👋

Exposing product trends using Product Hunt

Two years ago I built the Product Hunt leaderboard as a tool to identify influencers in Silicon Valley's startup scene (and made a blog post). Since Product Hunt has grown up, and with it an increase in the number of daily products submissions, votes and comments, and I thought it a fun idea to see what data I could extract from all this activity.

Data Treasure

Product Hunt submissions have an interesting characteristic of representing a new product or venture, of which its makers identified an market opportunity and invested time and energy on getting it shipped to the public. Next to that Product Hunt's activity of votes and comments reflect how much that product resonates with the actual market. Finally only featured products make it to the homepage and reach a wider audience, which represents another curation filter.

If you add those three product submission characteristics together and group them with Product Hunt's topic and submission date you can get a cumulative product trend representing venture activity. Which is not only a fun topic, but could also provides valuable market insights.

Obligatory note; Product Hunt's topic feature has only have been available for a year, and its quality is dependent to how they are manually curated. Also products on Product Hunt are heavily skewed towards the consumer market, and some topics are easier to release a product in (e.g. chrome extension) then others (e.g. hardware) which strongly impacts the submissions metric.

Methodology

[TL;DR: visualized submissions to Product Hunt grouped by topic]

The goal is to make a visualization of the Product Hunt trends, not an actual statistical analysis.

For every Product Hunt submission (70k) in the dataset I have the number of votes, comments, whether it featured on the homepage, and its connected topic(s). I filter out the submissions categories games, books and podcasts so I can focus on tech.

I group all these products under their respective topic, of which I simply calculate the topic's cumulative performance indicators (total votes, total comments, total featured).

Then the script takes into account time, by allowing a script to query my dataset by submission date in a certain period (week, month). Next to time, I add some filtering and plot it to a nice chart to cherry pick interesting exceptions.

The results

An beautiful interactive motion bubble chart which you can explore yourself! Bubble charts are perfect for visualizations of data points with multiple characteristics, over a time period. Use the filter on the right to select your trends of interest, and the time slider to see how it evolves.

My findings

To juice things up, I did some exploring of the data myself and made it a Q&A (like i did previously on TechCrunch trends crunched and Rip Tribute to web 1.0).

What are the biggest topic growers of 2016?

Comparing Juli '15 and Juli '16 against each other.

  • In Juli '16 Artificial Intelligence and Wearables were all over the homepage (featured). Artificial Intelligence also got a strong amount of comments (2x) compared to iPad (a good steady reference topic.
  • As no surprise Bots and Fintech already had strong attention (votes) in 2015, but became one of the top new featured topics as well.
  • Starting in beginning of '16 Developer Tools out grows every other major topic for number of submissions. When only looking at times featured it was also one of the strongest topics.
  • Developer Tools is quickly followed by Messaging as the biggest topic growers, especially in number of comments.

What leads to discussion?

Instead of looking at votes, if we sort by comment quantity.

  • Web, iPhone, Android and Productivity are the top topics on which most is talked about, but are also the topics which have the most product submissions. If you take that into account, Mac, Design Tools and Chrome Extension stand out.

Steady and strong

Topics which Product Hunt features continuously.

  • Submissions for Web and iPhone are not too far apart, if you only look at featured products they almost overlap. But Web get almost twice as much votes.
  • Android is at half the amount of submissions compared to iPhone topics, and also half the number of times featured.
  • For all three topics comments are strong compared to similar popular topics.
  • The iPad is one of the strongest topics in 2015 and 2016 with up to 57 times featured in a month (Juli '16)
  • Email and Music remain strong and have a relative high chance of being featured.

What trends are in demise?

  • In 2015 the Apple Watch is one of the stronger topics (submitted vs votes) but became only a small blink on the radar in 2016.
  • Virtual Assistants and Quantified Self don't seem to catch on.
  • Browser extension for Safari and Firefox are almost near zero, while Chrome still remains strong.

What is the best non-US location?

  • Using Product Hunt's "Made in :location" topics, India, Canada, UK and Germany stand out on top.
  • In March '16, Canada had a crazy month with 15 times featured.
  • France has a few spikes in features as well during the year.
  • Italy had a good start of '16, but has been in decline since.

Concluding

There could be a much deeper level in exploring these trends, using statistics or machine learning models. Another opportunity could be in both focusing on more specific trends (within web dev or design) and matching the data up with qualitative input.
I'm thinking of investors, developer advocates or senior product people adding their own insights and analysis which provide context to identified trends. Or maybe even be able to do predictions of trends that way, which could be used as input for companies to see what to invest resources in.

If you are more interested in the topic of tech trends, and/or get involved ping me with an short email at yvo@yvoschaap.com or leave a message via @yvoschaap.

Posted by on .