https://github.com/chrisclark/retentionizer

https://github.com/chrisclark/retentionizer

Last synced: 3 days ago
JSON representation

• Host: GitHub
• URL: https://github.com/chrisclark/retentionizer
• Owner: chrisclark
• License: apache-2.0
• Created: 2015-02-15T23:02:13.000Z (about 11 years ago)
• Default Branch: master
• Last Pushed: 2017-03-11T23:25:03.000Z (almost 9 years ago)
• Last Synced: 2026-01-15T08:04:59.788Z (about 2 months ago)
• Language: HTML
• Size: 506 KB
• Stars: 7
• Watchers: 2
• Forks: 2
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

• awesome-billing - `retentionizer` - Python package to show the projected retention rates for each cohort and calculate the LTV of a given customer in that cohort. (Business Intelligence / Customer Lifetime Value)

README

          
[Try it live!](http://retentionizer.herokuapp.com/)

_____

Using the techniques in [How to Project Customer Retention](https://marketing.wharton.upenn.edu/files/?whdmsaction=public:main.file&fileID=327) by Fader & Hardie (2006), and an implementation of those techniques by [JD Maturen](https://github.com/jdmaturen), Retentionizer will fit a shifted-beta-geometric distribution to the data, show the projected retention rates for each cohort, show the imputed beta distribution for each cohort, and calculate the LTV of a given customer in that cohort.

Basically, it turns a sample of cohort survival rates:

| t  | past 30 |

|----|---------|

| 0  | 1.0     |

| 1  | .81     |

| 2  | .80     |

| 3  | .76     |

| 4  | .75     |

| 5  | .72     |

| 6  | .70     |

| 7  | .67     |

| 8  | .66     |

| 9  | .65     |

| 10 | .64     |

Into this:

![sample-output](https://raw.githubusercontent.com/chrisclark/retentionizer/master/static/img/sample-output.png)

Retentionizer is built by [David Chudzicki](https://www.github.com/dchudz) and [Chris Clark](https://blog.untrod.com/).