Reverse engineering my #1 Hacker News article | Hacker News Reader

Reverse engineering my #1 Hacker News article

danielwirtz.com

123 pointswirtzdan44 comments

I don't want to discourage this type of retrospective, but we (science-type humans) love to look for causal explanations, and can do a great job of reverse-engineering rationale. Even though I think you did do a bunch of things to encourage engagement, it seems like it's more boosted your luck of getting traction from a very small fractional percent to a slightly larger but still very small fractional percent.

I think we should. This sort of hustling is exactly why most of the internet is now completely unreadable. I didn't see the original post but if I had, I'd have probably flagged it for the clickbait content-less title.

The fact that this article (which follows none of its own advice) is steadily making its way to the top of hacker news sort of invalidates all the actual recommendations in the article itself.

I find it kinda odd some people seem to be against this article.

Sure an off the cuff thought could be an unease with an “guide to going viral on HN” but OP isn’t disclosing any secret hacks or anything from what I’ve read, he just wrote down his creative process and reflected a bit on the virality.

I’m sure a bad actor wanting to go viral on HN could think of this themselves within a days work as well, it’s just a fun article.

I'm torn between "nice writeup" and "ugh, a guide to making viral content on HN?"

Yes, fair point!

I tried to reflect on some lessons from this, because I learned from the experience that small things (e.g. the title) can make large differences.

Which means that often good ideas or nice articles might get unrecognized, because they lack a good "packaging".

Making sure those things are on point helps.

But that should be it.

One just needs to write "LLM" to get the same effect ;-)

at this point in time, but the rule is too specific.

"One just needs to write $currentTrendingFad to get the same effect"

Not too long ago it could have been crypto/blockchain, or rewriting $app in $lang

That's the thing of viral is it is just trend following, so recognizing the trend helps the changes of going viral

"rewriting $app in $lang" will never go out of style

The general sentiment on HN against cryptocurrency and blockchain has always been pretty negative as far as I can can remember.

A quick algolia search seems to support that too. Most of the highest voted stories about cryptocurrency and blockchain are negative news and opinions. There are a few highly voted stories about cryptocurrencies and blockchain too, but far fewer.

If you search for something specific like Bitcoin, there looks at a glance to be more positive or neutral stories with high number of votes, and still a few negative ones about that too.

You’ll often find a few positive comments on the stories though, so it’s not like everyone hates it. Not saying that. But overwhelmingly I feel that HN readers as a whole range from disinterested to actively against anything blockchain and that this has been the case for a long time.

To go viral doesn’t mean it’s always a positive reception. If you’re shopping karma points, it just takes getting people engaged and posting something that is known to grate people is the easy play. Just look at all of the drama on Wordpress. Just look at all of social media algos.

Being viral, for content, is not always bad: it can make an otherwise boring or tedious subject engaging and enjoyable (and on the other side, a fun subject might appear boring if badly written), which I think is what this article is about :-).

The title "I've acquired a new superpower" might have the most virality, but it's not a great title for users. Chasing viral content doesn't have to be bad, but you do often get this imbalance between clicks vs UX.

I thought superpower was fitting when I read it, back when the referenced post was originally posted. I was already familiar with optical illusions that you cross your eyes to see some kind of shape emerge in 2.5D

But I had never heard of using that same trick to solve puzzles where you find how two pictures differ. And I tried it out and it worked for me.

It took something I had limited application for and turned it into a slightly less limited power. And a great party trick. That’s sort of a superpower :D

I don't mind a "viral" title for a good article (not those kind of "You won't believe what this woman did to..." or other clearly clickbait titles), but of course, you don't know before clicking if it's a good article, or a clickbait one, indeed.

In this case though, it's a dude posting on his own blog. I'm fine with whatever title he found cool.

Reminds me of guides for selling something online with say, craigslist. There is probably a specific day of the week, or time of day you should post. There is a way to describe your product. There is a title you should use. Maybe you should give a phone number, maybe not. Maybe you should use the craiglist email redirection.

But in some way, I hope people don't do that with hn. I sometimes read helpful product reviews on an amazon product. "I tried the XIWSHDX salad spinner, and I found it to be very useful. blah blah."

I think I'm afraid of "A foolish consistency is the hobgoblin of little minds", or possibly will just hide good organic content.

"How to growth hack HackerNews, without saying those words out loud so I wouldn't get flamed in the comments due to the nature of my audience".

And by the way, absolutely no hate to the author, I totally get it, and fairly ambivalent to it. It is what it is.

They mostly don't work.

Nice analysis. The point about titles is interesting: given how content spreads on the Internet, it's really an invitation to potential readers. If you take that too far, it can devolve into tacky link-bait, but it's worth thinking about how to summarize your post into a handful of compelling words.

Exactly!

As mentioned in another comment, I think it's sad if one writes an amazing article but it stays unrecognized or gets little traction simply because the title was a bad choice.

But at the same time, we have lived long enough with clickbait to know it works (unfortunately), and this shouldn't become a place where people are lured into low-quality posts.

I think there's a lot to be said for knowing your audience. I don't submit often, and most of my submissions only get a handful of points -- very rarely have I had one get onto the front page. (To be clear, I'm talking about submissions of article I found interesting, not things I've written myself, so this is a bit different than TFA's focus).

But, there was one submission I made I knew instantly would be very popular -- The elusive future of San Francisco’s fog -- 152 points, the most by far. Given that many on this forum live, work, or visit SF, and that people everywhere love to talk about the weather, and that SF's fog is an iconic aspect of SF weather, I knew there'd be interest.

Interestingly (or possibly not), I made this [0] comment here (on HN) some time ago and was quickly informed that my technique was used in the 1960s to identify moving celestial objects in successive astronomical photographs. [2] I kind of assumed is was a fairly well-known technique, but didn't realize it also had practical use!

[0] https://news.ycombinator.com/item?id=37981314 [1] https://airandspace.si.edu/collection-objects/microscope-ast...

My highest rated HN link was a sarcastic troll post about the MacBook Touch Bar in 2016. Nothing I have written since has even gotten close to the same level of attention. Someone reposted a more recent article from my blog over the weekend that gathered 4 upvotes and a few hundred views. It's nice to be noticed but I don't live for attention.

I think that unless you want to form a cabal to get something on the front page, there is a great deal of luck in what takes off on HN (and other similar sites). Unless a post quickly gets those all-important few dozen upvotes it will languish on the second or third page wastelands.

Looks like the best way to accumulate kudos is to have a bot that watches for new announcements from Apple, OpenAI, etc and then immediately submits the links to HN.

There are definitely accounts that do this. I occasionally submit stuff to lobsters, usually it appears in the feed here within a minute or two.

Ooops you did it again! Good analysis.

It makes me think: we may be living in a golden age when human analysis and thought still produce the most interesting content. Soon AI-assisted content generation processes will take the next evolutionary step: doing proactive analysis and concept development with the human being more like the back seat driver and director.

AI powered writing assistants, music generators, life coaches, etc, will get so good that authentic solo human creation may become a rarity.

(Tangentially, if you ask Perplexity for tips on writing blog posts that are likely to go viral on HN you'll get the above post, only 6 hours old, as a reference)

Don‘t you think there could be a world where the link to something human might triumph what an AI can generate.

For example: Music is not only about how good it sounds. But also a lot about who the singer is. Their history. What they stand for and what story is behind a album or song.

With AI, I don’t see that we will ever reach that dimension. Expect maybe when AI becomes sentient?

Yes I believe that authentic human creativity is mostly better than artificial. It's like when drum machines became popular in music in the early 80s there was consternation in some circles that this would be the end of drummers. But good drummers like John Bonham, Tony Allen or Vinnie Colaiuta are like magicians whose artistry cannot be replicated in machine form ... but on the other hand drum machines, and electronic 'machine' music enabled incredible vistas of musical creativity.

So maybe AI in music is just another creative toolset that can be used well or badly. My DAW for music creation has auto mastering powered by AI/ML, it does a good job although not as good as a skilled mastering engineer. AI plugins for music composition in DAWs must already exist, they will inevitably get better and more powerful. So I see it as a phenomenon of AI creeping in to our lives rather than a binary dichotomy.

I do think it's problematic - this whole area of AI disrupting the creative arts, and human endeavor in general. I think it's going to carry on creeping into our human creative processes, as new standards of practice and AI acceptance become the norm. We'll still appreciate live musicianship but we won't completely believe it unless we experience it in a live performance.

I agree, the human perspective will remain crucial. I'm torn between being excited by AI-powered creativity and being turned off by it. I will certainly carry on exploring advanced AI tools for writing, music, visual art, etc. for the time being.

For example, I would love to chat with a bot that has a huge memory and context, like a relationship and collaborator, and to train it over many months of conversations so that it becomes like an authentic extension of my creative brain. Then together we could write some serious shit, I'm sure. I think that in the future there may be books written this way that will come to be considered as great works of literature.

But after I play around with AI tools for some time I think I will get bored and will wish to return to a simpler more organic life. Throw away my phone and pick up a paint brush :-)

something not mentioned is the importance of timing. for example a link posted on saturday 11:00PM pacific will not receive nearly as much attention as the same link posted on a friday at 7:00AM pacific.

I’ve seen several blog posts trying to analyze HN data on the best time to post. However, the results are all over the place. For example, the below ones have different recommendations (weekend vs weekday).

- https://blog.rmotr.com/the-best-time-to-post-on-hacker-news-...

- https://medium.com/@mi.schaefer/what-is-the-best-time-to-pos...

But the precondition is that you’re submitting high quality content.

Yes, I think so too.

At least for this post, I made sure to post it in the afternoon. To have an overall between Eurpean and US time. (I'm based in near Amsterdam)

Love it, having an "aha-moment" is always a great indicator that it's something worth sharing! I encourage others to share their ahas so we can all learn together.

100%!

As somebody who has had stuff from my blog hit #1... if you write interesting technical content and have a straightforward path then it usually gets rewarded.

Hell, even the example article would have probably gotten as much (if not more!) traffic by being titled "you can cross your eyes to solve find-the-difference puzzles".

Stop trying to trick people into reading your content. Write something mildly interesting, and post it.

One day I told a friend I could make the front page of Hacker News by end-of-day. He didn't believe me. I won our bet.

The post: https://news.ycombinator.com/item?id=35906604

The "try-it-yourself element" was what got me. Very close to the top of the article was a simple exercise that the reader could perform themselves, and when it "clicked," it was magical. I upvoted it for this reason.

EDIT: Typo

Surprised this hasn't been taken down by the admins.

Not having a link to the original article in the write-up seems like a bit of an obvious oversight, in a piece about reader engagement no less ;)

the technique itself is very interesting and not very well known to many people. also the writing was clear and to the point

And you made it again! :D Nice writeup. Very interesting analysis.

I thought it would be nice to immediatly do a write-up.

Otherwise I push it into the future and it never happens.

Happy that I found some time this week.

alt title: Post-hoc Rationalization of my Previous Success on HN

This blog post is the proof that HN will become enshittified just like any other platform.

Not because of ads, not because of bots, but because of HN's ever growing userbase.

People publishing content too niche for Reddit, will start tailoring it for the HN crowd, socially engineering it to be posted here and garner more upvotes. HN's home will mostly become crowded with "artificially" upvoted stuff aimed at funneling the HN crowd into an ad-supported website

TL;DR trending posts on “dumb” Reddit do just as well on “smart” Hacker News.

Reply on news.ycombinator.com