It has never been easy to recommend something to other people, including your friends and not to mention people you have never met or talked to. Just like users in your online shop! If you run an e-commerce business, you should know how important it is to recommend the right products to the right people in order to encourage them to buy more and come back to you for more in the future.

Having said earlier how difficult it is to recommend, so, what should you do to avoid mistakes when displaying your product recommendations to complete strangers? What’s more, your recommendations have to hit the right target, but how?  It is not easy at all, but doable.

Although some things may seem rather obvious, but, when it comes to the real life, most of the e-commerce businesses tend to make one major mistake: They think, they know what suits their users most without actually asking them.

1 recommendation

Nothing could be further from the truth! Of course, we are not implying that you have no knowledge about your target group. However, in e-commerce, it is quite impossible to know every individual customers. Unlike an offline business, we are unable to meet and speak to them in person and hear about what they like. Nonetheless, we still want them to be satisfied with our service and most importantly, find and BUY what they have come for or even buy MUCH MORE than what they have come for.

„I know the best what my users want, so I can identify Their needs just rights” A big mistake of e-commerse business…

Let’s go back to our recommendation strategy. You might be assuming that users searching for a bottle of wine would most probably also search for the opener. But is that always the right approach?

A little QUIZ

Are you able to guess what other products the users who have bought a basketball ball would be eager to search for?


You might be thinking it’s a pair of sport shoes. Or maybe it’s a pair of basketball shorts? Oops…but the correct answer is DEODORANT!

Would you have thought about this by merely using your own ‘human’ logic? Very unelikely….

However, there is nothing to worry about. What may sometimes seem difficult for us to understand is easy for the case of technology – technology can learn about the users’ needs, preferences and the products they would like to buy based on their activity on the site. Although you are unable to follow thousands (or even millions) of your customers on the site, technology can achieve that. Furthermore, this technology analyses all the data and creates recommendations tailored to the user’s needs.

→Thus, our belief continues: Have trust in smart technology rather than your intuition.


Product Recommendations Best Practices

Technology is a ‘must have’ if you want to display product recommendations which will ultimately lead to higher sales. However, even if you use the best tools available or try to work this out yourself, you will still need to follow some basic rules in order to achieve satisfactory results.


Firstly, it is because users are already used to certain standards throughout decades of surfing the web. Secondly, it is due to the fact that there is nothing worse than displaying a bad product recommendation or making a poor placement. That is why we have prepared a list of rules you are advised to follow when using recommendation widgets on certain pages of your online shop.

There are 3 main Recommendation types:

  1. ALTERNATIVE RECOMMENDATIONS: recommending similar products and helping the user find what he is looking for.
  2. COMPLEMENTARY RECOMMENDATIONS: they encourage customers to increase their orders and buy other complimentary products that can be useful.
  3. GENERIC RECOMMENDATIONS: they serve as inspirations; these are best sellers, top viewers, or top rated products. The important thing to remember about this type of recommendations is to apply the right logic. Many online businesses make a major mistake by assuming that they know exactly which products are complementary to one another. However, users often surprise us and make choices we would never have predicted. That is why it is important to follow the users’ logic rather than ours and take advantage of the technology to discover and understand users’ way of thinking correctly. Types of recommendations are presented in the chart below.



Recommendations standards – where and how to use them?


In most cases, your homepage is the first thing users see upon their visit. Hence, it should either be kept general or if you are displaying it to a returning user you can make it personalized by displaying recommendations based on the user’s earlier activities.

HOMEPAGE Advised Recommendation Types:

  • Personalized Products for returning users.
  • Generic Products for new users.

Position on the webpage: depends on Homepage layout. Personalized products should be shown at the bottom or bottom-right side of the webpage.

Usual headlines: „Recommended for you”, „ Products that you might like”, „Bestsellers”

It may be the company’s strategy not to display any product recommendations on their homepage. Even with a very striking design, this is not the best strategy for the online shop. Users simply love to see what the hottest items are in this season according to the buyers and the items that other users have viewed on the site.


Category Page

Category page is a great but often underestimated place to display your recommendations.

The best logic to use on category page is the Generic one (bestsellers and/or best viewers filtered to the currently viewed category). Such a logic is the safest way to bring the best selling merchandise from your store within selected category. It is also acceptable to show personalized products based on the users’ behaviour and products they were interested in on their visit in the store, also limited to selected category.

However, you have to be cautious with this logic as it might occur that you end up with no recommendations displayed at all should there be no products meeting the filters and logics applied.

Hence, in general, we would advise using the generic logic for the category page. Users will appreciate if you show them the best selling products from the category. However, try not to exaggerate, make it 3-6 bestsellers rather than 20. It is also acceptable to filter recommendations to the main category. For example, ‘iPhone cases’ under the subcategory could be found under the main/mother category ‘iPhone accessories’.


→Filter by one brand! DO NOT recommend products from just one manufacturer in category page recommendations! The users will soon find out that you are trying to mislead them and by trying to sell what is best for you. This can do you more harm than good.


CATEGORY PAGE Recommendations – main tips

Advised Recommendation types:

  • Personalized Products for returning users.
  • Top selling Products for new users.

Position on the webpage: Recommendations are usually shown on the category page either at the very top or at the very bottom of the page. When placing it on the top, before products listings, you need to make sure that the recommendations do not cover the entire screen so that users do not need to scroll down in order to see the first organic product listings.

Usual headlines: „Recommended for you”, „ Products that you might like”,

“Top selling from this category”

Product page

For e-commerce websites, the product page is the most important in terms of placement for recommendations. This is the page where the sales process speeds up and user makes the most important decisions. Therefore, it is vital to make the most out of the product page in your online store and use recommendation widgets in the best possible way. It is essential to keep recommendations on the product page personalized for returning users. Such a strategy always guarantees the best results.

→From QuarticON experience we can state that the recommendations on the product pages generate the highest sales above any other recommendations placements.



  1. Always show the complimentary products at the bottom of the product page (for 3+ column websites).Different products which are complementary to the currently viewed product –based on the history and behaviour of other customers.
  2. Always show the alternative products on the right-hand side of the product page (for 3+ column websites). Different products which are similar to the actual viewed product (based on the history and behaviour of other customers, as well as the list of other products that customer considers to buy).
  3. Do not switch placements for the two above (for 3+ column websites) The outcome might not be desirable. Users are already used to this kind of approach to see similar products on the right and complementary products at the bottom. However, if there is a 2 column website, such as Victoria Secret’s on the image below, it’s acceptable to go either way.
  4. Filter products to the main category. Consider filtering recommended products to the main product category (Electronics, Clothes, Books, etc.) to avoid recommending clothes in electronics product page.
  5. Do not present user with too many recommendation widgets
. Having more than 2 widgets might result in an adverse effect than the expected one. It might confuse your customers which will make you end up with no sales at all. More recommendations do not always mean more sales. We have seen how some of QuarticON customers changed the layout by adding 4 recommendation widgets instead of 2 which they used to have. As a result, their sales went down instantly.
  6. Use standard headers for your widgets. It is important to be creative on your website. However, it is better to stick to the standards at times. Standardized headers for recommendation widgets is one such case. So you should rather say: ‘Products that you might also like’ or ‘People who bought this also bought” rather than ‘Check this one out’ or ‘Other products in our store’. The general rule is that the headers should inform the user why we are presenting him with this specific set of products.

PRODUCT PAGE Recommendations – main tips

Advised Recommendation types:

  • Complementary Products
  • Alternative Products

Position on the webpage: Alternative products are normally shown on the right side of the product.Complementary recommendations are generally placed below the product.

Usual headlines: ‘Products that you might also like’, ‘People who bought this also bought’, ‘Chosen for you’.

Shopping Cart Page

Do not be mistaken. When the product goes to the shopping cart it is not the end of the story. Actually, the cart is a place where the users tend to spend even more money than what they have intended when they started shopping. The key is to provide them with the right recommendations.



  1. Show complementary products in the shopping cart. The shopping cart is one of the best places for cross-selling. This is a great opportunity to inspire the users, or even remind them what they could have forgotten.
  2. Use the price filter and suggest complementary products which are not too expensive. Give impetus to the rapid purchase without much to think about. Keep the recommended products within the price range of 30% of the product/products added into shopping cart.
  3. Create the ‘Add to Cart’ button for the recommended products widget. Allowing the user to add products easily without leaving the cart is not only convenient for the him, but it can also improve your sales from the shopping cart recommendations tremendously.


  1. Suggest alternative products in the shopping cart. This may kill your sales! You do not want user to change his/her mind when products are already in the cart.
  2. Recommend a product which is already added to the shopping cart even if it is different color
  3. Show products which the user has bought lately or in previous session, unless it is not food, cosmetics, or other FMCG goods. Consider using recommendation engine that will never let you make such a mistake.


Error „404” Pages

On 404, you have to explicitly inform the user about the error but why not also recommend your best sellers or top rated products? The potential of 404 page is rarely taken advantage of. With QuarticOn recommendation widget, you can direct users on the buyer’s path even from this abandoned, ‘Page not found’ area.


No Search Results Pages

An approach with a slight difference but rather similar, could be used with the ‘No results’ page.

Most of the small and medium-size online shops search engines are too distant from being perfect. To be honest, the big e-shops often deal with the same problems. ‘No search results’, which is the most common situation is when user has misspelled the name of a searched product and the search engine brings him no results at all – this is a very bad practice for the ecommerce industry! Not everyone is wealthy enough to purchase advanced search technology which is quite an expensive investment for the majority of eCommerce businesses. Why should you show a blank page to the user rather than  ‘Personalized Recommendations’? It is not only a great way to improve usability and fill the gaps of your online shop, but it also helps to increase the sales with cost-effective personalization technology. Recommendation widget on “zero search results page” is the best you can get if you do not intend to build the coolest search engine ever which is both expensive and in most cases just unnecessary, if your eCommerce business is not that big.


Thank You (Confirmation) Pages

Do not forget to place some recommendations even when the user gets as far as the ‘Thank you page’.

This is an amazing placement for up-selling. Recommendations are based on the historical data of what the user has bought lately and what he has bought just now.

It is particularly important to take advantage of the up-selling potential of the ‘Thank you page’ in the food&beverages stores and online stores selling FMCG goods.

The only condition you need to follow while considering the ‘Thank you page recommendations’ is to create the possibility of reopening the shopping cart and adding more products to the same order. Customers do have last minute thoughts on what they have forgotten to put into the cart so let’s not discourage them by insisting them to make a brand new order.



Share this post with your friends

About the author

Marcelina Przybysz

Leave a reply

Most popular

DYNAMIC VS. STATIC – what’s the difference?
18539497_l — kopia
Defining programmatic marketing and its 5 main benefits
6 Voice of Customer Tools You Should Be Using Now