing firm Teradata issued a report based on a survey in late 2014 of over 1,500 marketing and communications executives worldwide which found that 90% of marketers believe that individualized marketing is the trend of the future, moving “beyond segmentation to true one-to-one personalization in a real-time context.”

What online shoppers think about personalized recommendations

Surveys and studies show 86% of consumers indicate personalization plays a significant role in their purchasing decisions. For online shoppers, 45% are more likely to shop on a site that offers personalized recommendations; and 56% of online shoppers are more likely to return to a site that offers product recommendations.

96% of the people surveyed agreed that personalization helps advance customer relationships

Amazon – Pioneer of personalized recommendations

From early on, marketers have created associations between a specific product and other products for cross-selling and upselling opportunities with customers. Today, with powerful computing technologies and more and more data available and easily accessible, creating product recommendations has now become very sophisticated.

Amazon pioneered automated, real-time “item-to-item collaborative filtering” in early 2000, recommending products based on actual data of customer browsing and buying behavior, personalizing the shopping experience to each user. Amazon developed algorithms that focused on finding similar items to recommend for each purchased or rated item by a user rather than finding similar customers with the purchase and rated items overlapping those of the user.

This approach allowed an immediate response to any new customer inputs or activities, real-time results, the ability to process huge amounts of data, and still produce high-quality recommendations even for new customers with few purchases or product ratings. There has been a discussion of just how crucial this extremely effective personalized product recommendation feature has been to Amazon’s remarkable success.

Today, Amazon’s site has multiple recommendation sections on almost every page, including:

• What Other Customers are Looking at Now

• Customers Who Bought Items in Your Recent History Also Bought

• Your Recently Viewed Items and Featured Recommendations – Inspired by your browsing history

• Customers Who Bought [the chosen product]… Also Bought

The variety of common taglines is quite large, limited only by creativity: Recommended for you; More like these; Brand top sellers; Complimentary products; Category top sellers; Recently viewed; Viewed also viewed; Bought also bought; Homepage top sellers; and so on.

Obviously, the more relevant the product is to a shopper, the more likely that person will be to purchase it. Targeted product recommendations create opportunities for sellers to upsell or cross-sell, bundle products together, include related products sections, increase average order value and accelerate sales. Today, technologies have advanced, and a number of solutions are available.

By 2020 smart personalization engines used to recognize customer intent will enable digital businesses to increase their profits by up to 15%

The effectiveness of personalized product recommendations

The research firm MarketingSherpa undertook an extended study encompassing 1.5 billion shopping sessions during the second quarter of 2015 from a group of e-commerce sites using personalized product recommendations. The recommendations used a variety of different common phrases on a product page, home page, shopping cart, category page, or throughout the site. The actual product recommendations were dynamic and personalized based on visitor data, behavior, and history.

The study found that on the whole, 11.5% of the revenue (whether from more volume or higher value of products) generated in the shopping sessions was attributable to purchases from the product suggestions. The companies that used the most common “visitors who viewed this product also viewed” on the product page had the highest success, with a remarkable 68% of those companies’ revenue coming from the product recommendations.

The research firm concluded that this reflected social proof, the tendency to assume that an action is more accurate if others are doing it. The phrasing “you might also like,” another common and highly personalized product recommendation formulation based on visitors’ current and past behaviors, was also successful, correlating to 16% of that group’s revenue to the recommendations. The study also showed that the popular phrasing “customers also bought” on the cart page generated only 8% of revenues through recommendation sales.

Product recommendation engines and increased data sources

Product recommendation engines are tools that personalize a user’s experience by leveraging on-site context and browsing habits (profile, clickstream, pages viewed, prior site visits, conversions, and purchase data) even within just a single or a few visits. The engines employ real-time analytics with self-learning algorithms to refine recommendations.

Advanced data science today provides marketers access to more and more contextual data from multiple sources outside of just the user’s activities on a website, and the sources and volume of consumer data will surely continue to increase in the future. These include in-store behavior, email engagement, geographic location, device usage, customer surveys, service logs, and others. As technologies advance, information from these other data sources will be incorporated into the platforms, refining the personalization process, allowing product recommendations to be more precise and relevant to the individual consumer.

Personal product recommendations are somewhat of a sleeper for e-commerce sellers. Even though Amazon was a pioneer in this area and has made continuous improvements, it is striking that recommendations generate over a third of its product sales. Such is the power of personalization! The research discussed above shows this success is not limited to Amazon: over 11% of the revenue came from recommendations, and in some subgroups, the figure was an astounding 68%.

This information, coupled with the clear trend toward personalization of the consumer buying experience, indicates a powerful incentive for e-commerce websites to use product suggestions to personalize and improve the customer experience. Of course, there is room for even more creativity and personalization to the advantage of both the buyer and the seller. It is more efficient and enjoyable for the buyer, and for the seller, it accelerates sales and enhances customer loyalty – a win-win situation for both.