Service Background

Recommendations that sell: helped GoodWine increase conversion and sales

Tags:
Business Intelligence (BI)Custom Software DevelopmentData EngineeringWeb application

About the project

Personalised advice and recommendations turned buyers into regular customers

The GoodWine brand has long been synonymous with a huge selection of quality alcohol and healthy products. The company is the largest importer and distributor of well-known brands. The store offers over 5,000 types of wine and spirits, 300+ types of craft beer, and over 8,000 items available for delivery.

Client challenges and needs:

GoodWine approached us with a need to improve the efficiency of its mobile application and increase sales. Our team had to overcome the following factors:

  • the app did not provide enough personalised recommendations. This, in turn, reduced conversion and average check. Of course, user loyalty was lost;
  • the lack of prompt promotion of products with a short shelf life, which significantly increased marketing costs.

There was an urgent need for a system that would analyse user behaviour and purchase history. And then, in real time, generate accurate offers.

How we solved the client's problem

The GoodWine mobile app has become an intelligent shopping assistant.

  • We successfully implemented a strategy that significantly increased the average purchase amount and improved product updates.

  • Our approach:

    At the initial stage, the customer suggested using a ready-made recommendation service from AWS. It is integrated with the transaction history of purchases. However, we discovered limitations in the model during testing. The fact is that it did not take into account the remaining goods in stock.

    These challenges became the basis for our work. We created a customised tool that combined all recommendations and took into account actual stock levels. This mechanism has flexible settings. They help the system work effectively and also correspond to the specifics of GoodWine’s business.

Work results:

  • The result of our interaction with the customer was the creation of our own personalised model. The system processes behavioural and logistical factors. It also covers three key scenarios: catalogue recommendations, selection of similar products, and cross-selling.

  • Not much time has passed since the model was implemented in the GoodWine mobile app, but the indicators have improved significantly:

    • the average check increased by 5%;
    • inventory turnover increased by 10%;
    • the average session duration decreased by 10%;
    • repeat purchases increased by 5%.

Key facts about the project

12 months

Project duration

Medium

Project size:

Project Detail

Project complexity

Completed

Project status:

Our team: Project manager Android developer iOS Developer Data Science engineer DevOps
Service Background

Similar cases

Contact Background

Contact Us

Looking for a reliable technical partner? We are ready to join your project at any stage – from idea to launch.

Information icon

Information

Fill out the form, and we will contact you to discuss the project.

Address icon

Address

Senator Business Center, 32/2 Knyaziv Ostrozkykh St. Kyiv, Ukraine, 01010

Contact us

Add anything that will help us get started faster.

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.