All Articles

I tracked & analyzed 13,159 Hacker News posts. Here's what I learned.

Dan Ankle
By Dan Ankle: Ex. founder. Frontend developer. Perfecting the art of imperfection. Going solo #indiemaker.Follow me on Twitter

Published 29 Sep 2020

Motivation

Hitting the front page of Hacker News is an incredible source of short-term traffic to your website or article.

According to this article, in 2017, being on the front page brings a sudden influx of between 10k and 30k visitors to your website/article, something many content marketers, product people, and entrepreneurs undoubtedly salivate over, including me.

I occasionally post to Hacker News (with random things and personal stuff) but have never reached the front page before. Therefore, I decided to dig a little deeper into the mechanics of Hacker News and the best approach to take if the sole goal is to hit the front page.

Obviously, a critical part of hitting the front page is submitting an interesting article which resonates with the HN audience. But beyond this, there are other factors at play—timing, title, upvotes, etc.

Methodology: A unique approach

I am not the only person to want to know the best approach to hitting the front page. A few people have analyzed data from Google’s Hacker News Dataset on BigQuery to find the best time to submit to Hacker News.

The problem with this approach is that BigQuery doesn’t know how long a post spent on the front page, nor how long it was there for or what position it reached. It can only provide the total number of upvotes.

With this in mind, I wrote a small NodeJS script to poll Hacker News every 2 minutes. Using a Postgres database, I stored every post, user, and upvote over two weeks.

With this information, I compiled a list of useful information and insights below.

Note: To filter out posts that are either flagged or transiently featured, I set an arbitrary duration of 60+ minutes as a time requirement in the top for a post to be considered a “front page post.”

Limitations

Obviously, the lack of data (13,159 posts) and time spent collecting data (2 weeks) is the most significant limitation of my results. Nonetheless, I believe the results are still relatively useful, at least for the next few weeks or months.

Results

Basic stats

  • I tracked a total of 13,159 posts over two weeks.
  • Of these, 1,073 made the front page for more than 60 minutes.
  • That’s a hit rate of 8.15%
  • Frontpage posts stayed at the top for an average of 483.3 minutes (8 hours)
  • 30% of posts reaching the front page were posted by users who posted more than once over the two weeks.

Takeaway: In my opinion, more posts reach the front page than expected and for longer than expected.

Types of posts breakdown

Hit rate % of different post types making the front page

Takeaway: Interestingly, “Show HN” posts make the front page by proportion very slightly less than “Links,” which goes against previous thought that “Show HN” posts perform better due to an algorithm bias. That being said, there may well be an algorithm bias, just many “Show HN” posts are simply not worth upvoting.

When is the best day to post to Hacker News in 2020?

The chart above shows that you should post on the weekend to stand the best chance of reaching the front page of Hacker News. The difference from the worst to the best day is only 3.2%, so there’s not a lot in it.

The weekend likely gives you a higher chance to reach the front page because fewer posts are submitted, probably due to lower traffic.

Takeaway: If your goal is to reach the front page regardless of traffic, posting on the weekend is your best bet. However, because the difference in chance is relatively low throughout the week, posting on a more popular day might be a better risk-reward ratio based on the considerable difference in the traffic you’d likely receive.

When is the best time to post to Hacker News in 2020?

Taking a more granular look at the combined hours of all the posts submitted, the chart above highlights a much more significant difference (10.9%) between the chance of reaching the front page of Hacker News.

The best time to post to Hacker News is between 6 am—12 pm UTC (11 pm—5 am PDT). Again, the reason is likely because this is when the website traffic is at it’s lowest, and therefore the competition is relatively low.

I imagine the US is responsible for most of Hacker News’s web traffic, which lines up with the results above. Therefore, if you are posting a link for a US-specific audience to HN, it’s probably best to pick a different time like 4 pm UTC (9 am PDT), or midnight UTC (5 pm PDT).

Takeaway: If your link requires a US-specific audience, post at 4 pm UTC (9 am PDT / 12 pm EDT), or midnight UTC (5 pm PDT). Otherwise, post between 6 am—12 pm UTC (11 pm—5 am PDT) for the best chance of reaching the front page of Hacker News.

When in the best day and time to post to Hacker News in 2020?

Combining both days and hours into a total weekly picture is even more evident when it is best to post. However, please note, the lack of data makes this chart vulnerable to skewed results.

The chart above illustrates wild fluctuations between the different hours of the week. The lowest chance of reaching the front page is 1-2%, and the highest is 18%-27%. That’s a massive difference.

Looking at the chart, below are generally the best days & times to post to Hacker News:

  • Monday: 1 am—4 pm UTC (6 pm—9 am PDT)
  • Tuesday, Wednesday & Thursday: 12 am—6 am & 9 am—12 pm UTC (5 pm—11 pm & 2 am—5 am PDT)
  • Friday: 5 am—12 pm UTC (10 pm—5 am PDT)
  • Saturday: 4 am—9 am UTC (9 pm—2 am PDT)
  • Sunday: 3 am—8 am & 10 am—6 pm UTC (8 pm—1 am & 3 am—11 am PDT)

And here are the best times specifically (bear in mind these may not be the case week-in-week-out because the amount of data limits the results):

  • Monday: 12 pm UTC (5 am PDT)
  • Tuesday: 9 am UTC (2 am PDT)
  • Wednesday: 10 am UTC (3 am PDT)
  • Thursday: 12 pm UTC (5 pm PDT)
  • Friday: 8 am UTC (1 am PDT)
  • Saturday: 9 am UTC (2 am PDT)
  • Sunday: 5 am OR 1 pm (10 pm OR 6 am PDT)

Takeaway: It’s pretty clear that the best times to post on Hacker News to reach the front page are usually around 6 am UTC (11 pm PDT), 9 am UTC (2 am PDT), or 12 pm UTC (5 am PDT), when the US is asleep.

However, given that 8 hours is the average time spent on the front page, it’s probably best to post at 12 pm UTC on Monday, Thursday, Saturday or Sunday to maximize traffic for when the US wakes up.

How many votes does a post need to reach the front page of Hacker News in 2020?

5.22 is the average number of votes a post needed within an average time of 27.1 minutes to reach the front page for the first time.

Below is a chart of box plots to illustrate the variation of votes required over 24 hours.

Interestingly, this graph clearly shows every post, regardless of time, required at least three upvotes, and for 75% of posts, they never needed more than eight votes to reach the front page of Hacker News.

Expectedly, the range of upvotes required is lower for the first half of the day than the second half, roughly matching up to the timings above.

In some cases, posts required a lot more votes to reach the front page. Presumably, because either the time it took to get these votes was over more time, or there was a lot of competition.

Takeaway: The votes required to be seen on the front page is relatively small. For the best chance of getting on the front page, you’ll need between 4-6 votes in about 30 minutes, posting in the first half of the day (12 am—12pm UTC).

How much karma do you need to reach the front page of Hacker News in 2020?

Finally, I thought it would be interesting to see the relationship between karma and reaching the front page of Hacker News.

This first pie chart demonstrates the proportion of posts being posted by users with varying levels of karma.

Surprisingly, over 50% of posts are submitted by users with over 1,000 karma points. However, the largest single segment is by users with little to no karma.

The chart above shows that newer users (0-50 karma) make the front page proportionally less than more experienced users. Interestingly, really experienced users (over 10,000 karma) reached the front page significantly more than experienced users.

Takeaway: It’s hard to say whether karma has anything to do with rankings. Likely, users with more karma are just better at posting content that resonates with the Hacker News audience.