Data and insights from the Product Hunt API
Getting insights before submitting my first product
Like you, I love Product Hunt.
After all, we are on a Medium post about data extracted from its API. It can only go so far.
It’s become my go-to site for 2 minutes breaks. I get inspiration about products, landing pages or tagline. It’s the best repo for stock photos, icons, sound files or any asset I need for tinkering. And the product comments offer great feedbacks for its makers as well as for anyone reading them.
Before trying to get feature for the first time on the site I used its API to have a look at what the data could tell me.
Here is the game I made: http://www.getdoubles.com/producthunter.html
NB: I’ll try to make another post with standard deviations, comments and user count analysis… Since writing this post made me wants to dive even more into the data.
Product Hunt is trending
Data show a steady increase in total number of product being posted, I suspect the team to tightly control the number of post every day.
There is also a clear shift in the site activity at the end of YC S14 seen through the score a product receives on a daily basis. It is apparently stabilizing. Is it the growth in users slowing or that the community gets harder to convince?
I know for one that I don’t upvote all that much (77 in total) since I use my upvotes as a way to bookmark the products I love.
With higher scores in average, let’s see how the products are distributed.
To illustrate the increasing average score I differentiate two period. Before and after YC. There is decisively more products with a higher score since PH got broad exposure.
I truncated the X-Axis for readability since there is a few products with a score higher than 620.
Another way to see it is by looking at how many products and votes there is each week (starting in 2014 for lisibility).
The number of products posted is now stable at 200-250/week. Certainly limited by the team. While there has been a noticeable increase in the number of votes. It will be interesting to have a look later down the year to see if the votes keep growing. Votes count seems to be stabilizing around 25k/week.
I’d like to look at the number of users too, to see how the three correlate. But that’s for another post.
Looking at the average product score, we also see it stabilizing at 110.
The following is based on the last 6 months/post YC
Impact of the day
Let’s just look at the performance of a product given the day of the week it is posted:
I wasn’t surprised to see less post on weekends, more so by their average upvotes.
Keep in mind that even a higher score on the weekend is good, the fewer number of posts also mean there is less overall votes during the weekend, so probably less exposure and feedbacks. We’ll see in another post how the number of comments vary.
What is posted - What interest people
Let’s look at product taglines and what compose them. What words appears often and what “works”:
The two graphs below are the same, just reordered
and reordered
It appears that people tend to write detailed taglines, try to be explicit !
Thanks for reading
I’m looking for new challenges, you can contact me here: https://twitter.com/ASwiniarski
Products I love: http://www.producthunt.com/aswiniarski
Try my iOS game here: http://www.getdoubles.com/producthunter.html
And if you speak french, try OPUS: http://www.opusapp.co
Want to look for yourself, raw data are here
I’ve put them in a .xls here: https://www.dropbox.com/s/fq4px1de9p0xz6q/PHDataRaw.xls?dl=0
There is a couple of post already on the matter and I’ve tried not to overlap. You’ll find those posts here http://tinyurl.com/punydsu and here http://tinyurl.com/outqb3r . They offer a great read.
