---
title: "Discussion on \"What 72% Repeat Purchase Rate Taught Me About Who Not to Market To\" | Hashnode"
description: "Discussion on \"What 72% Repeat Purchase Rate Taught Me About Who Not to Market To\". A Lagos-based analyst examines a UK retail dataset — and discovers that the business never really had a retention problem to begin with.\n\nI tested seven hypotheses against the Online Retail II dataset"
url: "https://hashnode.com/posts/what-72-repeat-purchase-rate-taught-me-about-who-not-to-market-to/69ff52e5f239332df4c80a49"
---

[Feed](https://hashnode.com/feed) /

Discussion

# [What 72% Repeat Purchase Rate Taught Me About Who Not to Market To](https://tobijoshua.hashnode.dev/what-72-repeat-purchase-rate-taught-me-about-who-not-to-market-to?utm_source=hashnode&utm_medium=feed)

A Lagos-based analyst examines a UK retail dataset — and discovers that the business never really had a retention problem to begin with. I tested seven hypotheses against the Online Retail II dataset

May 9 · 6 min read · tobijoshua.hashnode.dev  
Read the article

-   [#retention](https://hashnode.com/tag/retention)
-   [#analytics](https://hashnode.com/tag/analytics)
-   [#product](https://hashnode.com/tag/product)

## Written by

[Joshua Paul-lasisi](https://hashnode.com/@tobijoshua)

@tobijoshua · Lagos, Nigeria

  
[View profile](https://hashnode.com/@tobijoshua)

Sharing thoughts on analytics, development and business perspective

Data &amp; product analyst based in Lagos, Nigeria. I write about what I find when I go into the data — and what it means for the business decisions that follow. Most of my work sits at the intersection of behavioral analytics, product thinking, and the occasional counterintuitive finding that changes the entire recommendation. Day job: Process Engineer at Dangote Fertilizer, monitoring operational data at one of Africa's largest industrial plants. Building toward full-time remote work in product and data analytics. Current focus: funnel analysis, retention analytics, hypothesis-driven analysis, and BigQuery. All project code lives on GitHub.

## Responses

Markdown supported · ⌘↵ to post

No responses yet.
