Skip to content

A/B testing that was done on a course from Tinkoff. The tasks and data are as close to real cases as possible.

Notifications You must be signed in to change notification settings

aleksandrGlebov/Tinkoff-AB-tests

Repository files navigation

Data Analysis Project - A/B Testing

This repository contains the analysis and implementation of A/B tests aimed at optimizing product sales and service offerings.

Task 1: Optimizing Product Pricing

Context for task 1

The hypothesis is that the current pricing of the product might be on the higher side, which could be impacting sales. Reducing the product price is believed to potentially increase the sales frequency and overall profitability.

Experiment for task 1

An A/B test is conducted with two groups:

  • Control Group: Products are sold at the original price.
  • Test Group: Products are sold at a reduced price.

The objective of this A/B test is to validate if reducing the price indeed increases the product's profitability.

Statistical Parameters for task 1

  • Significance Level (Alpha): 5%
  • Minimum Detectable Effect: 5% change in the target metric
  • Statistical Power (1 - Beta): 80%

Task 2: Evaluating Aggressive Service Sales Approach

Context for task 2

The Central Bank has eased certain regulations that previously limited the options for selling services. There is an opportunity to adopt a more aggressive sales approach for the services. However, there are concerns that this might have a negative impact on the core product (credit card) economics due to possible customer dissatisfaction.

Experiment for task 2

An A/B test is conducted by dividing the customer flow into two groups:

  • Control Group: Service is offered in the current manner.
  • Test Group: Service is offered aggressively.

The objective of this A/B test is to determine whether a more aggressive sales approach for services leads to positive or negative changes in the core product's (credit card) economics.

Statistical Parameters for task 2

  • Significance Level (Alpha): 5%
  • Minimum Detectable Effect: 5% change in the target metric
  • Statistical Power (1 - Beta): 80%

Conclusion

It is essential to carefully plan and analyze A/B tests for optimizing product sales and service offerings. Adherence to statistical principles and significance criteria is key to obtaining reliable results.

About

A/B testing that was done on a course from Tinkoff. The tasks and data are as close to real cases as possible.

Topics

Resources

Stars

Watchers

Forks