Website product recommendations result in 11.7% revenue lift for Hoff online store
The homepage is the first place where a user decides to stay and carry on with their search or leave the website. Visitors will only continue their journey through the website if they see something that interests them, so it is crucial to show the right products.
Hoff is one of the big and fastest-growing companies. The brand first appeared as a result of rebranding the Australian brand “KIKA” in 2007. Since then, the company has grown by leaps and bounds and now has 33 hypermarkets throughout the country. In September 2011, Hoff launched their online store, which now reaches over 4,7 million users per month.
The online store offers a complete interior design solution based on thousands of different multi-brand products. Visitors to the store can find virtually any product that they may need for their home. Hoff was looking for a solution to help make the customer’s product search fast and complete, and to significantly increase their online revenue.
Retail Rocket helped the company to implement a personal product recommendations project to their website in May 2016.
Analysing the efficiency of product recommendations at Hoff
The process of optimizing the recommendation system for the Hoff.ru online store started with an effectiveness study of various recommendation algorithms on the website’s Homepage.
Hoff wanted to focus more strongly on the “Home Products” segment, and particularly on those products that were advertised on TV.
To achieve this, personalized recommendation blocks were displayed on the homepage. Retail Rocket’s algorithm analyses visitor behavior and adjusts the display in real time so that each customer sees a personalized version of the online store. This helps to highlight the products which they are most likely to buy, based on their individual preferences and purchase history.
An efficiency study of the Retail Rocket mechanism was conducted using A/B testing. Website visitors were randomly divided into two groups:
1. A group that were shown the personalized version of the homepage. Three blocks were displayed (from top to bottom): personalized bestsellers based on the visitor’s short-term interests, popular products from the categories that the visitor is interested in, and personalized bestsellers from the category “Home Products”.
2. The control group. These visitors were shown two recommendation blocks (from top to bottom): non-personalized recommendations of popular products, and popular products from the category “Home products”.
Efficiency analysis results
Over 112,000 website visitors were included in the efficiency analysis. This is a 17.8% share of all of the traffic to the website. A history of the user’s interests and recent browsing available was required for inclusion in the A/B test.
The test showed the following results:
Conversion Rate | AOV | Revenue | |
Control group | – | – | – |
Personalized popular products based on the short-term interests of the visitors (at the top), popular products from the categories users are interested in (in the middle) and personalized best-sellers from the category “Products for Home”(at the bottom) |
+5.6% |
+5.8% |
+11.7% |
The results show that using personalized recommendation on Hoff.ru’s homepage increases conversion rate by 5.6% and average order value by 5.8%. The combination of these two increases leads to a revenue increase of 11.7%, with a statistical significance of more than 90%.
Comments from Hoff
As well as using online channels to attract traffic to the website, we actively employ offline channels such as TV advertisement, promotions in our bricks & mortar stores etc. These methods help to increase online visitors.
It is very important to us to communicate effectively with visitors to our homepage using both online and offline marketing activities, and we do not want to miss the opportunity to make use of the data we have about the interests of our regular users.
Integrating personalized product recommendations on the homepage is just one of the stages of our project to personalize our website together with Retail Rocket. We plan to squeeze the maximum benefit out of the Retail Rocket platform.
Dmitriy Dvoretsky, eCommerce director of Hoff