In online commerce, more often than not, the product recommendations that shoppers are presented with can actually be quite limited, and don’t actually recommend relevant products that are going to boost a given session’s order value. Think of your own online shopping experiences – how many times have you been recommended a desk lamp to go with a pair of pants, or a scarf to go with that wall clock in your shopping cart?
When you consider how each individual consumer shops, with their own style and preferences in mind, it should come as no surprise that shoppers want to be shown items they’re truly interested in personally – and certainly not products without relevance to them. This is why ensuring that your brand’s product recommendations are relevant to shoppers is crucial, especially when considering the significant impact they can make on the shopping experience for consumers and overall metrics for brands like yours.
To make this common – yet often underutilized or not properly utilized – practice of product recommendations occur as it should for shoppers (with them actually being introduced to products they’re unconsciously looking to buy), the question is clear. How can product recommendations be approached in a way that accurately hits home for shoppers repeatedly?
Lucky for you, Lily AI is here to help with our practical guide to product recommendations in e-commerce. In this guide, we’ll take a look at the most practical steps to take as a brand in order to serve up more relevant product recommendations that are most likely to appeal to shoppers and drive key metrics forward. Plus, we’ll share results from a large multi-brand retailer who uses Lily AI product attribution to help them think like their customers. Here’s where to start expanding your product recommendation capabilities and keep from missing out on opportunities to create a better shopping experience that converts.
For brands, the impact of providing relevant product recommendations can be substantial. Let’s take a look at some of the most noteworthy metrics that jumped out at us:
Shoppers that click on a product recommendation are nearly 2x as likely to come back to the site.
37% of shoppers that happened to click on a recommendation during their first visit returned, compared to just 19% of shoppers that didn’t click on a recommendation.
Product recommendations are an important part of any e-commerce brand’s strategy. Not only can customer satisfaction be improved, but AOV can also see an increase if you’re presenting the right products at the right time. Although product recommendation engines exist for this purpose, they can easily miss the mark when they don’t take product attribution data into account. Don’t leave money on the table or limit yourself to only the basic ways brands often recommend products, like the following:
1. Displaying Products Based on Browsing History
The Pros: Honing in on a customer’s search, browsing, and purchase history to recommend relevant products (“Recommended For You”) is an often-used and relatively effective way to recommend products.
The Cons: Many consumers are beginning to have issues with this because of privacy concerns. They don’t want their personal information shared with third parties and regulations are getting stricter for this, leading to a decrease in tracking information available to retailers.
2. Featuring Only Best-Selling Products on the Homepage
The Pros: Showing your most “Popular Products,” “Customer Favorites,” or “Best Sellers” can be enticing for shoppers who always want to get their hands on the latest and trending products.
The Cons: Most shoppers are going to be visiting your site with specific products in mind to potentially purchase. Although recommending these best-selling products can potentially hook the attention of a consumer, it’s not likely to reel them in for the full hook, line, and sinker if they have no personal interest in these items.
3. Highlighting New Arrivals to the Exclusion of Other Items
The Pros: “Featured Recommendations” that highlight new arrivals can introduce shoppers to items they might not originally have thought about shopping for and can serve as a way to bring new styles to the front and center of their carts.
The Cons: Although new arrivals are typically the latest products in what’s considered fashionable and “in,” there’s no guarantee that a consumer is going to be interested in any of the new arrivals in stock and ready to be sold.
4. Promoting What Others Frequently Bought
The Pros: “Frequently Bought Together” recommendations exist for a reason. If a bundle of items are regularly bought together then there’s usually a good reason for it – i.e., athletic leggings for a woman, followed by a sports bra recommendation, would typically make sense here.
The Cons: Just because two items are “frequently bought together” doesn’t automatically imply that they’re going to be relevant for shoppers. Shopper A may love your athletic leggings, but Shopper B may hate how your sports bra fits – or doesn’t fit – both their body and their personal style.
5. Showcasing Similar or Related Products
The Pros: When a customer looks at a specific product, this can indicate that they’re generally interested in that particular product category overall. Showcasing similar or related products in that same category can potentially help to boost sales in the case they’re not completely satisfied with what’s in their cart.
The Cons: Although this can be a great opportunity to recommend similar items, it’s important to make sure that the items being recommended are actually similar. Consumers will most likely have no interest in these recommendations if they’re not accurate.
When it comes to driving better product recommendations, Lily AI comes in to play (and win) with our expansive product attribution data. With our emphasis on customer-centric product attributes and closing the gap between brand-speak and customer-speak, we allow brands to more effectively recommend products for shoppers based on style, trend, and occasion. And not to mention – help to increase AOV, AOS (average order size), UPT (units per transaction), and RPV (revenue per visit).
Just as each shopper has their own unique language for searching for products, they’re also going to have their own style preferences and they want to be shown items that actually appeal to them.
Matching consumers with style-driven recommendations, rather than merely focusing on tagging those products with generic, legacy, “out of the box” product attributes, is a great way to show shoppers that you care about their personal style and preferences. It confirms you understand them and have the ability to match them with products that actually hit the mark – even if they’re looking for a sequin blazer or eyelet top that requires a robust embellishment taxonomy.
From Barbiecore to Dark Academia, trends in fashion are continuously evolving and are increasingly being defined by what consumers are seeing on TikTok and other social platforms. While this ever-revolving trend cycle can be hard to keep up with, being able to recommend shoppers the trending products while they’re actually trending can make all the difference.
Capitalizing on what’s trending by recommending products that fit within the trend can not only increase AOV and RPV, but the overall demand for items as well.
When a consumer is met with a trend-forward shopping experience, they’re more likely to return and think of your site as the place to go to ride the latest trends. Maybe you’ll even get a special shout out on TikTok from a satisfied shopper who wants to share how others can “Complete the Look!”
With occasion-driven recommendations, retailers can keep up with what’s happening externally in the current moment by getting the most relevant products and recommendations out there internally – of course at the most likely time consumers are going to be shopping for them. From winter weddings to New Year’s Eve dresses, occasion-driven recommendations are an organic choice for brands to make – as long as the recommendations accurately line up with the upcoming occasion the shopper has in mind.
Lily AI is an easy-to-use platform built to bridge the gap between brand-speak and customer-speak and enrich your entire attributed catalog, instantly analyzing and managing product attributes. It allows brands to easily discover, edit, and inject enriched attributes to their existing workflows and commerce systems. The platform also allows brands to take action based on analytics, such as customers’ current search trends, conversion metrics, and even new trend identification with our recommendations dashboard.
This extensive product data enrichment and AI-powered product taxonomy helps bring online shoppers product recommendations that meet the moment. We ensure your customers are being presented with items that truly match the attributes that make up their personal style and preferences, giving them all the right reasons to make that second click.
We work with commerce industry leaders to ensure that the language of their customers informs and enhances every point in their commerce ecosystem, from existing site search engines to recommendation systems to SEO/SEM tools and more. It helps brands deliver better product recommendations with visually-similar matching and algorithms based on Lily AI’s 15,000+ product attribute taxonomy – built by understanding how shoppers really describe the products they’re looking for.
Lily AI’s team of domain experts bring their hands-on experience and expertise to help build and develop our taxonomy, essentially guiding our automation to “think and behave like a stylist” for each individual shopper. This team of in-house domain experts are not only committed to researching short-term and long-term trends, they also identify the relevant tags that embody the many products and styles shoppers will want to be recommended. They then collaborate with our AI-powered automation engine to pull and highlight the specific attributes that resonate, allowing brands to take advantage and inject those attributes to their relevant products to make them much easier to recommend and discover.
The challenge of matching shoppers to the products they actually want via recommendations is a constant across the e-commerce landscape. When a large, multi-brand retailer became concerned that their product recommendation engine was not able to take advantage of granular product attribute data, they knew they needed to find a better way to ensure that customers were being presented with items that truly matched their personal style and preferences.
In the initial proof-of-value test, Lily AI was able to provide more granular product information, which in turn provided more context and relevant product recommendations that could then be input into the retailer’s existing recommendation engine. This allowed the retailer to develop a more hyper-personalized product set, thereby increasing the number of relevant products shown to a customer. Lily AI’s output for the recommendations test was deployed into Certona, the company’s product recommendation engine.
KPIs that this retailer sought to measure included RPV (revenue per visit), AOS (average order size) and overall demand – all of which were hypothesized to increase with an increase in relevant products recommendations and views.
One interesting discovery during the retailer’s proof-of-value tests was that visits, orders, CVR (conversion rate) and AUR (average unit retail) all remained constant, which indicated that the basket size increase was what was driving the demand. Customers were adding more “quality” products to their shopping bags, based on seeing more relevant product recommendations.
Did you love our guide and want to save it for later, or share it with a friend? No problem! Just fill out the form and we’ll email you a PDF of the The Practical Guide to E-Commerce Product Recommendations for easy future reference!