Combine product recommendation and personalization to boost your conversion rate

E-commerce websites often offer a huge range and number of products. Brands therefore face the challenge of quickly highlighting the right products to the right consumers to make the customer journey as seamless as possible. How can you ensure that each visitor easily finds the product that they need and that best meets their expectations?

To enable website visitors to just be offered products that are relevant to their interests, brands have two invaluable and complementary methods at their disposal: product recommendation and personalization.

1 Product recommendation: highlighting complementary products from your catalogue

What is product recommendation?

Product recommendation enables you to promote products to your visitors that complement those already viewed - in other words allowing you to suggest items that are usually bought together.

This is an invaluable tool for retailers whose catalogue contains a lot of products. Recommendation algorithms enable certain kinds of products to be aggregated and then displayed to visitors.

For example, if a visitor buys a bike, the website will then suggest that they buy a helmet, lights or sports clothing.

On the Fnac website for example you see the "complementary products" slider on every product page 

​

This approach is product-centric and enables brands to logically promote other, relevant items in their catalogue. However, this is not enough to entirely address consumer needs.

How to use product recommendation?

Product recommendation is somewhat like a window display in a store. It presents a selection of items chosen by the brand and targeted at all visitors.

Grocery e-commerce players can, for example, suggest products that complement those the visitor has already added to their cart while browsing the site. It is the equivalent of positioning products in a strategic place, just as a bricks-and-mortar store would when designing its layout.

Luxury brands can use product recommendation to showcase their brand concept, clearly providing an overview of their current collection.

For marketplaces, product recommendation can enable them to display similar products from other sellers, or to show the full extent of a single seller’s catalogue.

2 Personalization: adapting the content of your catalogue to visitors’ interests

The difference between Product recommendation and Personalization

Unlike product recommendation, personalization is fully customer-centric and seeks to address visitors’ needs. The experience offered is therefore dependent on the interest of the visitor rather than products you offer.

With personalization, it is possible to contextualize all the elements of a website. For example, this can include spotlighting products via banners, pop-ins or emails.

Let’s get back to the example of our visitor who bought a bike. They return to your website. This time, they want to buy a washing machine. Depending on their behavior on the website, personalization will enable you to display domestic appliances or relevant services that complement their search at any given moment.

Kameleoon’s solution enables you to personalize visitors’ experience based on the products that interest them. A wide range of criteria are available, including: 

• types of product page visited
• a price decrease or increase on a product since the last visit
• most/least expensive product in its category
• visitors’ price sensitivity
• the color or size of a product.

The real difference between personalization and product recommendation is that we display the products that actually interest the visitor at the time of their visit and not the products that complement those they have viewed in the past. So our returning cyclist looking to buy a washing machine won’t be continually offered more cycling gear!

On the Fnac site, the homepage is personalized around the products that are of most interest to the visitor - it isn't just a selection of complementary products. 

​

Personalize in real-time with Artificial Intelligence 

AI-driven personalization, which uses predictive algorithms, enables you to identify the conversion intention or interest of a visitor (even an anonymous one) in a given product category, and to personalize their experience in real-time.

With Kameleoon, it is possible to personalize any visitor’s experience after 15 seconds on the website, based on their behavior (such as the product category viewed, price sensitivity, time of day or weather).

Kameleoon’s solution translates a visitor’s interest in a product category to actionable insight thanks to the Kameleoon Conversion Score™ (KCS), which uses a scale of 0 to 100 to indicate a visitor’s probability of converting. This can be used by the brand to decide what actions (if any) are triggered in real-time.

For example, you could reorganize the menu categories or the different parts of the home page based on the individual visitor’s KCS. So a consumer with a high KCS for domestic appliances, IT products or video games will be shown these categories as a priority on the navigation menu or above the fold on the home page.

How to use personalization 

To draw another parallel with bricks-and-mortar stores, personalization is akin to the advice of a sales assistant who will find the ideal item for a shopper based on the intentions they have gathered from chatting with them.

Brands must therefore get to know their audiences if they are to offer them experiences and products that meet their needs and expectations. Drawing on visitor data is the best way to offer them a relevant experience, something that is not possible by using recommendations based on showcasing the product catalogue.

It is then possible, depending on the preferences of each visitor, to adapt the products that are featured and also how they are presented.

With personalization, you can effectively make the experience of each and every one of your visitors unique, and not only via the products you display to them. For example, you can:

• adapt the home page or the featured product blocks according to the product categories that most interest individual visitors
• display the number of views for a product that day, or the number of items remaining in stock
• display social proof messages to highlight popular products
• offer personalized sales for certain product categories
• offer a discount coupon for a product category depending on the amount in the visitor’s cart (for example, if the amount is $50, offer them a discount coupon if they spend $60 or more)

3 Personalization and product recommendation: complementary practices

Personalization and product recommendation complement each other, just like in a store where the window display is equally as important as the advice provided by sales advisors.

Ideally, you want to be able to combine product recommendation algorithms, which help you cover all products in a sometimes very large catalogue, and personalization algorithms, to ensure that the products featured match visitors’ expectations.

Personalization enables you to boost all your marketing actions and product recommendation is one of these.

Visitors convert more when actions are more targeted while at the same time personalization let you optimize your costs by concentrating on actions with a high added value. Finally, automation through AI personalization frees up time for your teams to focus on other strategic tasks.

At scale, personalization can thus increase the efficiency of marketing expenditure by 10 to 30% and increase revenues by 5 to 10%. Combining personalization and product recommendation can truly guarantee better conversion performance and a successful client experience.