Essential to meeting e-commerce performance goals, product recommendations are a proven way to increase revenue, conversion rates, and average order values. On average 2 percent to 5 percent of total website revenue (and sometimes up to 20 percent) can be attributed to recommendations.
However, online merchandisers often make seemingly minor mistakes when it comes to product recommendations that can have a negative impact on the bottom line. Here are four ways to improve your integration of product recommendations that drive business results.
1. Use what you know. The sad truth is that a lot of e-commerce businesses don’t deliver relevant product recommendations. Instead, they implement machine learning and algorithms that spit out basic recommendations that aren’t relevant to the clues the customer is providing about their intent.
There is no need to go into product recommendations in the dark. Leverage relevant details about customer purchase histories to serve relevant recommendations as a model, and then take your recommendation strategy a step further by examining your customers’ browsing history to identify items of interest.
A great way to add relevance to recommendations is to tailor recommendations based on whether the customer is new or returning. For example, a new visitor who is looking for a pair of running shoes should see recommendations for the “best-selling” or “most popular” sneakers you sell. For returning visitors – where you already have information about the specific brands they’ve previously browsed or purchased – you should highlight new products you’ve added since their last visit.
2. Quantity doesn’t always mean quality. A common mistake is simply showing customers as many product recommendations as possible. More is not always better, and by testing and refining your recommendation strategy you’ll find that it’s not always prudent to offer new recommendations every step of the customer journey.
While it’s logical to deliver recommendations on product pages, the shopping cart and checkout page aren’t always the best places to show multiple product choices. Instead, look for upsell opportunities when customers add an item to their carts by recommending the top one or two items that truly complement the main item they intend to purchase.
Take a customer that’s buying a flat-screen television, for example. Showing them five recommendations for similar TVs on the checkout page doesn’t make the most sense. A better idea is to highlight the items that complement the television, such as your top-selling TV mount or the HDMI cables everyone who purchases a new TV needs. By putting more thought into the items that complement the products a customer has added to their cart, you will improve your chances of upselling and boosting the value of that order.
3. Don’t rely on black box tools. Currently there is an overreliance on “black box” or machine learning to drive recommendation strategies, but there isn’t a lot of transparency in terms of how algorithm-based tools are creating and delivering recommendations.
To combat this phenomena, focus on providing straightforward recommendations and testing what you can track. For example, try showcasing a list of the three to five most popular products that customers have recently browsed on your website and test whether new visitors ultimately add these items to their carts.
For example, in the catalogue retail environment customers are shown complete outfits – and they typically see items like a hat, earrings, shoes, and other accessories that complement the main item (a dress) on the page.
This principle is the same online. When a customer adds a dress to their cart, you should recommend a bracelet, necklace, and handbag that perfectly complement the item they selected. By recommending the top three parallel components that go with the main item, you’ll improve your ability to drive larger transactions.
4. The lost art of merchandising. Most marketing teams possess a substantial amount of internal knowledge about what products their customers are searching for. The key is to empower your merchandisers to help influence the types of products you’re recommending across the entire website experience and complementing your algorithms.
Product recommendations are a natural extension of a buyer’s or merchandiser’s skill set, so take advantage of these skills by introducing a level of manual curation into the website experience. One way to do this is by highlighting items that go with a customer’s purchase.
For example, a buyer may know of a new product line that is being introduced to the organization, and regardless of history needs to be prominently promoted. This will not always be taken into account by the machine-only process. Margin, “regionality,” and liquidation factors are all additional things that manual curation would need to introduce into the automated process. Use that internal knowledge to drive your business.
Get the Most Out of Existing Resources
The most important step toward boosting the value of your customers’ carts is to clearly outline your merchandising goals. Whereas some companies use product recommendations to increase order value by getting their customers to buy a more expensive item, others aim to raise the number of total items in their customers’ carts. It’s critical to understand what is important to your organization to help define which tactics to use and measure success.
Once you’ve established clear goals, you can take your recommendation strategy to the next level by combining your internal knowledge and expertise with technology. Getting the most out of your existing resources will help your business boost the overall value of a customer.
When measuring the effectiveness of discount codes, retailers often get it wrong. In this article, we'll look at how data-driven attribution can help businesses better understand where discount codes produce the best ROI.
Data. It’s the latest ‘buzzword’ in the digital marketing world when it comes to content.
Digital has quite forcefully overturned the entire media industry, causing even the most traditional companies to adapt or be left behind.
What’s behind a successful data-driven marketing strategy?