NEW
Boost Sales with our Interactive Product Recommendations App!  |TRY IT FREE
Ecommerce Product Recommendation Guide
By
Michael J. Prichard
June 25, 2021

Did you know that 63% of consumers expect a personalized shopping journey on an ecommerce store? That’s not all. About 31% of the consumers get frustrated when the site can’t recognize them as an existing customer.

And I’m not done yet. Consumers would go so far as to share their data for personalized offers and product recommendations. 

Needless to say, personalization is a fan favorite – for good reason too. Personalized product recommendations on the Shopify store help customers to easily discover products that they’re likely to buy. 

It also enables them to find new products that they didn’t know they wanted. At the end of the day, it’s all about providing a stellar experience to the customer, and product recommendations aid that at all times. 

The majority of top online stores today have implemented some form of personalization. Amazon has been able to grow to such a giant and provide an extremely personalized experience using product recommendations. You might have noticed it too – a tripod stand being recommended with a DSLR camera, a pair of socks being recommended with leather boots, and so on. 


The bottom line? It works and everyone’s on the “recommendations” train. Awesome, but what exactly is product recommendation? 

What’s product recommendation all about?

To put it simply, product recommendation is an ecommerce personalization strategy that allows a merchant to dynamically populate products on their site’s page, email, or an app based on intelligent data such as customer attributes, browsing behavior, situation-based context, and/or purchase history. 

For large Shopify stores with thousands of products, personalized product recommendations help shoppers find the right products and save them countless hours of scrolling. 

For small online stores with a handful of SKUs, shoppers don’t just drop off after perusing through a page. They keep browsing the site, and maybe add more products to the cart if they are served with contextual & relevant recommendations (i.e., relevant products) that fit their needs. 


“We are seeing customer engagement numbers rising (upwards of 119%) and conversions growing (upwards of 165%) using product AI-based recommendation quiz.”
- Michael, CEO at Skafos.ai


Needless to say, brands can benefit exponentially if they apply the right product recommendation system. 

The impact of product recommendations

Product recommendation helps your site’s visitors navigate to the right products based on contextual data like browsing history, purchase history, time on page, etc. It’s helpful but without quantifiable data, it’s hard to see its efficacy on the growth of a Shopify store. 

Fortunately, I gathered some stats from the corners of the internet. Let’s check ‘em out:

  1. On AOV: When shoppers engage with even a single recommendation, the average order value increases from $44 to $162, i.e., a 369% jump. 
  2. On Revenue: On average, the revenue attributed to product recommendations is 12% and this number can go up to 31%.
  3. On Conversion Rates: A research found that shoppers that clicked on recommendations are 4.5x more likely to add items to the cart, and 4.5x more likely to complete their purchase.
  4. On cart abandonment: It was found that implementing personalized product recommendations can improve cart abandonment by up to 4.35%.  

By now, it should be clear to you the impact of product recommendations. But how do they work in reality? What’s the secret sauce?

How do product recommendation systems work?

While you can simply use the [recommendations] object to automatically generate a list of related products, you would need a holistic product recommendation engine to deliver the best experience. 


There are three basic approaches to craft a product recommendation system for your Shopify store:

  1. Content-based filtering: This method analyzes customer data based on the likes and dislikes of each shopper – cookies aid in tracking the same over multiple visits.  The engine takes this analyzed data and programmatically makes recommendations based on the browsing history of each user.  It’s much like Netflix – when you like a particular genre of movies (say thrillers), their system accounts for that, analyzes that data, and then recommends you other shows or movies that are identical to your choice. 
  1. Collaborative filtering: This method collects data from shoppers who have purchased similar products in the past, and then accumulates & analyzes that information to provide recommendations. In other words, you will see recommendations based on the browsing behavior and purchase history of other shoppers. For example, if a shopper visits a product page, they may be shown a recommendation section titled “Customers Who Viewed This [ProductName] Also Bought” where similar products are displayed that might lead to additional sales.
  1. Hybrid filtering: The hybrid method, as the name suggests, is a combination of content & collaborative filtering. The logic behind the hybrid method is that it incorporates group decisions but focuses on the result based on individual shopper’s attributes.  Spotify aces it with their “Discover Weekly” which is a complex hybrid system combining data from your listening habits as well as users who listen to similar songs. 
  1. Intent filtering: This is a special type of AI-based filtering where the shopper is shown different products and is asked to like/dislike them. Based on their choices, an extremely personalized set of products is displayed as the final result.  This type of filtering is especially helpful for sites that have hundreds of products on the catalog. Needless to say, this (intent filtering) saves time for the shopper to sift through those products and find the right one; just like an in-store sales associate. 


Skafos is capable of delivering such personalization for a Shopify store that comes with a large catalog of products. 

The five different styles of product recommendations

Now that we are aware of how product recommendation engines work, let’s look at the frontend side of things – how they’re displayed (aka styles of product recommendation).

1) Inspired by your [X history]

This is a special kind of recommendation engine used by ecommerce giants like Amazon to recommend products. They have a few different variations of “inspired by your ____” pasted across the site. I’ve found a few:

  1. Inspired by your purchases: This recommends a set of items that match your purchase history. 


  1. Inspired by your wishlist: As the name suggests, this displays products based on what you have wishlisted on Amazon. 



  1. Inspired by your shopping trends in [category]: As and when you browse through a category on Amazon and make purchases from that category, Amazon is capable of displaying similar products based on that data. 



There could be more by these are the ones that I found. Oh, and did you know that 35% of Amazon’s revenue is directly attributable to its robust recommendation engine? 

2) People also bought

This is a classic tactic that takes the best of both – recommendation & bundling. If you’re purchasing a condenser microphone for your podcast, you’d also want to get a boom stand to set up your mic in the right position. 

Best Buy does it well where all their product pages come with a “People also bought” section and the results are surprisingly accurate to entice a buyer to purchase more, thus increasing the basket size. 


3) Bought together (aka product bundles)

Much like its cousin “People also bought”, Bought together with is a similar technique to increase AOV by displaying complementary products with the one you’re about to buy. 

BestSelf, an e-commerce store selling physical productivity stuff is great at this. For example, if you add the Gratitude Journal to the cart, it recommends buying the Impact Deck during checkout. This is also a form of cross sell that increases basket size significantly.


4) Featured products

Want to showcase a specific set of products? Add them to a “Feature Products” widget and place it on your homepage. Featured products are more likely to draw attention as they are positioned to be better than the rest. 

Bambu Earth does it pretty well. They have added a section called “Spring Featured Products” which adds a feather to the cap by making it even more personalized, simply by adding the word “Spring”.


5) Bestseller

This is a common tactic used by Shopify stores across the globe to increase sales and grow their revenue. If you haven’t been living under a rock, you’d notice that the “NYT Bestseller” tag to books is a matter of prestige and social proof. It also fetches more sales than the thousands of other books in the same category. 

Well, bestsellers on ecommerce sites are pretty similar – the ecommerce merchant wants to push their best products out because they get more sales. Hence, adding a tag on a product or creating a whole new section is logical. 

The folks at Vahdam Tea are experts at it. Owing to their large number of SKUs, they have a dedicated page for Bestsellers. This makes it easy for folks to find that out from the simple navigation menu on the left.


6) Quiz-based

Last but not the least, there are product recommendation quiz engines that leverage product data and help you discover exactly what you need. 

There are different variations to this. For example, Bose assists you to select the best headphones by asking a series of simple questions to accurately determine your needs. 



And there are stores like Lights Online that rely on Shopify apps like Skafos to build an engine that will recommend you a set of products based on what you like and don’t like. 



There are a bunch of others that you can add but the above are the commonly used ones by Shopify stores across the globe. But what’s the strategy to implement them?

3 best practices for using product recommendations to grow your sales

You can’t just wing it when it comes to personalized product recommendation widgets on your website. Too many and you risk spamming your shoppers, too little and you might lose out on sales. So, here are a few best practices that you can follow to ensure a well-oiled, efficient, and ROI-focused recommendation engine. 

1) Show “New Arrival” and “bestseller” widgets on your homepage

You’d want to show your best and most recent products to your new shoppers and you want to show it fast before they drop off. 

That’s why it is always advisable to place New Arrivals, Bestsellers and Most Recommended sections right on the homepage itself. Simply doing this can dramatically increase your chances of getting a conversion in the first go.

2) Display “Frequently Bought Together” & “Related Items” on Shopify product pages

PDPs are where your shoppers are most engaged – they’re just over the fence, checking out your product before hitting the [Add to Cart] button. This is where you can leverage upsell and cross-sell by showing them recommendation widgets like “Frequently Bought Together” and “Related Items”. 

3) Add “Frequently Bought Together” on the cart page

If I were to pick one page with the highest engagement, the cart page always takes the cake. If the shopper has added the product to the cart, their intent to buy is the highest you can expect. You can capitalize on this valuable opportunity by displaying complementary products to the one that’s already in the cart. If all goes well, your average order value increases then and there. 

Ready to accelerate sales with product recommendations?

With endless options in the ecommerce industry, competition is at its peak and the only way to differentiate is through a better customer experience on your site. 

To ensure that you stand out, you need to personalize an individual shopper's journey. Good news is that you can effortlessly enable a good part of that through product recommendations.

I hope this guide helped you understand the basics of this vast concept. If you need more help, feel free to drop an email at marketing@skafos.ai.

Till then, here's to more sales in your store! 

Featured photo by bruce mars on Unsplash