Skip to content

Personalized tasks & achievements based on customer retail transaction data

Notifications You must be signed in to change notification settings

ilya16/pepe-lenta-quest

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

13 Commits
 
 
 
 
 
 

Repository files navigation

Lenta Quest - personalized tasks & achievements based on customer retail transaction data

Solution produced during "Lenta Hackathon" as part of the "Hack Lab" course in Skotech, September 18-20, 2020

Team PEPEtoners: Ilya Borovik, Vladislav Kniazev, Vanessa Skliarova, Bulat Khabibullin, and Zakhar Pichugin

Solution and demonstration: Open In Colab

Problem statement

Current system of personalized recommendations in Lenta app:

  • Complicated system of special offers: personal offers in the app + general offers in SMS + surprize cards in the stores
  • No possibility to affect on special offers

Concept

The solution is a symbiosis of two pillars:

  • Personalized tasks based on the history of transactions for engaging customers:
    • tasks based on previously bought and novel products
    • tasks in different product categories
  • A system of achievements with rewards and bonuses:
    • category based
    • product based
    • location based

Task personalization is provided by a multilayer recommendation system working with the following levels of item granularity in transaction data:

  • materials (products)
  • hier_level_2 categories
  • hier_level_3 categories
  • hier_level_4 categories
  • vendors

Recommendation system

Each model is a Collaborative Filtering for implicit feedback datasets based on Alternating Least Squares.

Using multiple models we can offer personalized recommendations with various levels of product description: from raw products and their properties to vendors and categories.

See demo for more details. The business side of the solution is shortly presented in the video.

Tasks

Example tasks shown in the demo:

  • Buy and rate: recommending known or previously bought products to buy and rate them in the app
  • Fan of Lenta's goods: recommending to buy Lenta's products
  • New vendor: recommending to buy goods of several vendors
  • Diversity in category: recommending to buy goods from one of the most likeable categories
  • Diverse Basket: filling your cart with products from at least 7/10 previously bought (or 1/2 novel) hier_level_2 categories.

Achievements

Example achievements shown in the demo:

  • Krusenstern: visit more than 3 Lenta stores
  • PartyGoer: buy alcohol on more than 10000 rubles
  • Sweet Tooth: buy sweets on >500 rubles
  • Greta Thunberg: buy 10 reusable bags
  • Milk King: buy all brands of Milk
  • Lucky Devil: buy products with discounts 300 times
  • Loyal Customer: buy something in Lenta 4 weeks in a row

About

Personalized tasks & achievements based on customer retail transaction data

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published