Two visitors land on your store. One just bought a candle from you last month. The other has never heard of you. Both see the exact same product page, the exact same “You might also like” section, the exact same homepage.<\/p>\n
That is the problem. Your website treats every visitor the same, whether they are a first-time browser or a repeat customer who already knows your brand.<\/p>\n
AI product recommendations fix this. They look at what someone has browsed, what they have bought, and what similar customers ended up buying, and they show each visitor different products based on what is most likely to convert them.<\/p>\n
And you do not need a developer to set it up.<\/p>\n
You already know that showing related products increases average order value. “Frequently bought together” sections, “Customers also bought” carousels, and manual product recommendations all work.<\/p>\n
But they have a ceiling. Manual recommendations show the same products to every visitor. They do not account for browsing history, purchase history, or what similar customers actually bought together.<\/p>\n
AI recommendations learn from actual behaviour. They see that someone who bought your lavender candle also bought the reed diffuser 68% of the time. They see that first-time visitors who browse gift sets convert at a higher rate when shown your best-selling bundle. They adjust in real time.<\/p>\n
The result is not just higher average order value. It is also higher conversion rate, because each visitor sees products that feel personally relevant instead of generic suggestions.<\/p>\n
This is the simplest recommendation type. It looks at what someone has viewed and suggests similar products. If someone spent time on your candle collection, it shows them more candles or complementary products like holders and matches.<\/p>\n
This works because it reduces the effort of finding something else to buy. The visitor does not have to navigate back through your menu. The next product comes to them.<\/p>\n
This looks at actual purchase data. Not what people browsed, but what they actually bought in the same order. If customers who buy your soy wax candles also buy your wick trimmer 4 out of 10 times, that pair gets recommended.<\/p>\n
This is the most powerful type for increasing average order value because it targets people who are already buying. A “bought together” suggestion appears on the product page or in the cart, right when someone has their wallet open.<\/p>\n
This recommendation targets repeat customers. If someone bought your monthly candle subscription three months ago, it suggests they reorder. If a product they viewed came back in stock, it tells them.<\/p>\n
This type is the most underused by small businesses, but it has the highest conversion rate because it targets people who have already demonstrated purchase intent.<\/p>\n
Shopify has built-in related product recommendations on most themes. Go to Online Store > Themes > Customize, and look for the “Related products” or “Product recommendations” section on your product page template.<\/p>\n
For smarter recommendations, the Shopify App Store has several options. Look for apps that use machine learning rather than manual rules. The best ones require zero configuration and start learning from your store data immediately.<\/p>\n
Key features to look for:<\/p>\n
Install a recommendations plugin that integrates with your product catalog. The best ones use your order history to build “frequently bought together” suggestions automatically.<\/p>\n
Set up three placements: below the product description, in the cart, and in post-purchase emails. These three positions catch people at different decision points and increase add-to-cart rates by 15-25% on average.<\/p>\n
Squarespace does not have native AI recommendations. Use a third-party tool that injects recommendation widgets via a code snippet. Most tools provide a single line of JavaScript you paste into your site header, and the recommendations appear automatically based on browsing and purchase behaviour.<\/p>\n
Below the add-to-cart button, show three to six related products. This is where “browsing-based” recommendations perform best because the visitor is already in shopping mode.<\/p>\n
Before checkout, show “frequently bought together” suggestions. This is where “purchase pattern-based” recommendations perform best because the visitor has already committed to buying something. Adding a second item feels like a small extra step rather than a new decision.<\/p>\n
After someone buys, send a confirmation email that includes three products they might want next time. “History-based” recommendations work best here because you know what they just bought and can suggest logical follow-ups or replenishment items.<\/p>\n
After setting up AI recommendations, watch these metrics for two weeks:<\/p>\n