Four Product Recommendations Mistakes and How to avoid Them

By Paul Terry

December 18, 2020


Including product recommendations on your retail site seems like a no-brainer for increasing conversions and customer satisfaction, right? Not exactly. It’s true, many brands see huge upticks in their key metrics by implementing and A/B testing some product recommendations. And overall, if you can implement them you should. But there are a few major mistakes I see brands making that you want to avoid. Here are the top 4, and how to remedy them.

1. Putting the wrong recommendations in the cart.

The problem with product recommendations in the cart is that they have the potential to distract users, pull them away from the checkout process, and lead to higher rates of cart abandonment. This is especially true when those product recommendations are for things like popular or trending items, or similar items to what is already in the user’s cart. The cart is not the time to start comparison shopping — this just introduces more confusion and opportunity for the user to leave.

Instead, product recommendations in the cart should be reserved for complementary items, upsells and add ons, or buy again items. As a bonus, you may want to optimize the add to cart process. Can a user click on an item and add them to the cart without actually leaving the page? This streamlines the process and increases the probability that they will, in fact, check out with those items.

2. Only using product recommendations on product detail pages.

It certainly makes sense that product recommendations should be targeted based on the other products a user is looking at. For example, on a product detail page for a sweater, you could recommend similar sweaters, accessories, or outfit ideas. However, this only allows you to target customers who have already begun to find what they are looking for, and you miss out on users who may be browsing with less direction.

Instead, keep those product recommendations on your product detail page, but add some on your homepage as well. These could be personalized, featuring complementary items for past purchases or items matching the user profile. For users you don’t have data on, you could feature popular items or seasonal items. This will help users start browsing, spend less time navigating through pages they’re not interested in, and give them shopping ideas.

3. Not A/B Testing Your Product Recommendations.

Just because product recommendations are a best practice doesn’t mean you can set them and forget them. The wrong algorithm delivered on the wrong page at the wrong time could actually cause additional friction for shoppers rather than reducing it.

Instead, you should be A/B testing your product recommendations placement and algorithms and measuring conversions. And, you shouldn’t focus only on purchases. Make sure you track metrics like total product detail pages viewed, searches, total adds to cart, items added to cart from product recommendations, average order value, and clicks on product recommendations.

4. Not including recommendations on your search results page.

Search results pages are often overlooked, but with some attention they can make or break the customer experience. Customers who perform searches are giving you clear insight into what they are looking for and what they’re interested in. This is especially valuable for new customers, about whom you don’t have any historical data.

Instead of neglecting your search results page, include some product recommendations based on the item searched. Especially when a search yields no results, you’ll want that product recommendations module to redirect the customer and salvage the experience. Consider displaying similar or complementary items, and even popular items when the search yields no results.

To learn more about SiteSpect, visit our website.


Paul Terry

Paul Terry

Paul Terry is a SiteSpect Consultant in Customer Support, guiding SiteSpect users on the road to optimization. He has over 15 years experience in optimization, testing, and personalization. He is based in Duluth, Georgia.

Suggested Posts

Subscribe to our blog: