ePRICE and prudsys: unique performance for unique requirements

Sales increase by 25 % for Italy’s largest online generalist.

ePRICE Onlineshop mit Recommendations mit der prudsys RDE Recommendation Engine

Every day millions of users trawl the internet looking for enormously different products and offers. ePRICE as an all-rounder is designed to meet the needs of nearly all users and provides its customers a wide range of goods. Every year over 12 million people visit the site which offers a vast product catalogue and an enormous variety of goods. No surprise then that ePRICE has decided to rely on prudsys services to manage its online product recommendations and to make the most of the potential for personalization offered by the prudsys system.
ePRICE Srl started in 2000 and is part of the Banzai Group, a group which includes other leading names such as Saldiprivati and Giallozafferano. Banzai is one of Italy’s main online operators. Since its foundation ePRICE has consolidated its position as an e-commerce leader in Italy. Its strong points include seamless site navigation, competitive pricing and a broad ranging catalogue. The site offers over 400,000 items across the board from leading edge electronics through to inexpensive household goods. An ever-expanding catalogue with more new products and updates being added daily does however have its downside. There is a real risk for ePRICE that users will be confused by the wealth of material on offer and that the company will lose sight of the real, individual needs of each visitor.

ePRICE Onlineshop mit Recommendations mit der prudsys RDE Recommendation Engine

The objective

ePRICE had outgrown its existing recommendation system and needed a modern system in line with its current needs. Today, a recommendation system has to meet two main requirements. Firstly, it has to have the rule flexibility required to tailor a recommendation to the part of the site where it will appear, from the home page right through to the shopping trolley page. Secondly, it must be relevant and achieve the maximum level of personalization if it is to keep pace with the increasing tendency towards one-to-one relationships in e-commerce. E-commerce market leaders are increasingly implementing one-to-one.
It is of fundamental importance that a recommendation has real relevance for the user. Real, appropriate recommendations will lead to a real increase in upsales and the number of pages visited by users. The result is an all-round increase in turnover.
The central feature of the system specified by ePRICE is the ability to understand and adapt to the interests of site visitors, immediately. This is based on a knowledge of visitor behaviour patterns and interests both past and present. The self-learning function enables complete automation of the process. Self-learning enables the system to improve its knowledge of individual user needs, click by click. As the user clicks through, the system processes this information improving its understanding of the relations between products – automatically and in real time.

The solution

In 2011 ePRICE Srl decided to entrust prudsys with the system for producing the recommendations to be displayed on the ePRICE site. Within the space of a few days the self-learning system had already collected enough data on user behaviour to be able to display the first relevant recommendations. Clearly the prudsys system does not stop with the display of the first recommendations. The system implements an on-going learning process which tracks user navigation behaviour so that recommendations are updated, optimized and made more efficient day by day.

The project started with a kick-off workshop during which marketing personnel decided the business rules to be applied to recommendations. Decisions were made about the articles available, sales priorities and other factors. At the same time, ePRICE technical staff met with the prudsys team to review all the issues surrounding integration of the project. The versatility of the system and the new technical knowledge made it possible for ePRICE to exploit the full potential of the system and to implement it in all site sections, from the product info pages right through to the category landing pages. At the moment the ePRICE team has implemented approximately 30 different recommendation rules using the client software supplied by prudsys.

The result

The results have not been long coming. In 2012 ePRICE was able to display 40 million recommendations which influenced 13 % of orders. prudsys demonstrated that it is able to increase the permanence of visitors by up to 50 % on the sites where it is implemented. It does this by guiding visitor navigation. Increasing the number of pages visited and the average permanence time has also increased the conversion rate.
During the coming months ePRICE will be releasing a new version of its site with a completely redesigned home page. The prudsys presence on the site will be augmented to ensure that each customer receives a different, personalized shopping experience which takes into account his or her history of visits over the last year of navigation.

At a glance

Objective

  • Individual recommendations for each visitor
  • Increase conversion rates, revenue and residence time

Solution

  • Long tail recommendations, too
  • Real time calculation and adaption to new customer needs
  • Personalized recommendations for each individual customer thereby increasing customer retention rates
  • Very complex algorithms which are easy to manage
  • Increase in returns and sales measured by A/B testing
  • Omni-channel solution to include recommendations on all customer touch points

Result

  • +18 % click rate compared to the previous system
  • +25 % revenue by prudsys
  • +50 % residence time

Further case studies of our clients

N11 Onlineshop mit Recommendations mit der prudsys RDE Recommendation Engine
nu3 Onlineshop mit Recommendations mit der prudsys RDE Recommendation Engine
Produktempfehlungen im Newsletter, Thalia Newsletter mit Recommendation Engine prudsys RDE
holyart Onlineshop mit Recommendations mit der prudsys RDE Recommendation Engine