Imagine walking into your favourite store with the intention of buying a new dress shirt. You might find yourself in the shirts section, staring up at the overwhelming magnitude of options – where do you even start? Brand? Color? Maybe the gingham pattern looks nice but you start wondering if there are different collar options. Luckily, a sales associate quickly notices your predicament and gives you more recommendations from their selection of gingham shirts based on the shirts you’ve been looking at. In fact, a really good sales associate would not only offer you similar shirts in various colors but also a few ties that will compliment the look.
The late, great visionary leader Steve Jobs once said, “A lot of times, people don’t know what they want until you show it to them.” Customers may love your movie or product – but are they even aware it exists? In recent years, a lot of online retailers have adopted the approach of implementing a system to help customers discover new products while browsing online shops. While a critical step in the right direction, a retailer’s attempt to integrate a personalization system to improve customer experience is only the first step. Online retailers should consider deploying recommendation systems that take advantage of predictive analytics of customers’ data to improve cross-sell and upsell significantly.
Studies show that nearly half of all online stores offer only one type of recommendation strategy, thus missing the opportunity to stimulate product discovery and optimize conversion through cross-selling and upselling. For instance, American clothing brand Gant relies heavily on recommending the same product in different colors; failing to cross-sell accessories or introduce shoppers to other items based on their browsing behavior. In addition, having only one kind of recommendation strategy can result in bad recommendations if shoppers are exposed to items that they have absolutely no interest in. After a few of these recommendation “fails”, the shoppers’ trust level diminishes. They either might leave your site for good or resort to ignoring all recommendations and prefer to find their products of interest by sifting through the product line. Therefore, failure to take advantage of the recommendation system creates frustration and resentment, and affect negatively on customer loyalty. In some cases, not having a recommendations system may be better than having a bad recommendation system on your site.
Typically, online shopping experiences don’t match that of a face-to-face customer service. But with a sophisticated recommendation engine you can use analytics to cross-sell, up-sell and give a similar customer experience like that of an in-store Sales Associate. A good online recommendations system should leverage behavioral data to make at least four types of recommendations to shoppers: complementary items, items similar to what a consumer is browsing, items based on shoppers taste and behavior, and recently viewed items.
A “Related Products” bar suggests other items from the brand being browsed. It cross-sells to brand loyalists and helps increase shopping cart volume.
A “Similar Items” bar appeals to shoppers still in the research phase and help them in selecting the perfect product.
A “You May Also Like” bar produces unique suggestions of the items that the buyer is most likely considering to purchase.
A “Recently Viewed” bar archives products of interest for shoppers to revisit when they are ready to make their purchase.
The recently-launched Trouvus Recommender plugin for Magento immediately improves your online store. This free extension uses advanced machine learning algorithms to make better product recommendations to site browsers. It looks at what the user has browsed or purchased previously and compares that information against commerce data sets to predict what a shopper will most likely want to buy next. This kind of technology used to be available exclusively to the large e-commerce giants. Now that the Trouvus Recommender plugin is available in the Magento Extension Marketplace, even smaller stores can take advantage of a dynamic personalized recommendation solution.
Consumers are becoming increasingly more accepting when it comes to personalization and brand interaction, while brands are getting better at using past consumer behavior and other data to ensure that the interaction is relevant. In fact, shoppers are now demanding personalization and they expect brands to make their best efforts in providing the best customer experience possible. The onus is on the organizations to get the right recommendations to customers. Those that get personalisation right can inspire shoppers and have a positive impact on customer loyalty. For brands to produce suitable personalization, they need to understand and respond to customers needs at an individual level. A broad segmentation approach in today’s highly competitive market is no longer sufficient.