E-commerce quick-win

Action list for the A/B test to measure the revenue uplift using the one-to-one recommendation in real time in eCommerce.

1. Introduction

Pleasepoint is the AI-based software to personalize interactions and increase the Customer Lifetime Value of all customers. We provide all the tools to personalize in real time and implement a predictive marketing strategy based on quick-wins.

We provide a +57% average revenue uplift and a +53% average orders uplift.

In this document you will find all the necessary information to prepare the A/B test in 5 days.

The goal of the test is to measure the revenue uplift and the orders uplift by comparing the conversion between a manual selection of products (or bestsellers) and a hyper-personalized selection with one-to-one product recommendation.

These are the results for this kind of A/B test:

A: Bestsellers products selection

B: One-to-one product recommendation

Uplift results

The hyper-personalized product recommendation using the one-to-one engine has delivered these results:

+92% interactions uplift with the product module.

+ 78% orders uplift compared with the bestsellers selection.

+ 59% revenue uplift compared with the bestsellers selection.

The higher the interaction with the products module, the greater the impact on revenue growth.

2. Test A/B: Revenue uplift

The revenue uplift A/B test consists of measuring the results of one or more product modules located on one or different eCommerce pages.

First of all you will need to decide which product component you will use for the test.

Next you have to define the A/B percentages on the traffic. For a more robust results test use 50%-50% for each case.

3. Roadmap

We have planned the task in order to prepare for the A/B test within 5 days. These are the steps that we will follow.

• Day 1: Data extraction.

  1. Extract all the historic data (CSV files) from your eCommerce. Follow the Data extraction documentation.

• Day 2: Import data.

  1. CSV files validation before being imported.

  2. Import the CSV files into Pleasepoint. Follow the Datalake documentation to import the CSV files.

  3. Google Analytics tracking workshop.

• Days 3-4: One-to-one integration.

  1. Use the getRecommendations API call for real-time one-to-one recommendations.

     • Read the Real-time one-to-one recommendation user manual.

     • Read the getRecommendations API call documentation.

  2. Use the registerEvent API call to import new interaction data in real time. Use the one-to-one user manual to understand how-to register the user interactions.

     • Read the registerEvent API call documentation.

  3. Business rules definition workshop.

• Day 5: Start the test.