90 comments

[ 2.8 ms ] story [ 157 ms ] thread
(comment deleted)
(comment deleted)
Nice post! I think it's interesting how every website has their own culture and what is totally fine on Product Hunt wouldn't be fine here etc.
Agreed! We had quite a few false starts on HN before getting any traction.
This isn't the first post of TS either, but it looks like it is his first successful one.

FWIW, had a few failed "Show HN's" myself. You certainly need a certain amount of luck in order to get enough upvotes to be noticed.

Author here. Ever since I knew of Hacker News, I've always been interested in the massive leverage it provide to both startups and individuals.

Choose the correct title and your startup can be projected to thousands. Write a thoughtful comment and you can make 20 sales. Ask the right question and you can end up with a Silicon Valley job (that actually happened)!

I came across Simple Analytics [1] a privacy-focused startup which pulls in huge traffic from HN. Fortunately all their traffic was all "/open" so I could compare traffic spikes with their founders [2] Hacker News activity and understand what really does works on Hacker News.

[1] - https://simpleanalytics.com

[2] - https://twitter.com/AdriaanvRossum

Thanks man, that's really nice insights :)
Yeah, it's just the title. I was on top for like 16h or so because I had a catchy title and some way of supporting that title within the content.

Comments were far different. Almost all of them were negative and highly unconstructive. I don't mind criticism, but I really mind criticism from which I can't learn anything. One HN reader actually took the time to write a constructive comment. I don't know who it was nor did I thank him at the time, but that one constructive criticism was way more helpful than all the traffic boost I got.

Thank you for sharing this - Almost all the info in the article is new to me. Definitely worth reading
The contrarian in me is fighting the urge to click on this story.
Finally, I can legitly skip content and read comments first.
That's what I did. The most ridiculously obvious click-bait.
Click-bait implies deceptiveness, but the article content is consistent with the promise in the title.
It is actually very interesting and relevant read. He specifically mentions that deceptive or click-bait titles don’t go far on HN.
I have to agree, it reminds me of this book called "How to make $1000 on Amazon", the book cost $1000.
Choose the correct title and your startup can be projected to thousands.

If you're selling a product that HN readers will buy then that's awesome, and well worth doing. Those of us who sit and read /newest will happily upvote things that are great.

If you're selling something that will appeal to a different group of people who don't read HN, or to the mass market so HN readers won't make much of a dent in the metrics, then it'll just spike your bandwidth costs for a while with no upside. If that's the case then focus on marketing to your potential customers. Ignore everything else. 30K views from random people who won't buy what you're selling is vanity.

The founder of Mailparser in this episode of Indie Hackers[1] talks exactly about this. When he did Show HN, His thread was on the front of HN and he got so many views and advice but zero sales. But ultimately his Show HN thread lead him to a partnership that made him a solopreneur.

[1]-https://www.indiehackers.com/podcast/024-moritz-dausinger-of...

Adriaan from the article here. You are right. HN is the right target group for Simple Analytics. If I had posted a totally different product with no privacy angle or interest for HN readers, it would have little result. You need to find you target audience and I believe HN is one for Simple Analytics.

I also hear a lot of people knowing Simple Analytics from Hacker News. So maybe they don't convert to customer yet, they know about my tool. Which is super useful and got me talks with DuckDuckGo and other great companies. It's not only for selling your tool, it's also great for getting your name out there.

Yeah, I couldn't agree more. I do in fact make this point in the case study:

"The point is that once you start hanging with your perfect audience, conventional marketing goes out the window. You don't have to try so hard.

Find the place where your tribe hangs out. Show up, help people out, talk about what you're doing, and the rest will take care of itself."

Hacker News is useless, unless you blog / product is aimed at the Hacker News archetype.

> 30K views from random people who won't buy what you're selling is vanity.

Not sure how far you would get in the world of advertising sales with that observation!

30K views from the nit-picking HN crowd could be very useful if you want to refine your product for a wider audience. Things like whether the page loads without errors or spelling mistakes will be useful to correct with the HN audience. You might not get sales but you will get your sales pitch debugged.

There is definitely a set of topics and themes that are particularly popular on HN, ex. privacy, de-Googling, write-it-in-Rust, Plan 9, etc.

If your product fits into this niche (or better, you can create a product specifically for the niche), I doubt you will have many problems. Otherwise, good luck reaching even page 5.

Actually, Patio11(ie kalzemeus/Patrick) showed/proved a simple SEO strategy:

Create good content for one group of people (ie HN readers) while generating SEO juice for the real buyers(teachers).

PS It only works if you are earnest about both sides of the equation.

That is super useful, thanks for sharing this!
Super interesting, thanks for sharing. I had never read about anyone managing to get consistent traffic from HN.
I think a few sites get consistently highly voted, because they're known for, and produce, good content. Gwern.net might be one.
I m quite baffled by meta posts claiming that hacker news has a different culture that anywhere else. Sure its not facebook but it s very similar to mid-level moderated subreddits. Major plus is there are no images.
I recently posted one of my silly side-projects to HN and it reached the front-page; I'm not really into analytics and user tracking, but I was curious about how much traffic was generated.

So, I just awk/grep'ed Apache's access.log [1]. The results were about 23k requests from 17k unique addresses in a couple of days and then a longer tail for about a week, but haven't checked since.

[1] https://tpaschalis.github.io/show-hn-sunlight-live/

Thanks for posting your stats. Your results match almost exactly what happened when a (trollish) article of mine[0] was posted a couple of years ago.

From memory, I got about 20K hits in the first 24 hours, and another 5K over the next few days, trailing off to nothing after about a week. It was exciting at the time but I had no evidence of any retention.

[0] https://sheep.horse/2016/11/rip_apple_macbook.html

Out of curiosity, do blog owners expect retention from a popular article? I just don't consume content this way. Unless the article is covering a long-term project I might be interested in (following game development from inception to release, etc) or the article sells something to me and I convert to a client, why would I start visiting a blog on the off chance that the author strikes lightning again? I much prefer curation services like HN.
I think they hope for retention and therefore measure it, rather than expecting it.
If I come across I blog I like then I tend to add it to my RSS reader but perhaps I am in the minority.

As for my blog, it is very unfocused and for every well-thought out post on technical matters there are half a dozen "check out these photos" posts.

Hey Matt. Nice work with osdev.org, I'll be checking it out. Allow me to offer my point of view on your question.

I initially started out this blog to improve my writing skills; I don't have any tracking, ads, analytics. I don't know (and don't care, tbh) if and how many people read it. It allows me to write down things I've learned, really cement that knowledge, and it's also hella fun!

Many people that I know, which maintain their own blog are on the same page. Writing in public also allows other interesting people to offer their opinions, and start a conversation. This is the best thing I've got from it, helpful comments, criticism, and different viewpoints from strangers!

Ever seen https://www.die.net/earth/ ? I ask because it's really similar to OP's product Sunlight.live/ but came out in the late 90s early 2000s. It uses a satellite image of the weather, and a version of earth by day and earth by night to generate a composite that looks really sleek. Thanks for all the downvotes/fish.
Hey! Yes, I have seen die.net, and it was a source of inspiration. I wanted to build something original, use Python, learn about astronomy, and maybe help someone else with their ideas!
This matches up with what I saw. I don't have an access.log, but I took the total network traffic for that day and divided that by the size of one full page load including resources, which came out to about 10k to 15k page views for the first day.
I posted a question [1] to see if there is any interest for octopus deploy alternative (fdeploy.com) but got zero replies. I'm still developing the tool for my personal use though.

[1] https://news.ycombinator.com/item?id=20506088

Edit: just noticed that there was indeed 1 comment. Made my day :)

I also got 30k views for No CS Degree the other week. Should have answered every comment but hard when there are like 200!
FYI if the comment count goes above the number of points (after a certain number of points, ~45, I forget) - your submission gets penalized and drops off the front page really fast.

So there are times when you might want to hold off on adding comments if it's close to that threshold.

Interestingly, I wrote a similar analysis 7 years ago. 30k visits was also what I saw! https://shkspr.mobi/blog/2012/11/whats-the-front-page-of-hac...

I had a couple of moderately popular posts in June this year. HN sent 31,588 views directly - plus associated views from apps. A post of mine in March got 36k.

I wonder if HN has reached a local maximum of readers?

I take it you're always reaching some small subset of HN readers - filtered by the kind of people who visit links by the kind of people actually interested in the topic, etc. I know that of the links that I do upvote, I probably only visit a tiny fraction of them.
Interesting! I had a post top front page earlier this year and it generated 31,029 pageviews - almost exactly what you experienced.
Perhaps by posting an article titled “How to get 30k Hacker News visitors to your website”?
Yes, my first though was: Is this title meta? Am I being baited?
This is plain garbage. The best way to get visitors to your website is to create good content. No marketing, no click-bait or other garbage, just good content.
Hm, it is very doubtful that good content is simply enough nowadays. With tons of information in the internet, "just good content" can hardly be discovered by itself. There are should be ways for target audiences to find it, share it and engage with it.
It's the eternal debate: is content truly king? Can you just push awesome content out there, and get floods of traffic in?

Or is it a "If a tree falls in a forest..." type situation in that an unknown website with great content will continue to be an unknown website?

Since Google can't truly understand content quality/semantics, a certain level of marketing does seem to be needed to bring traffic in.

That is not garbage, that is exactly what he's saying: do something interesting, then share it with the people who are interested in that kind of thing.
You'll need to attract attention to it in some way. HN, when done right, can be a way to do that.
actually, i'm not clicking on that.
Write an article with the title "how to get X users to your website" and pretend you're not click baiting
I think the most important point is in the final paragraph:

"Find the place where your tribe hangs out. Show up, help people out, talk about what you're doing, and the rest will take care of itself."

The article starts out sounding like it's about effective marketing, but as it progresses, the unavoidable conclusion becomes that it only works because he's sharing content relevant to HN, in a way that appeals to HN, because he himself is a HN user.

Marketing as an insider of the crowd you're marketing to is trivial. Marketing as an outsider is hard.

This exactly. I've tried couple of times with my side project on show HN - https://tardis.dev (aimed at algotraders) and it didn't pick up, but on the other hand on reddit algotrading sub I received useful feedback and even few paying users.
Sites that display content based on upvotes are inherently chaotic, you can get published or denied for no reason at all. They are just biased into being useful, but there are no guarantees.

A set of small communities, like on Redit, have a much more predictable behavior. If you have content that interests one of them, it is very likely that it will get picked.

I think there are ways to game it, actually. You need someone else to give you that first upvote. I think/assume that those first few upvotes will make it visible to more people which will hopefully attract more upvotes. Of course the content still needs to be good enough to get those upvotes; but it makes it more likely to dodge the risk of good content going ignored because nobody sees it.
I've tried that in limited scope, didn't change much, but perhaps content wasn't good fit for the audience and I accept that.
> LESSON 4 - HOW TO SHARE

> ...

> What I’ve noticed some people do is put out a tweet directing people to the newest page. This gives their followers an easy route to find their post.

I'm pretty sure this method gets flagged. It's easy enough to detect a bunch of users landing on the new page with Twitter as a referrer all upvoting the same article. The mods and the community don't appreciate sockpuppeting (unless you're in YC of course).

Oh oh oh, so true. I love that you just called this out.
It's so true. Every time a YC startup launches, there is an upvote ring from YC I think. They reach front page so quickly. Sick practices :(
Posts that start with "Launch HN" get placed on the front page similarly to how job ads do. These are the two formal things that HN gives back to YC in exchange for funding it: https://hn.algolia.com/?sort=byDate&dateRange=all&type=comme.... The difference is that job ads can't be upvoted and decline in rank as a function of time, while Launch HNs can be upvoted and get ranked like regular stories after the initial placement.

Posts by YC startups after their initial launch don't get any special treatment. If a submission doesn't say "Launch HN", it's subject to the same rules as everything else.

Edit: HN also displays YC alumni names in orange to other YC alumni.

I would like to hear what you think you know about this, and how.

A large part of our job is trying to prevent HN from being gamed. YC startups don't get to do that any more than others, and if we notice egregious cases we email them and give them an extra scolding.

> don't appreciate sockpuppeting (unless you're in YC of course)

That's not true. Votes from YC alums get no special treatment, and HN's voting ring detection software works the same on their posts as anyone else's. If you or anyone has concerns about this, I'd be happy to address them, but specific links would be helpful. It's pretty hard to respond to drive-by accusations like this, which have no information in them.

The first thing we tell YC startups is not to solicit upvotes and comments. If they email us complaining that their post didn't get traction, we tell them the same things we tell others.

Maybe I heard wrong then, but I was told that links posted by YC members are colored differently visible only to other YC members. Perhaps that was just a conspiracy.
YC alumni usernames appear in orange to other YC alumni, but HN's ring detector prevents that from having much effect on the voting process.
Thanks for confirming this feature. I mean, if someone's name stands out in a nice bright orange color, I might feel just a bit trigger happy to upvote it. Unless you're just completely ignoring votes by other YC alum altogether, I can't see the ring detector being able to discern from authenticity vs bias towards liking orange colored names. Even just a few upvotes can give someone an advantage.
We don't see voting rings involving orange accounts getting more traction than voting rings in general. Our systems certainly aren't perfect, though. If you think you see a post that is being affected by sockpuppet voting, you're welcome to send us a link so we can look into it.
Use the title “How to get 30,000 Hacker News visitors to your website”
What the hell is going on with this website? The post pops up in a modal (which is not focused, so my cursor keys and space bar didn't work to scroll at first) that only takes up about a third of the screen. Close it and there's a bunch of article links with giant thumbnails.
Except that people have admitted to gaming the upvote system here many times in the past so its not all organics-based. I wouldn't be surprised if the majority of top articles had been shilled to N 'friends' / 'associates' / 'colleagues' etc before they had reached number 1.

If you look at all the new submissions there are many interesting articles and most of them only get like 1 or 2 upvotes. So either its all random luck or people are coordinating it like they do on Reddit. But ofc: I will trust this totally legit website about marketing that everything is hard work, organics, yada-yada-yada, here's my email bro and ill be sure to click your referral links too ;>)

This making it to the top posts pretty much drives the point home ^^'