Online marketers can choose from many tools to provide customers who are shopping on your ecommerce site with relevant products recommendations. For example, on Amazon, customers are shown items “inspired by [their] shopping trends,” products within the genre in which they typically shop, “recently viewed items,” “featured recommendations,” “customers who bought items in your recent history also bought. . .” and the list keeps going.
And, Amazon certainly isn’t alone in this. Ecommerce sites across the Web have embraced these tactics to increase sales for some time.
In my opinion, traditional algorithms like “bought this, bought that” and “viewed this, bought that” are getting old. And, although Amazon has branched out from this strategy a bit with different wording, it’s the same idea. Consider all the behavioral data from the customer—as well as the customer’s own past purchase history and customer details from loyalty programs and customer relationship management (CRM) systems. But, even when you can leverage all this rich data, you still need to test your assumptions to optimize the results to improve your conversion and, most importantly, average order value.
Most importantly, however, is where these recommendations are placed and what they look like. We have to move beyond the skyscraper-type banner ad or horizontal banner ad that includes recommendations. Instead, we must intersperse the recommended product into our message, storyline, and overall customer journey.
Are your product recommendations working as well as they should be across this entire journey? In a recent article, my colleague Gina Casagrande discusses various ways automated, real-time recommendations can make a large impact on a customer’s purchase decisions. Gina gives insight into some of the important capabilities to seek out when choosing a recommendation tool. I’d like to add to her insights and discuss how online marketers can institute some of these tools across the customer life cycle by joining a customer in his thought process and producing relevant product recommendations and content (ads) across his journey.
For example, if we have a customer journey that shows someone is searching for a place to go hiking, how do we merchandise products and recommend the right gear for the hike?
First, we must embrace programs that allow us to make multichannel recommendations across devices. In today’s multichannel world, online marketers can’t afford to leave out mobile or tablet users, for example.
Second, we must utilize helpful tools that offer full marketer control. We should be able to modify recommendation layouts at any time, based on the information we obtain about our customers through A/B testing. Testing, as I mentioned earlier, is a crucial part of a successful product recommendation strategy, as data can only be trusted when it has been thoroughly verified as helpful and relevant. Auto-optimization delivery of ads to the appropriate customers and data sourcing are also important parts of the process.
Google remarketing lists for search ads (RSLAs) can play a big part for customers who have already visited your site and looked at particular products, as you can easily retarget them. For new customers, however, it’s about doing the research behind which types of customers are searching for particular products (based on their shopping personas). Grounded in this data, keyword patterns become apparent for these different shopping personas, and it’s easier to serve the appropriate products and product ads to customers who are likely to convert.
It all comes back to a unique experience for the customer. With the amount of customer data available online today, it’s easier than ever to determine which buyer personas are interested in which products and make the most relevant suggestions. It’s time to go beyond “viewed this, bought that.”
Darin Archer is head of product strategy and marketing for Adobe Commerce.