Hacker News new | past | comments | ask | show | jobs | submit login
Most promoted and blocked domains on Kagi (kagi.com)
379 points by lucgommans 23 days ago | hide | past | favorite | 182 comments



What I gather from this is few if any non-software devs use Kagi

8 of the top 10 "raised" sites are software dev sites and with #6 being MDN I'm guessing it's not just any software devs, is web developers specifically that use Kagi.

Am I drawing the wrong conclusion? Does that mean Kagi's days are numbered? What would it take for them to get enough non-web-devs that the top 10 raised looked more representative of the average internet user?


I think it is also a consideration that more technically inclined users are more likely to spend time tailoring their experience in this way.

Technical users also likely have specific sites (namely documentation) that they wish to bump in their results where other professions might not.

I'm not disagreeing with your point, just offering some thoughts of my own.


I think it’s more word of mouth because many professions use the Internet and could surely make use of site preferences.

To be honest, I’ve only heard of Kagi on HN and nowhere in any of my friend groups.


> Does that mean Kagi's days are numbered?

Why would it mean that their days are numbered? Nothing wrong with having steady income from a loyal customer-base, even if that customer-base is niche.


It would mean that if they are operating at a loss and hoping to capture more of the market later. But if they're profitable then the main danger is a competitor coming along.

Growth mindset is a big part of our sick society, unfortunately. It's the only thing our politicians like to talk about, after all. Being a stable business delivering value is as good as dead.


Kagi is the only paid search engine and they are already profitable: https://blog.kagi.com/what-is-next-for-kagi

I think the bigger danger is the competition around AI chat assistants. There are other paid assistants, but you already see that even the paid assistants are getting trained to promote certain corporations.


Kagi were an early adopter of LLMs. You can use Kagi assistant for AI needs.


Kagi Assistant is awesome. It uses their search for context. I got a free year of Perplexity and liked Kagi better.


Bing Chat (chatgpt without login) has replaced more than 80% of my searches. Still with kagi for the remaining 20% but if my searches run out I probably use duck again for those


Because devs (or maybe just web devs?) will be replaced by AI soon, I guess.

edit I do not agree with this, but this is what i assume OP was referring to? Because right now web devs is a quite populous and active crowd.


Someone still has to write the prompts though, someone who knows what he is doing.


The average internet user is never going to pay for search. Trying to target them would likely be a mistake.


The average person is far more willing to pay for anything computer-wise than the average software developer / hacker.

Or how about the average knowledge worker? There's hundreds of millions of them.


I am an average internet user.

I pay for Kagi and it's refreshing to actually see real results rather than sponsored slop from the others.


  I am an average internet user.
How many average internet users comment on HN?


The average Internet user doesn't contribute content/comments at all.

The average Internet user doesn't visit HN at all.

The intersection of these groups, HN visitors that comment? Distinctly not average.


it's a nice round number.


Goalposts


Not every business sets out to be the next Google

Loyal Kagi customer here, based on their posts and in my dealings with them, they are doing their thing and doing well

They are focused on privacy, do a great job of it, and their AI assistant is top notch (highly recommend you take a look, can choose from many models and swap out responses instantly, not even getting into the awesome search features)

Not commenting on your (good) who is the main audience question, rather the other point about if Kagi is doing well

I subscribed my girlfriend to it as well and tell people whenever the moment is appropriate

Really rooting for these guys to succeed long term

As an aside, when I got my Kagi subscription the first thing I did was lower Pinterest results


I hear a lot of good things about kagi but privacy isn't one. They need to record every search and connect it to your account. Duckduckgo is more known for privacy and searches well through tor.

Here are the stats: https://kagi.com/stats

How many are paid vs trial accounts?

We know family doesn't offer a trial nor teams.

126 teams x 5 or 6 members = 1,000 accounts at 10 per day 10k

4500 family plans: most will take the 20 a month plan 100k

45000 individuals lets say they are all paid most on the 5 dollar plans lets assume on average 6.50 is earned 300k

Then you have orion+ members at 2000 giving an extra $15 per account. 30k

They probably make 450k a month

They have 19 employees on linkedin and they are listed at under 50 everywhere else. Lets give them 25 employees at 100k average salary which would be 2.5 million in salaries which might be low.

Add on costs to actually run the website (paid search, servers, office costs) which hopefully cost less than 3.5 million.. the rest is profit.

I'd say they are doing well enough. My average of 5/6 per team might be much higher if they have a few 100+ sized teams. I think the mode would be 5/6 regardless of the average.


> I hear a lot of good things about kagi but privacy isn't one.

I don’t use it, but the Privacy Pass [0] thing should actually make it great for privacy if you care more about that than personalization.

[0]: https://blog.kagi.com/kagi-privacy-pass


> I hear a lot of good things about kagi but privacy isn't one. They need to record every search and connect it to your account.

Looking at their privacy policy they state the following:

> We may store web requests made by user browser temporarily, with strict retention periods, for debugging purposes, and in a manner that they are not linked to an account.

https://kagi.com/privacy#server-data


Is it actually true though? They've lied about other things, what stops them from lying about that?

For 2 years they've said "Kagi has zero marketing budget and spends zero on marketing", that's a lie and they still haven't changed it. They've sponsored over 200 youtube videos https://filmot.com/search/%22brought%20to%20you%20in%20assoc...


Do we know that they spend money on this or is that inferred?

I asked because the search you've linked to only seems to list relatively niche chess videos on one creator's account. And the videos themselves are also ad-supported. It's (at least) not impossible that the sponsorship association is not budgeted in the way that (I think) you infer.


https://news.ycombinator.com/item?id=40013889

https://kagifeedback.org/d/3822-kagi-is-sponsoring-youtube-v...

"I watched Daniel King for 10 years and have previously donated to him, now I am happy that we are in position to sponsor his work which I believe is extraordinary.", this also definitely sounds like monetary compensation to me.


Thanks! I see why they may not consider this "marketing", even though it is obviously a sponsorship based on that information.



That’s a great segment to target. I’d take that any day over most others. It’s also a pretty hard bunch to please, so pretty defensible unless you somehow beat Kagi by a lot.

It’s also much bigger than their customer base. Keep going, Kagi.


> I'm guessing it's not just any software devs, is web developers specifically that use Kagi.

I do both, and I have MDN pinned. The thing is, if I search for C#, Python, Java, I get great results on any search engine. If I search for web stuff, I get tons of crap and mediocrity. MDN is almost always what I want instead of some other stuff.

Regarding "days numbered", I agree with the gist of the other replies ;)


I have MDN pinned and I’m not even a software developer, just a mathematician who likes programming.


> Does that mean Kagi's days are numbered?

Perhaps it just means technically inclined people are the first to see value and use in some product. Whether word of mouth spreads and people will find value in paying for search engine + goodies - time will tell.


Kagi's days are probably not numbered because of web devs (web devs are not going away any time soon, even with advancements in LLMs) but rather because of the indexes they use, or rather don't use.

Kagi is fairly good at ranking results and essentially making do with what they have, but in my experience it does not seem very good at all for searching for anything particularly obscure. It's like how DuckDuckGo uses Bing - nearly useless if you're not searching popular sites (like Wikipedia, MDN, etc.)


DuckDuckGo likely also had almost exclusively tech people as users because the average person doesn't even think about search engines and more importantly doesn't want to. Alternative search engines only have a chance at getting a wider audience when Google messes up enough that people need to think about it.

With Kagi this is more significant because paying for a search engine is an even larger obstacle than simply switching.


> Am I drawing the wrong conclusion? Does that mean Kagi's days are numbered?

It is developers who are more likely to take digital privacy more seriously than the rest. And so, it seems like Kagi is indeed on its way finding product fit with the most demanding segment (from a digital/information privacy standpoint) there is.

For any tech, it is fully expected that early adopters belong to a niche. Whether Kagi wants to make the leap or stay in this segment is upto them. There is no indication that they have saturated the market even in their niche ("developers"), while this niche itself on its own might be lucrative enough.

Besides, Kagi may very well venture into other products to sell to the same market segment.

If anything, to me, the writing that's on the wall for Vlad (Kagi's creator) right now is... "The world's your oyster".


This good be a good sign.

It’s better to “own” 10% a known, faithful customer base than 0.1% of a floating one.


I have a hard time figuring out how to pitch a paid search engine to "normal people", when Google works and is free.


You literally just wait for them to complain about Google, then say “Well, you get what you pay for.” They will be curious because everyone knows search is free.

IME, this works well.


The problem is that I’ve never seen a “normal” person complain about Google search.


Their days aren't numbered because they've built a sustainable business with a small team of developers that doesn't require a huge user-base to fund. Arguably if Kagi were more popular it would be worse. Just the fact that they aren't Google and use their own ranking while being small enough to avoid the notice of SEO hackers carries them a long way.


I’m one of them. I love Kagi, although the first week or two had a ton of !g. Now I only really bang for local areas, conversions, or shopping (maybe).

If they can sustain, maybe they can takeoff. Search in GenAI world is hard, and Google has other focuses with talent and inference chips too.

I hope their days aren’t numbered!

I think it needs some UI improvements. It’s ugly, and I find it can hinder actual use.

More usability improvements on features. There’s a lot I’m still not leveraging because I haven’t bothered learning. Maybe they can build an LLM tool to help with this?

And I don’t care how, but make it easy to make it default on all browsers, mobile mainly. Maybe they fixed this recently, but when I was swapping browsers recently, this was annoying. If they can’t fix this, they probably won’t make it.

Just some top of mind thoughts as high-usage, 95%+ non-coding user.


I'd rather gather that the normal people subscribing to Kagi are simply satisfied with the better quality search results, and don't bother so much with domain ranking.

The "hacker" type of computer user is zealous in everything they do.

What's important is that Kagi refrains from letting hacker zealots influence their business too much, because normal people also deserve high quality search results, and most knowledge workers are not software developers.


Well I am a webdev who more or less exclusivly uses Kagi so I guess I support that point; but I also havn't bothered raising any sites and I would guess a lot of people don't.


I'm not sure about numbered, all I see is great arguments to try it and thus pay for it.


Most pinned website by far is Wikipedia tho, by a large margin.


Pintrest owns all the top 7 blocked results. Good, they've earned it.

I never understood why Google let them destroy Google Image search results.


Apparently Pinterest has half a billion users[1], presumably significant number of those appreciate those search results. I am also going to guess that there is little overlap between those 500M users with HN's 5M users[2] or Kagi's 43K users[3].

So while those results-gated-behind-logins might seem annoying to us, we are in the minority.

[1] https://www.statista.com/statistics/463353/pinterest-global-...

[2] https://news.ycombinator.com/item?id=33454140

[3] https://kagi.com/stats


I remember a few years before Pinterest’s IPO, there were random statistics flowing around about how unique of a demographic Pinterest had compared to the rest of the internet. It’s mostly female dominated compared to the rest of the male dominated internet. Not just any female demographic either, but upper middle class, high income, young women †. There was simply no other online service that had that as its main demographic to the degree that Pinterest did. I remember thinking that must be a really attractive target for online advertisers and maybe I should keep an eye out for the IPO. Luckily I forgot. Its stock has been mostly lackluster.

†: hell son that was during my dating years. I was going on a lot of dates with women specifically in that demographic. Once I got to know a woman I would research her on google while narrowing my search to Pinterest. Out of 20 or so women I recall there were only 2 that didn’t have Pinterest accounts. The rest had, very active, Pinterest accounts.


Interesting. I didn't even know pinterest is a legitimate site. I just assumed it is som SEO scam from looking at the website for a few seconds.


It's a great tool for finding inspiration and creating moodboards.

Tangentally, IIRC the setting for the next Elder Scrolls game was all but confirmed after some creative at Bethesda accidentally made public a Pinterest collection with weapons and armor from the medieval arab world.


Try it, you may like it. I use it for collaborative visual brainstorming with real life projects (e.g. fashion or photo ideas) and it works well for that purpose.


It's actually a fantastic site. I use it for bookmarking images every day for future projects.


Man I hate to say you're completely right, but you're completely right.

I remember years back being in some social thing where someone was railing on social media and pontificated "absolutely noone wants what Facebook is", and I grudgingly had to retort "you don't want Facebook, but a half a billion people (at the time) obviously do".

Read the room.


Many of us are reluctant users. There because our friends our communities are. I would love FB to go away.

I am the admin of two FB groups, both UK home education related (one about exams and qualifications, the other for single parents). I do it because I feel I should help the community (especially as the crappy commercial groups that target the same audience), which is made up of non-technically inclined, mostly middle aged women - i.e. FB's core demographic. I cannot remember the other demographic nos offhand, but its 95% women in the exams group, and 98% in the other (and both have 1% other/did not say), and the age profile reflects the fact that people have school age (or just over, in the 16 to 18 age group) children.

I had actually planned to try and push the community towards forums, but with the Online Safety Act in force that is not a risk I am inclined to take.

My cousin runs a family group where she shares photos and wishes people happy birthday. I would prefer her to use WhatsApp but she is the one running it.

A lot of my friends post family news, important things like births and deaths and weddings on FB.


Many of us are reluctant users.

I understand what you're saying, but my own observation is "many of us" == "a tiny minority". "There because our friends our communities are" is exactly it...our friends and communities don't care as long as FB and the other social media cesspools are the default. I too have tried to get various relations to not make me log into social media to be part of their life, and I get some variation of "oh, you weird little infosec guy, that's way too much work and it's good enough for all of us".


I like Pinterest a lot.

I find art, tattoo ideas, home decor, concrete/wood builds, clothing fits, wedding themes…

It’s like a one-stop-shop when you need some fresh ideas on design.


It also has a ludicrously smooth and fast UX and makes it possible to access the high resolution original photographs (up to 100+ megapixels) by modifying url parameters.


I use Pinterest a ton. What param can I use to extract the original res?


Images are served from i.pinimg.com/ and the url params can be changed from "[0-9]+x(\/[0-9a-f]{2}){3}\/[0-9a-f]{32}\.jpg" to "originals/(\/[0-9a-f]{2}){3}\/[0-9a-f]{32}\.(jpg|png)".


Awesome, thank you.


I'm surprised - simply because I never get Pinterest results on Google. Now admittedly most of my searches aren't the kind where Pinterest is likely to have relevant results, but even then, surely I'd at least see them _sometimes_. But I literally can't remember the last time I saw a Pinterest search result.

Unless, as you suggest, they take over Google Images but not text search results? I could believe that I use Image search sufficiently rarely that I wouldn't have seen a Pinterest result.


I only get Pinterest results when I'm searching for something generic enough, and in those cases, why not use an image from Pinterest. I don't really understand the hate.


It's a nightmare for finding the original sources of images. For example, I was looking for a new sink basin and doing some quick image searches for various styles.

All the ones I liked were pinterest posts with zero attribution. A reverse image search then just brings up dozens of ripped and reposted copies of that pinterest post, also without attribution.

It gets frustrating


I assume with the 'popularity' bias (probably not the right phrase) in the modern internet this is pretty much the future of search. Someone comes up with something cool, posts a pix, and someone else puts it on Twit/Face/Tube/whatever and it gets reposted over and over and over and since the original is some worthless peon as far as the algorithm is concerned you'll never, ever find them.


I wonder if that's something that can be addressed by embedding the right metadata into images/videos? Most people don't bother even checking e.g. Exif data (let alone stripping or otherwise altering it) when reposting content they find online.


I can't speak for every platform but when I was working with frequent photo posts, most in-camera or post-editing metadata was stripped out on instagram and facebook. Some smaller sites like Gab didn't seem to mess with it as much, but the bulk did. I wouldn't be surprised if all of the other big ones did, too.

It was incredibly disheartening to have no recourse to attribute my own work, other than to smear some gross watermark on it. The automatic removal of that metadata, along with AI image generation, are some of the reasons why I gave up on the hobby entirely.

It's incredibly hard and stressful to derive any sort of pleasure or interest from something when the second it's exposed to the internet, any sense of humanity you tried to attach to it is stripped away, burned, and commercialized for the monetary benefit of some ethereal financier. It's the sound of an invisible vacuum cleaner, whisking away any sense of joy or life you wanted to share with the world for common love; the death of sharing. For-pay hugs.


Pinterest kills meta and exif of all uploads.


Well fuck them and the horse they rode in on, then.


usually that's a good thing for privacy


From my experience it’s only because many times the original copy is gone. Without Pinterest, you wouldn’t even find it anymore.


Pinterest is always a dead end for me. I don't have an account, so I can't actually access anything that the link is taking me to. It's a giant turd in my search results.


The hate is because Pinterest is shallow but shows when folk are searching for depth. The engine can't tell and these results are tricksy and false.


Mostly because if you try and visit the page that the image was supposedly on, it won't actually be there.


Because Pinterest doesn't let you see anything without making an account and logging in.


And even if you are logged in, good chances you get redirected to some other useless page rather than the image you were trying to view. Or if you're not logged in, by the time you do get logged back in, you lose the original link and you're dumped on a random feed.


All of those top domains (of which the majority are pinterest) have more than earned it.


I'm one of the minority that loves Pinterest. Happy to have them in my results, honestly. A lot of times they are the only remaining source for a specific image that has faded from the rest of the Internet.


When I've been reverse-image searching for obscure things, sometimes it's been the only result (despite not being the original source). I'd rather see pinterest than nothing - but I suppose you can fix that with downranking rather than banning them outright.

Compared to the AI slop flooding image search results, pinterest is increasingly looking "better than the alternatives".


Because they sued.


Did Pinterest sue Google? Do you have a link for that?


I love that wiki.archlinux.org is a top-Pinned.

I'm firmly a Debian shop, but I find that the Arch Linux wiki often answers non-distro-specific questions that I have about how to do something on Debian.

(Especially since I'm usually using Xmonad without a lot of the "desktop environment" stuff.)


The Arch Linux Wiki has some awesome information, but it's a little too informal for it to be a go-to for me. It's information is incomplete, opinionated, and sometimes has a "works on my machine" sort of a vibe.

But sometimes it just has the answer you need in an easily digestible format. Top 10 source for me, but not a top 3.

-- Some nerd with almost two decades of distro-hopping experience.


Yeah, I don't expect comprehensive and definitive documentation from it, nor copy&paste answers.

But for getting pointed in the right direction about things that have been obscured by the desktop environments, and then left largely undocumented nowadays, the Arch Linux wiki usually points me in the right direction.

Much of it would be pretty confusing to someone who only wanted high-level documentation in terms of the Gnome Desktop or KDE Plasma, though.


> But for getting pointed in the right direction about things that have been obscured by the desktop environments, and then left largely undocumented nowadays, the Arch Linux wiki usually points me in the right direction.

Even as an Arch user, probably at least 80% of my usage of the Arch Wiki is just going through "Troubleshooting" list of previously seen issues and solutions for whatever thing I'm dealing with. I don't go in expecting that everything will work exactly the same for me, but over the years I've ended up with quite a few headaches solved by pasting the right line in the right conf from one of those sections.


Oh it's really cool that you've used the arch wiki in that way. I had already done Linux From Scatch before I ran across the Arch wiki, so I was already familiar with concepts like boot loaders, kernel modules, and daemons. I mostly used it to find some sane config file values.


Similar here (25+ years Linux experience, including making my own distro). Personally, I use the Arch wiki for pointers to the current way to do system-level things, which has changed over time, with kernel and userland (e.g., all the things systemd changed, and for various kinds of devices), and sometimes for applications (e.g., what programs are currently available to do some small thing).

Maybe the value comes not just because they bother to maintain a wiki, but that Arch Linux tends to select for above-average technical people, even more than Linux in general does. (Even if Arch people strangely don't run Debian goodness; but we benefit from a diversity of perspectives, even if they are unexplained. :)


I often wish there was a date attached to the articles there. I get that a wiki format is ever evolving and as a result there isn’t necessarily a meaningful date that could be added. However, unlike an article about Jupiter, carbon fiber or WW2 on Wikipedia, the date for when a guide was written about Power Management on Linux is very relevant. I often find myself trying to sort of deduce that from the history page, then I fail then I have to go look up if that information is in fact the most recent.


I agree and I have to tell newer users all the time the wiki is very valuable regardless of if you use arch or not btw


It's so good it's made me question if it would be easier just to switch to Arch (so all the help would be directly applicable)


The Arch Linux wiki is an absolute goldmine.

I briefly used Manjaro, Debian and Ubuntu before that, and now am primarily a NixOS user, yet I still find myself coming back to the wiki.


I'm not surprised to see w3schools.com up there. I haven't come across it recently because of shifts in what I do, but it used to come up so often when I was looking for coding documentation. It was almost always useless.


If only it was merely useless. I know its reputation but it had some information that MDN did not have, so I used it this once in recent years. Turned out, the information was simply wrong and so I made a wrong decision based on that. (It might have been about favicon format support in different browsers. Presumably it was Safari that never had support for vector graphics whereas w3schools listed it as such, and it's not like you can just download Safari to double check.) Regardless, what I'm sure about is that I alerted them to whatever the problem was, but for me it was the last nail in that coffin


Oh yeah, I always had that plugin that removed it from Google results. What trash.


It's better than it used to be, but I've long since had muscle memory for appending "mdn" to any webdevy or javascripty search.

I wish geeksforgeeks would die in a fire, no offense to its operators.


Two different views for logged-in users and public. I realize that it has a different view from a different non-logged-in browser (or incognito).

I stopped wearing T-Shirt swags from companies quite a while back. Recently, I thought of promoting Kagi and wore the T-Shirt they sent at a few meet-ups, office spaces with lots of tech-people and no-one recognize it. A few of them thought, when we talked, if the logo is for a Golfing group/community!

Personally, I was thinking I’m proudly promoting something akin to ‘Wearing Google T-Shirt in 1999’ but this time, “Humanize the Web.”


Such a huge miss not including the name of the company on the shirts. It could be a fitness logo for all anyone knows.


First rule of marking. Say your name, loud & proud! Bragging doesn't take care of itself. Gotta be like the Beastie Boys, say your name 3+ times in every song.


Bleh, that attitude ruins merch. If all you care about is maximizing the number of walking billboards then you're satisfied because it's impossible to calculate the ROI of making available designs that people actually want to wear in public but reduces walking billboards.

The only tech merch I've ever worn was back when firefox had a good logo, goodwill, and no text on the front of the shirt.

Lots of people just want a simple graphic tee and IFKYK. It's not sports. Bands can sometimes get exceptions, but often approach wrestling tshirt levels of gaudiness.


You’d be right if it wasn’t a search engine.


At the airport, on the way to a Linux conference, I wore the Kagi shirt and people initiated conversations due to it.


I don't understand. Opening the page in a logged-out state, it looks the same to me (just that there are not buttons to pick whether you want to raise/lower/block a domain when you're logged out). What's the different view you mean?


When I subscribed to Kagi, my block/lower/raise/pin lists were very highly correlated with these aggregate ones.

It makes me think that for Kagi customers, search engine rankings optimize for something other than useful sites such as docs.python.org and cppreference.com


Google Ad team can dictate to the Search team. We have the leaks. Google has sold out.


Looks like alternativeto.net wins the most divisive prize, for being on the top boosted list but also heavily banned/demoted.

Runner up to the NYT


It's probably got to do with the bizarre design that hides the actual link to a productms website as a small element on the page and makes the most prominent links and buttons go to other things instead.


They're not very good at identifying what's actually an "alternative". By way of example, some of their top alternatives to OwnCloud include Dropbox, MEGA, and Google Drive. (All of which completely miss the point - OwnCloud isn't just a file storage service, and one of its key features is that it's self-hosted.)


I really like alternativeto. It's not always good: sometimes there are simply no good alternatives, or the community hasn't voted for the ones I'd have voted for and so a good option is way, way down. But if I want to know alternatives, I go there directly, so I guess that's maybe why people block it from appearing in random search results? I found it puzzling to see a useful site blocked (especially when I haven't seen it appear much in search results, but then, I've also been using DDG primarily, which ranks things rather differently)


Why is healthline so high on the block/lower list? (Excluding Pinterest)

IMO I like the fact that they link sources to their claims, which is very rare on the current web. I think of it as a somewhat trustable source of information. Am I wrong?


Healthline is not a trustworthy site- See Wikipedia discussion: https://en.m.wikipedia.org/wiki/Wikipedia:Reliable_sources/N...?


I think it's reasonably trustworthy for a popular health site, but if I used Kagi I would block Healthline myself. Their Top N lists frequently appear on top of my health related searches when I'm looking for something more scholarly.


I block them, but I block most common health sites. When I'm searching a health condition, I almost always only want Wikipedia. But Wikipedia's search is pretty bad unless you know the exact name of what you're searching for.


Could be because healthline blocks most VPNs and manifest-based ad blockers. Kagi users are basically HN users.


I have it blocked because I had them show up with auto generated (and, obviously, wrong) AI slop results. Anything like that gets an instant block.


Not the point of the post, but i’m surprised to see less than a million queries per day. Is that all of Kagi, or just some subset?


They only have 43,000 members (stat from the same page), so that's 20 queries a day per member, which doesn't seem that crazy.


So, only us HN members are members of Kagi.


You are (maybe) joking, but I’d guess there’s an extremely high overlap between HN users and Kagi users. I converted someone from a metal discord I’m in to Kagi with a promo gift last year, and he’s also the only HN user on the server ;)


“Many a true word is spoken in jest.”

;-)

https://en.wikipedia.org/wiki/Many_a_true_word_is_spoken_in_...


IIRC, nerds were the primary drivers of the search engine wars.


That said, 43k paying users is quite a bit of money, so if their team is small enough, that’s good enough, stable and steady income


Yeah, my understanding is that they are trying (and I think I remember succeeding so far?) to be revenue positive/sustainable from the beginning, and not trying to rely on "grow and make money later" that other companies used.

Which I prefer; I got burned by Neeva shutting down shortly after they started (for a pretty mild version of the word "burned", it cost me very little, but it was annoying).


What I gather from this the single decision Google could make to improve Google Search is not to give Pinterest preferential treatment as most people hate it (not the site itself, it has its uses, but the way it's promoted in Google Image Search results).


I feel like I haven't seen a pinterest link for ages, and I use google.


Interesting that TikTok is the 3rd highest on the block list (considering Pinterest links as one). I wonder why since it’s not often that I get it as a search result on Google . Is TikTok showing up as a result common on kagi or is TikTok just that bad?


Tiktok has a lot of deep-ish search keyword spam that doesn't actually lead to anything related to the query, just some generic videos and a pitch to install the app. It's basically useless as a website.


Tiktok is just really annoying, they rank high for a lot of keywords but the results will either just redirect you to tiktok's homepage or a bunch of "discover" pages with random videos that contain those keywords.

If you're on mobile you'll get asked to install the app, if you try to view a video on mobile you'll probably get asked to solve a captcha and again asked to install the app.

Say you search something related to repairing a bmw, you might get results that are "bmw mechanic", "bmw repair inclusive", "bmw maintenance meme" which just lead to those tiktok discover pages.


Requires login to actually do anything, like Twitter.

Making both completely useless as sources


You can get a login if you want.


What bothers me the most is getting an image result I want the full res version of, and being redirected to a discover page that doesn't have what I wanted


Likely because it requires a login?


This is convicing me to try Kagi for a while... search without pintrest would feel good.


Just as I suspected, fandom.com appears on the top "lowered" list.

So many games have moved to different wiki sites but because of SEO the Fandom.com wikis still appear near the top of "normal" search results, unfortunately.


I run a competitor to fandom (in kpop) and fandom isn't even the worst offender by a mile. Google has made my niche dominated by porn ads and underaged nude leakers for the past decade. Even gaming is completely fextralife enshittificated. Fandom at least has some professionals behind it.

The sad thing is we should be living in a golden age for these sorta websites, but Google has to promote dogshit so nothing cool gets built.


How pinterest got this big in the first place is only explainable by corruption/kickbacks within Google antispam team.


I don’t like Pinterest, but it’s easier to explain than conspiracy: a valuable demographic was underserved because people tend to overlook that women are half of humanity.

Mood boards do nothing for me, but that isn’t true of everyone.


Is there an opt-in for sharing such analytics/telemetry/statistics/insights (or whatever the euphemism treadmill is at now) -- or did all customers "consent" to this simply by using private settings "used by at least 20 users"?


Why should this require opt-in? I don’t see anything wrong with publishing interesting statistics.


I'm not seeing a way to opt-out on any settings page


I was confused why all the blocked results said “reset” until I realized I was logged in and had them all blocked…

I don’t recall using this list to populate mine but I think I must have!


Wouldn’t such a statistic cause a feedback loop?


That's a concern as old as search engines themselves. https://tomslee.net/2008/03/mr-googles-guid.html


These rankings have no effect on how Kagi ranks domains, so not in that sense. Maybe in the sense that a user will go down the list and also block the top blocked sites and uprank the most upranked sites, but that's just idiots being idiots.


I want to like and pay for Kagi but from australia it feels like it isn't behind a CDN.


That was my experience with them as well. There is a noticeable latency rendering the search results with Kagi as opposed to alternatives.


I guess if you could easily block sites as easily from the image search, too, at least alamy.com and dreamstime.com also would compete for the most blocked domains


I get the most value out of this feature by blocking fandom.com, fextralife.com and pinning quality wikis on gaming/nerd shit, fwiw.


Some people seem to be blocking `w3schools.com`. What is the rationale?


It has a pretty bad reputation from an earlier era when it was actively providing bad information [0]. That's largely been resolved from what I understand, but the reputation lives on.

People may also just find that they don't get a lot of value from it, but if that were the only reason I'd expect geekforgeeks to rank equally high with it or higher.

[0] https://www.w3fools.com/


For what came up to me when searching for Python info before I blocked it, they still have plenty of incorrect information or information that's useless for the tutorial page it's on


Is w3fools now defunct, or just broken on mobile? All I'm seeing is a message explaining that w3schools doesn't suck as much anymore.


Yes, that's all that's there now. It has a link to the archive version that detailed the sins of the old w3schools.


Ah, missed the archive link, thanks.


This will be controversial, but w3schools isn't bad for the niche it targets today.

It, however, has a long, well-earned reputation from being likely the worst site on the net for a decade, back in the early days. Even if I have a question that I think it would answer well, I still recoil at the thought of clicking a link to it.


It's a garbage site and there are far better non-profit web specification and utilization sites like MDN Web Docs. https://developer.mozilla.org/


It's targeting a different market. The MDN web docs are excellent reference material, but if you're learning or early in your career they're impenetrable. W3Schools has a reputation of over-simplification (to the point of being wrong sometimes), but that's not at odds with being useful to people at a particular stage in their career.

For another example, I gather that most of the high school chemistry I learnt is wrong, but it was mostly pretty useful in understanding the physical world as far as I needed to progress in my education at that age. W3Schools can be useful in that respect.


Education is a series of smaller lies. It's one of the corollaries to "all models are, wrong some are useful."

Some lies are actually better than the truth. My favorite example is that shooters are taught to imagine their sights being in line with their gun and that's they're shooting horizontally. But they're not, they're shooting in an arc so over short distances bullet just rises a little. Don't ask why. Why it's cool is that people using this model are better shots.


I've used w3schools a fair bit, but hardly for their web stuff.

I've used it for looking up SQL, XML/XSD/XPath and such when I had some specific question. Found it quite helpful, and still use it as a quick reference every now and then.



I haven't gone to that site in a long time, but back when I did work that was related to its content, it was somewhere between useless and actively bad. From reading this thread, it seems to have improved since, but it has a tarnished reputation. I use Kagi, and if w3schools had come up in a search I'd likely have set it to block before even considering clicking the link; it's got a tarnished reputation.


Content that kind of looks like technical docs, but without linking authoritative sources, date when it was last reviewed, software it was verified against etc.

Nowadays with AI slop websites all over the place, plus formerly legitimate newspapers not attributing sources & edits either, this may not seem that special. But w3schools earned their reputation when that was exceptionally rude. Also, straight up wrong all over the place. But that is more of a symptom. And to be fair a lot of the software back then was straight up wrong, too. Tough job documenting the reality of the WWW correctly while IE was around.


Seriouseats in the top pinned results. So dank


Why do people block ny times and Washington Post newspaper?

Is it because it is paywall or for political reasons?


[flagged]


I think it’s more likely that Kagi users value relevant search results & accurate information.


The anti-Pinterest political party


That's the only party I can get behind!


Every news source blocked generates more noise than signal as its main feature. I doubt that even a regular Fox News reader wants their results full of articles they’ve already read.


yes they do mate - yes they do… :)


I found that interesting too. Not a Kagi customer yet but I always feel the need when googling to be able to blacklist SEO spammers, and that would include Pinterest as an image SEO spammer.

But blocking political sites you don't agree with? I don't get it, it's not like they are polluting my search results.


> blocking political sites you don't agree with?

There's "don't agree with" and there's "website is known for spamming false information". Fox, breibart and daily mail all fall into the second category, whether you agree with their political ideas or not.

Edit: I wanted to link the original "things that cause and prevent cancer" list page, but it's not online anymore unfortunately https://web.archive.org/web/20140807144235/http://kill-or-cu...


I think there's another commonality in the 'lowered' sites that has nothing to do with political leaning. Especially for the top two lowered ones (foxnews and dailymail)

You don't go there for news. You go there for your daily set of outrage and panic.


They are news sites, not political sites. Maybe some people do not want them in their news searches because they consider them low quality, or for some other reason I guess?

I would def not consider sites like dailymail.co.uk a good source of information.


The Daily Mail and Fox News are more purveyors of fiction than anything else (just ask News Corporation’s lawyers if you don’t believe me: https://www.npr.org/2020/09/29/917747123/you-literally-cant-... ). Many people, on searching for current events, prefer news to news-based fanfic.


How would you be able to tell if it wasn't that some of the search results have such an obvious leaning that they would not be considered useful search results?

Other than that there is also a chance that certain websites have a high hit rate because of the words used in the articles or the amount of articles put out. If certain publications clutter your results they are likely to get banned.


What direction was the skew?


foxnews.com is near the top of the block list.

Of course, Fox News has made the legal argument that people shouldn't believe what they say (https://www.npr.org/2020/09/29/917747123/you-literally-cant-...), so it's not surprising that people paying for a service to look for actual information would block them.



There's a couple "left leaning" media sources in the boost list and 3 "right leaning" in the lower list. There's none going to the other way for sources and their generally acknowledged "lean".

I interpret that as a skew.


Skew in the user base at least


That's natural. Kagi users are mostly techies and techies skew left.


[flagged]


How do you mean? I’d love to know more about your theories!


[flagged]


I don't think the rank and file employees are going to blindly follow where their CEOs go on this.


I mean the other way to look at it is that there are two quality newspapers plus the BBC in the boost list, and a notoriously dodgy tabloid, a TV news site, and whatever the hell Breitbart is (I must say, I hadn’t realised it was still a thing) in the lower list. You’re not really comparing like for like.


I found it interesting too, not sure why you're being downvoted.




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: