Today, our Inboxes, social media profiles and messengers are flooded with commercial content from countless companies. With information noise growing, companies are looking for new ideas on how to sell a product and stand out against competitors, reach the audience and increase sales.
The secret may be more precise products segmentation and personalization of the shopping experience. Brands that approach it with both creativity and technical compliance have every chance to succeed.
Research support this assumption:
- According to numerous surveys and tests, 45% of users are more likely to buy on the website that offers personalized products, and 56% of users are more likely to return to the website with product recommendations.
- 88% of 300 marketers surveyed by Evergage reported that their customers expect online stores to provide personalized product recommendations.
Think for yourself, what would impress you more: offers based on your shopping history or random promoted goods. The answer is obvious. What’s more, if you doubt the purchase, you'll probably take a look at the block You may also like or Similar products to see alternatives to your selection.
With the above said, you can conclude that customers like to receive product recommendations based on their interests and needs because it saves their time and simplifies the buying process by showing relevant and personalized offers from the start.
However, such recommendations aren’t just a few similar items displayed in the footer. To increase purchase probability, it’s important to properly build recommendations based on the collected user data such as preferred price, brand, most often viewed items, etc. Think about that if you want to start an online store.
Recommendation blocks help achieve different goals:
- increased conversion. The visitor can buy the recommended product instead of or in addition to the one they are browsing for, and is less likely to leave the site without a purchase;
- increased average session duration and depth. The visitor views more pages per session and spends more time on the website which contributes to the ranking by search engines;
- internal page linking;
- simplified navigation. It is easier for the customer to find related or similar products in store catalogs;
- trust. The visitors feel like their interests are taken into account;
- soft promotion of diverse product categories;
- increased average check with cross-sell and upsell. (recommendations of more expensive products and alternatives for items that are not in stock);
- increased purchase frequency.
Let's take a look at how eCommerce stores can implement product recommendations and what methods can help improve sales of a product.
There are several ways to configure personal recommendations on the website:
- A tempting but complex one is to write an algorithm on your own. You know your customers better than anyone else and understand the decision-making process from the inside. But such an approach requires time, money and human resources – professional programmers who can do quality work fast.
Pros: you create the algorithm based on your needs.
Cons: cost, time-consuming process, the result may not meet expectations.
- A simple but not the most effective one is to sell on the marketplace that already has recommendations configured.
Pros: fast implementation, simple process.
Cons: recommendations also feature offers by competitors. It’s an additional promotional channel rather than a ready solution.
- An optimal and safe one is to use a specialized service/product recommendation system. For example, our CDP has recently implemented Product Recommendations for Website which you can already use to offer your audience the products they will love.
Pros: quick script installation, machine learning-supported recommendations, flexible pricing. You can also combine website recommendations with personal recommendations in emails.
Cons: clients don’t have access to the algorithm.
Website recommendations are built based on the user's personal info:
- contact data (e.g. location);
- website behavior;
- campaign activity;
- products and categories viewed during a current session or previous session;
- order history;
- channel that generated clicks (Email, Mobile Push, Web Push, social media, contextual advertising, etc.);
- social media profile.
Blocks with recommended products can be placed on following pages:
- product category page;
- personal account;
- pages of search that reply to the request.
- product page;
- page 404;
- out of stock.
Recommendation placement depends on the website's algorithms. If you use a specialized service, you can select the placement on your own, for example:
- at the top of the main page: the best place for best sellers;
- at the bottom of the page under the catalog: the most common placement for similar and also bought items;
- at the bottom of the cart page: so as not to interrupt the purchase of the already selected product.
Product recommendations can be divided into two main types – bulk and personalized.
Bulk recommendations are formed without considering user behavior and preferences. They include blocks with arrivals, best sellers, or sales. They are often placed on the main page or combined with personalized recommendations on the product page.
Personalized recommendations are formed based on the user's actions and preferences: previous orders, page views, other personal data, products properties and purpose. They can prompt the customer to make a decision and help choose the relevant product. The most common blocks are Recently Viewed Items and You May Also Like:
To get a better idea, let's see how major eCommerce sites place blocks with product suggestions.
The main page often displays best sellers from all categories. For example, Estee Lauder shows best ranked products straight after seasonal exclusives.
People often visit the online store without a specific purpose, so it is important to tell them about best sellers and promotions. Inkbox shows four different blocks with recommendations – Best Sellers, What’s New, Bundles, Finger Packs – to direct the visitor to the right one from the start.
Such a block may feature viewed products, best sellers, new products, and products that may be of interest to the visitor (e.g. accessories to the currently displayed items). You may use one block type (for example, viewed products or popular products) so as not to overload the page. Banana Republic does so by demonstrating first only discounted items.
Onzie opts to show Best Sellers and New.
Such blocks can have numerous designs and appearances. Companies come up with different block names and layout to catch the attention. The number of blocks may also differ.
You May Also Like + unusual vertical placement on the right
If you like this, you’ll love
People also viewed
Three blocks of recommendations by J.Crew that are easy to switch between.
Two blocks by Pilgrim: Try One Of These Instead + Recently Viewed.
Food52 also gives two recommendation blocks although one of them is disguised under maker introduction.
Aerie offers to complement the purchase with the most relative item.
Some marketers think that any product can be included in cart recommendations. But we’d recommend placing products carefully so as not to distract the customer from the buying process. Add items that can complement the selected product rather than push the visitor to browse for more. That’s how MeUndies does it.
Here you can show recommendations based on previous search, personalized offers, promotions or sales.
When creating product recommendations, keep in mind the following:
- offer relevant items;
- don’t combine unrelated products or services in order to draw attention to the worst-performing products;
- don’t show products where they can distract from the purchase;
- ensure that recommended products are available and prices are up to date. Otherwise, this block would generate confusion rather than additional sales.
Below, there several examples of how not to manage your recommendations.
When you show in cart recommendations the item similar to the one already added (and with a similar price), chances are good the visitor would abandon this cart and will browse the recommended item. However, they may not add it to cart and go looking elsewhere.
Your cart recommendations should complement the choice not encourage to change or replace it.
There is no need to repeatedly show the item that has just been added to cart. Show similar or complementary products that can go in hand with the current item.
Now you can create personalized product recommendations in your account.
Note. To use Recommendations for Website, you need to set up web tracking and subscribe to the corresponding pricing plan.
Log in to the system, go to Site > Recommendations, click New recommendation and proceed with settings.
Here you can
- choose the page type for your recommendations;
- set the data source based on which recommendations will be built (visitor’s purchases and views, category, description, and name of the product);
- configure appearance and placement. You can have several placements depending on the recommendation type.
You can also opt for a bulk algorithm. This option is used when there is not enough data for personalization. By default, bestsellers of all categories will be displayed.
You can do all the settings on your own as every step is provided with explicit guidance.
In our system, you’ll also be able to track performance of your recommendations with the help of all-in-one reports that would include:
- analytics on views, clicks and purchases;
- activity dashboard;
- change history.
Personalized product recommendations have long become a kind of rule of good tone for eCommerce. This tool allows you to implement the WIN-WIN strategy: customers see and buy exactly what they’re interested in, and the company sells more and gets a high profit.
Updated 3 days ago