Analyzing Product Hunt Data

During the last few months, I’ve become a big fan of Product Hunt, a site that showcases newly released products. Now that Product Hunt’s been around for a little while and is getting decent traffic, I thought it might be fun to look at recent submissions to see which products and product descriptions get the most attention. I downloaded the last ~6 months of submissions and spent several hours exploring the data. This post summarizes some of my findings. If you don’t want to read the whole thing, the most interesting section is probably “Miscellaneous Observations.”

Basic Stats

Here’s a plot of of the vote distribution (each bar represents a single product):

Common Words in Product Names

When looking at words in product names, I noticed 3 main themes.

Theme #1: Trendy naming conventions

Theme #2: Software for startups and for product development

Theme #3: Common actions and broad product categories

Most Upvoted Words in Product Names

The following are a few of the words most correlated with high vote counts:

Common Words in Product Descriptions

Theme #1: Common English words

Theme #2: Adjectives

These are exactly the adjectives I would have expected at the top of the list. :)

Most Upvoted Words and Phrases in Product Descriptions

The two big themes here seem to be simplicity (“in seconds”, “the easiest”, “Tinder”) and products for developers and startups (“for developers”, “for startups”, “entrepreneur”, “your app”). The latter is not surprising since there is a lot of overlap between the startup community and the Product Hunt community.

Miscellaneous Observations

Some conclusions based on frequently occurring terms in product descriptions:

(Most of these conclusions have at least several dozen samples for each term.)


This analysis is meant to be more fun than rigorous. The sample size is not huge and I’m not a statistician. I checked my code for obvious bugs, but it might not be perfect. YMMV. IANAL.

So What?

This was fun to do with a dataset of 3000 products, but what really excites me is the dataset that Product Hunt will have in a year or two. So many things could be done with a database of tens of thousands of products tied to popularity data and user comments. Are some people great at finding popular products before everyone else? They might make for great VCs. Is “Uber for X” a gimmick, or a product category that everyone values? What are the most compelling adjectives for describing a product? A lot of these questions could be answered with enough data, and I’m looking forward to seeing what the data shows.

Tags: Data Analysis
If you enjoyed this post, please share it with your friends!

Subscribe by Email
Copyright © 2013-2017 Leo Polovets