It's not you, it's SERP
January 17, 2024 10:50 AM   Subscribe

Research confirms that search is getting worse. We all feel it. Some scientists have measured it. "We can conclude that higher-ranked pages are on average more optimized, more monetized with affiliate marketing, and they show signs of lower text quality." Link above is to an article on The Register. Link to original paper.
posted by moonmoth (95 comments total) 39 users marked this as a favorite
 
We believe the issue of advertising causes enough mixed incentives that it is crucial to have a competitive search engine that is transparent and in the academic realm.

— Sergey Brin, Larry Page, taken from The anatomy of a large-scale hypertextual Web search engine
posted by bombastic lowercase pronouncements at 11:09 AM on January 17 [37 favorites]


Accelerationist morality tale: Search Engine company makes ML model, ML model is used by spammers to make higher-ranked spam pages, Search Engine company goes bust.
posted by k3ninho at 11:10 AM on January 17 [6 favorites]


In case you, like me, had to google the acronym: SERP = "search engine results page." (Oh, the irony.)
posted by minervous at 11:11 AM on January 17 [24 favorites]


And it isn't limited to regular old search, either. In my work, I have to occasionally have to avail myself of doing a reverse-image search and, hoo boy, have results gone downhill in the past year, or even just the past six months.
posted by Thorzdad at 11:12 AM on January 17 [19 favorites]


I really wish I were still an insider at Google to hear the internal conversation. Historically the company kept very close watch on search quality. I doubt that's changed but maybe the focus has.

For awhile I'd been optimistically assuming maybe their own internal quality measures were saying things were fine despite my own experience and that of many others. But if a small university group can measure search quality failure this directly I have to assume Google's team of lots of people with vast resources whose only job is measuring quality can too.

My second thought is maybe Google realizes there's a problem but doesn't know how to fix it. That's a scary thought. In particular they now dominate the English search ecosystem. In so many ways being a successful website now means "be good at SEO hacking Google", maybe there aren't any real quality signals left.

The third possibility is Google knows this is going on and doesn't care. Because the executives and business are focused on the ad business instead.

Google's been here before. In 2011 Google was awash in affiliate spam. The Panda update changed things radically and improved quality for many years. If we're lucky maybe they've got something similar coming. I have no evidence they do though.

There's interesting competition now. I like Phind as an LLM-enhanced research search tool. Microsoft Bing is remarkably interesting right now too, they are being very aggressive in adding AI capabilities and it's often useful. I hear lots of good things about Kagi but I haven't tried it. I've never liked DuckDuckGo but it's popular.

I wonder if quality on the Chinese search engines has had a similar problem recently. They've got an entirely different ecosystem and market dynamics but excellent technology and investment.
posted by Nelson at 11:20 AM on January 17 [44 favorites]


But I love scrolling through sixteen pages of meaningless drivel and affiliate links to read a four ingredient recipe that doesn't even work.
posted by 1adam12 at 11:21 AM on January 17 [27 favorites]


The fact that I can currently tell within two sentences whether a page was content-milled virtually guarantees that an LLM can be fine-tuned to pick up on whatever that je ne sais quoi I’m twigging to is, and filter it. The fact that Google hasn’t already done this means they are making money hand over fist on SEO bullshit spam, to the point where reputational damage is brushed off as a “for later” problem.

I want to predict that within a year or two correctly-tuned LLMs on the SEO side will be sufficiently integrated into the content mill pipeline that I will no longer be able to distinguish spam. But maybe not; it could be that scummy spammers already know people who can distinguish scummy spam are not their target audience, and can’t be bothered chasing the vanishingly slim hope of getting me to actually purchase anything.
posted by Ryvar at 11:22 AM on January 17 [18 favorites]


Der boffinß should do Amazon next.
posted by gottabefunky at 11:39 AM on January 17 [4 favorites]


Dump Google Search, once you leave it you will wonder why you didn't do it sooner.

DuckDuckGo has been my pick for years, and is the top pick among alternatives listed on alternative.to. It doesn't track you, or bubble your results, and bang commands are really handy. You can even go to DDG and type "!google is duck duck go a good search engine" to search right on Google.

If DDG is for some reason unappealing, Bing! is pretty nice as well. But srsly, get off the Google, it has nose-dived from what it once was.
posted by mcstayinskool at 11:42 AM on January 17 [17 favorites]


Google is organizationally sclerotic and addicted to throwing their most productive developers at The Next Doomed Project. Even core products like Maps are unusable now. I don't think they're going to fix this in a meaningful way.
posted by phooky at 11:44 AM on January 17 [10 favorites]


In a former life I worked for a large SEO company, and I remember wondering how long Google could keep ahead of the cat and mouse situation. So much of the content our Content Creator* was writing for clients was largely meaningless keyword-loaded drivel. Now it's almost an order of magnitude worse than it was then.

*I always felt bad for her, as I think she was a talented writer who was stuck writing mostly nonsense.
posted by BigHeartedGuy at 11:48 AM on January 17 [5 favorites]


> Google's been here before. In 2011 Google was awash in affiliate spam. The Panda update changed things radically and improved quality for many years.

was that the one that murdered metafilter?
posted by bombastic lowercase pronouncements at 11:51 AM on January 17 [17 favorites]


Obviously they can pivot to AI and, while not making results less garbage, at least make them in house generated garbage that is chasing tye hype of the moment. Line go up!
posted by Artw at 11:52 AM on January 17 [1 favorite]


anyway, to bang on my standard drum: search is an inherently bad method for surfacing web content, and if all attempts to devise an effective search engine fail due to chatbot spam and associated fuckery it will on the whole be to the world's benefit.
posted by bombastic lowercase pronouncements at 11:54 AM on January 17 [2 favorites]


Last month I wasted an hour when I tried to look up something in an area I'm broadly familiar with, but to do something a little esoteric that I'd never tried (in this case, setting up a virtual environment in Python to install a package that doesn't currently run in the most up-to-date Python version). I found a page on the first page of search results, read the instructions which seemed reasonable and complete, and then banged my head against the issue until I realized I'd been snookered by a confidently expressed but completely incorrect and impossible step which only an LLM could come up with.

I'd difficult to express how annoyed and horrified I was by this. Getting accurate information about Python, of all things, out of a Google search didn't used to be a problem. In conclusion, we should bring back gibbeting.
posted by figurant at 11:56 AM on January 17 [40 favorites]


Our kitchen sink was leaking and there were a couple of things in our faucet design that were slightly off from the general "how to" guides and my god, trying to wade through the SEO garbage to figure out how to get at the cartridge so we could take it off and replace it was maddening. Just, page after page of stuff formatted as Q and A lists of phrases people would search for.

I finally found a YouTube video that matched the model when we got to something that had a part number but of course, I had to click on a bunch of videos that were of slightly different faucet models and watch a bunch of ads first. And then, trying to search for the part number on the cartridge and figure out if there were equivalent parts took us back into SEO garbage land. We ended up going to Home Depot and just wandering around the aisle looking for the part, which they had, luckily.
posted by damayanti at 11:56 AM on January 17 [21 favorites]




Is there a half decent replacement for Maps? I was in an unfamiliar city for a few days recently and Maps was atrocious at finding a place to eat - it wanted me to drive 45 minutes for vegetarian food when there were plenty of options within a quarter mile. It's not just about how they're ranked in the results; it wouldn't even include pins on the map for most places (unless I zoomed way way in, but that's not helpful if you don't already know which block you're looking for).
posted by echo target at 11:58 AM on January 17 [6 favorites]


search is an inherently bad method for surfacing web content

serious question: if not search, what is the effective alternative for surfacing web content?
posted by mcstayinskool at 12:04 PM on January 17 [10 favorites]


seconding that Maps question... tired of Google Maps considering street names a tertiary priority to letting me know about businesses I don't care a whit about
posted by kokaku at 12:04 PM on January 17 [57 favorites]


Getting accurate information about Python, of all things, out of a Google search didn't used to be a problem.

When I search for basic python stuff, the official python documentation is no longer even on the first page of Google results. I keep on thinking I've forgotten the method name, but no, it's just not there.
posted by Hermione Dies at 12:08 PM on January 17 [18 favorites]


I wish I could suggest Apple maps but they have idiots in charge of visualization and are essentially unreadable. Remember when you could buy paper maps in a clench-bound book printed in black and white and you would know exactly where everything was? This is the complete opposite of that. Open Street Map is probably the least bad.
posted by seanmpuckett at 12:10 PM on January 17 [7 favorites]


The third possibility is Google knows this is going on and doesn't care. Because the executives and business are focused on the ad business instead.

Somehow Kagi manages to remove a good deal, though not all, of the SEO spam from search results. And they are even using Google as one of their providers. So I am sure Google could remedy this if they wanted to, but Googles customers is the advertisers not the users, so their priorities lies elsewhere.
posted by winther at 12:10 PM on January 17 [3 favorites]


trying to wade through the SEO garbage

You might be interested in - A storefront for robots: The SEO arms race has left Google and the web drowning in garbage text
posted by airing nerdy laundry at 12:12 PM on January 17 [6 favorites]


Google sucks for sure, and I am way happier with DDG. In my own personal funny little metric though, Google sucks a little less now than they did some years ago. Back then, if you searched google for "elytra", Minecraft content filled the whole first page. You wouldn't even know it was a term that existed outside of a video game. Now they have at least managed to put scraped content from Wikipedia's article "elytron" at the top.

Any current examples to share of particularly bad Google search results?
posted by SaltySalticid at 12:18 PM on January 17


It's not even hard to remove the SEO spam. Just look for affiliate marketing, trackers, and advertising on sites and give any page with it a great big downvote in the relevance-ranking algorithm. The more cruft, the bigger the downvote.

THIS ISN'T HARD.

But Google doesn't make money by running a search engine. It makes money by advertising and user surveillance. Including on SEO spam pages. That's a conflict of interest (*points up to first comment in thread*) and Google has decided that its search engine loses.

I'm curious whether downvoting cruft is what Kagi does -- any intel, winther? If it does, I'm a-switchin' from DDG, which I currently use but which still has significant SEO-spam problems.

"But it hurts small businesses on the web!" Probably not worse than getting elbowed out of search results altogether by LLM sludge.
posted by humbug at 12:22 PM on January 17 [10 favorites]


I just wish the Google would honor its filters, mainly the minus/dash which is supposed to filter out any results containing the dash's argument. In my experience this just stopped working sometime during the Obama administration.

DuckDuckGo has been my pick for years

Mine as well, sure we don't like how Google tracks us but when comparing basic search results I'm not seeing much difference between the two, these days. Really dislike how the Duck limits results to a single page, then you have to scroll past the "Searches related to" section and click More Results; also I find Google's maps the winner because more detail. And the Duck ignores my attempts at filtering, just like Google.

Now, Google Advanced Search is still useful (I use it most often for a specific domain). And this just in -- Lycos is still around! I'm using it again, hoping for improvements.
posted by Rash at 12:28 PM on January 17 [15 favorites]


Good timing for this post, as Google laid off over a 1000 people last week and then yesterday "hundreds" more from their sales team. Clearly all is not well in Do No Evil town.

I've mainly transitioned over to using Ecosia, which donates money to planting trees throughout the world every time you search. It's really just Bing in disguise, but I like the ad-free interface.

Speaking of ads, I noticed a few months ago that Google made a decision that certain search terms should funnel you directly into a shopping interface. Here's a screenshot of what I mean: when you search for the phrase "swing arm wall lamp", this is what you get.

How they decide which terms get this shopping treatment, I have no idea, but the wild thing to me is that the "traditional" search results are buried so far down the page that they might as well not exist. So if you were anyone who simply wanted to understand what a swing arm wall lamp was, there is no definition page aside from the ones trying to sell you one.

And the icing on the cake is that even after scrolling down, the very first "real" result is not even correct! That result is a "swing arm wall LIGHT", which may functionally be the same but it is nowhere near the best answer.
posted by jeremias at 12:32 PM on January 17 [9 favorites]


I just wish the Google would honor its filters, mainly the minus/dash which is supposed to filter out any results containing the dash's argument. In my experience this just stopped working sometime during the Obama administration.

I don't think Google does site-specific searches anymore, either. I haven't been able to get any results in a good long time, anyway.
posted by Thorzdad at 12:36 PM on January 17 [1 favorite]


If DDG is for some reason unappealing, Bing! is pretty nice as well.

If this is the same paper that was circulating yesterday, it confirms Google sucks but says Bing and DuckDuckGo suck more.
posted by dobbs at 12:40 PM on January 17


I've certainly found times when DuckDuckGo results are utterly abysmal and Google results are merely slightly shitty by comparison.

I'd consider signing up for Kagi, but I have no idea how many searches I actually perform in a month, and if I started using the Starter thing I'd probably suddenly find myself averse to searching for stuff. Or using DDG except when it's bad and then switching to Kagi and not really getting my money's worth out of it. I shouldn't have to think about it that much. Plus it seems like a downer to start subscribing to get a service that used to be both free and pretty good.
posted by Foosnark at 12:43 PM on January 17 [4 favorites]


Money you send to Kagi at least goes to the people who are de-shittifying Google's results using the same API, something Google is unwilling to do. I think it's a worthwhile endeavor, and the use case for me is definitely to have it as my default search engine for the dozens or hundreds of searches I might do in a day, depending. The cognitive load that's removed by not having garbage injected into my search results is substantial.

To inject some levity instead, the glorious apex of google's usefulness can be summed up by this image.

Years after coming back from holiday in Mallorca, I remembered a song emanating from the clubs there that I didn't recognize. It sounded like the lyrics were describing a very unfortunate, but nevertheless giggle-inducing beekeeping incident.

Turns out, "THERE'S BEES ALL OVER MY BODY" might not be the actual lyrics of the song.
posted by tigrrrlily at 12:48 PM on January 17 [12 favorites]


I don't think Google does site-specific searches anymore, either. I haven't been able to get any results in a good long time, anyway.

Absolutely it does. That's how you find information these days (by appending site:reddit.com to your search). This is necessary because of everything above. It's impossible to find information about a product or service without pages of LLM-written trash and, yeah, Google has obviously chosen to do nothing about it, because it's impossible that they haven't noticed.
posted by uncleozzy at 12:56 PM on January 17 [11 favorites]


I've got Kagi, but when I wanted to find the list of things a soldier must not do, I found it with Google, not Kagi.
posted by Nancy Lebovitz at 12:59 PM on January 17 [1 favorite]


Using kagi with the site:amazon.com, site:etsy.com, etc. filters has solved many a logistics problem (#prelarpfreakout). Although I do have site-specific block filters for the garbage injected by those sites' internal searches, as well.
posted by tigrrrlily at 12:59 PM on January 17


oo00 I thought lemme try Lycos again

its telling me its 20F in Helsinki... what does F even mean?

Google and Yahoo used to do this too around 15 years ago when I was flying between SF and singapore frequently enough to notice how American the world is when seen through the lens of Silicon Valley

Welcome to our hegemonic billionaire lords
posted by infini at 1:04 PM on January 17


Confirm that the Google site: operator still works, but negation and exact match have been buggered.
posted by Aardvark Cheeselog at 2:09 PM on January 17 [8 favorites]


I have been experimenting with DDG as my default search and found if to be, if not hugely better than Google not actually all that worse? Which frankly says things about how far from grace Google have fallen given their old level of superiority to anything else. Also not having quite so much sponsored content and bullshit before you get to the first result is refreshing.
posted by Artw at 2:16 PM on January 17 [1 favorite]


Panda update... was that the one that murdered metafilter?

No, first of all "murder" implies intent; I think Metafilter was collateral damage. Second, the dates are wrong. Panda rolled out starting Feburary 2011. Metafilter's dive in Google referral traffic started in November 2012.

Has there been a good postmortem ever written on what really happened? My memory is the answer is vague. This article from 2014 has Matt Cutts (at Google) explicitly saying it wasn't Panda but something else. I'm having a hard time researching this further because Google is so bad for this kind of search anymore. And I don't have the patience to wade through 10 years of Metatalk.

The Metafilter damage is a good warning for how hard it is to manage search quality through algorithm changes. My guess is they made some changes that started penalizing link farm type sites, things with lots of text and interlinks that SEOs were using to artificially boost their standings, with a heavy side-dose of affiliate links generating revenue. Metafilter looked very much like one of those sites, structurally, the difference being we were sincere and wholesome in what we were doing and not just spam. But the algorithm failed on that.

The problems with Google today seem cruder. It's very focussed on recent pages. And on very wordy pages that nevertheless seem to express one nugget or fact in just the right format to fit in a search result snippet. SEOs build whole sites structured to target that.
posted by Nelson at 2:31 PM on January 17 [9 favorites]


Is there a half decent replacement for Maps? I was in an unfamiliar city for a few days recently and Maps was atrocious at finding a place to eat - it wanted me to drive 45 minutes for vegetarian food when there were plenty of options within a quarter mile. It's not just about how they're ranked in the results; it wouldn't even include pins on the map for most places (unless I zoomed way way in, but that's not helpful if you don't already know which block you're looking for).

For your particular use case, Happy Cow!
posted by hydropsyche at 2:32 PM on January 17 [2 favorites]


I was in an unfamiliar city for a few days recently and Maps was atrocious at finding a place to eat - it wanted me to drive 45 minutes for vegetarian food when there were plenty of options within a quarter mile.

One time we were trying to find gluten-free restaurants that were on our route, that were NOT IN SAN FRANCISCO, which we were briefly forced to pass through/over on the way out of the South SF area. It just gave me tons and tons and tons of San Francisco for like an hour, even after we'd passed out of San Francisco/the bridge area.
posted by jenfullmoon at 2:33 PM on January 17


Since the paper is specifically about the problem of searching for product reviews I took a look at how the researchers categorize some of these review sites. This is a pretty nitpicky thing, but I find it strange the researchers put OutdoorGearLab in the category of
"Review Content Farms producing low-effort product listicles, pseudo-reviews, and buyer’s guides in large quantities, but with (superficial) editorial content on the side"
While I wouldn't rely solely on that site if I were in the market for a new bike helmet, for example, it seems quite a stretch to say they put out low-effort, superficial content. I mean, these are verifiably real people doing fairly substantial comparisons, with links to more in-depth reviews of each product. What are the researchers seeing here that I'm missing?
posted by theory at 2:43 PM on January 17 [5 favorites]


The third possibility is Google knows this is going on and doesn't care.

Having worked at Google, I'm pretty sure they care, but there's a caveat. All the measurements i recall hearing about were relative to their competitors. So, if, as the article notes, Bing and Duck Duck Go are impacted, but Google is still better, they won't be as motivated as they would be if one of the other search engines was kicking their butt.
posted by CheeseDigestsAll at 3:04 PM on January 17 [5 favorites]



Is there a half decent replacement for Maps?


I actually googled "Albany NY" as part of the thread the other day, and Google refused to even show me an option for an actual google map of Albany without another click. It showed me the minmap but that thing is worthless and clicking on it just made the jpeg bigger. I don't know if they were doing some kind of in prod testing thing because when I google now, it's got the option, but that shouldn't be hit or miss. I've defended them a lot in the past, but that's just a terrible user experience.
posted by The_Vegetables at 3:21 PM on January 17 [1 favorite]


Google's been here before. In 2011 Google was awash in affiliate spam. The Panda update changed things radically and improved quality for many years.

I'd just like to remind that Panda also severely hurt Metafilter's findability in search, and caused ad revenue to tank.
posted by JHarris at 3:51 PM on January 17 [4 favorites]


> Since the paper is specifically about the problem of searching for product reviews I took a look at how the researchers categorize some of these review sites. This is a pretty nitpicky thing, but I find it strange the researchers put OutdoorGearLab in the category of

"Review Content Farms producing low-effort product listicles, pseudo-reviews, and buyer’s guides in large quantities, but with (superficial) editorial content on the side"

While I wouldn't rely solely on that site if I were in the market for a new bike helmet, for example, it seems quite a stretch to say they put out low-effort, superficial content. I mean, these are verifiably real people doing fairly substantial comparisons, with links to more in-depth reviews of each product. What are the researchers seeing here that I'm missing?
Having some passing familiarity with the site (mostly because I've wanted to buy some things that site would cover in reviews, but also because I've worked in this space before) I think the reason why it gets classified as a content farm is because... it kind of is a content farm.

In terms of site structure, OutdoorGearLab largely matches the description, except perhaps the part about "low-effort" (which we'll set aside for now as that's a subjective judgment). Listicles and buyer's guides make up the overwhelming majority of their homepage. There do appear to be individual reviews of items but it's hard to tell if they're actually just repurposed from group comparison reviews or independently tested and then scored in such a way that allows for easy comparison between products. Content is used repeatedly across multiple pages (the same product review text will show up in an individual product review as well as multiple listicles/guides).

The only real point of contention is whether the content should be considered low quality and/or low effort. The paper's authors note that there were several disputes between the pair about how to classify certain sites. None of the disputes apparently involved whether to classify a specific site as a legitimate review site or a content farm, but nevertheless they did note that it showed the difficulty in trying to distinguish quality content from poor content in the space.

Anyone who's ever read a "Top 10 [PRODUCT CATEGORY] As Reviewed By Amazon Buyers" article knows why this happens: there are too many products to review and not enough resources given to sites to properly review them all. So a lot of content gets generated using shortcuts: summarizing Amazon customer reviews is a common tactic. Often you don't even get that; instead it's just the product DESCRIPTION pulled from Amazon instead. Google introduced several algorithm updates in an attempt to distinguish between reviews that contained first-hand experience with products from ones that just regurgitated information from other sources, and you can see some of the adaptations sites like OutdoorGearLab have made in an effort to be recognized as the former. For example, all the product photos in this bike pannier article are credited to the author of the article itself, indicating that they didn't just pull the photos off a store or manufacturer website, they really had the product in hand at some point. Having worked on similar sites in the past, I can tell you that photo credits became something we specifically thought about as a way to improve the authority of an article when improving sites.

But because everyone's following Google's playbook in this regard, it also means all these sites increasingly have the same structure and grab the same low-hanging fruit in terms of site content and structure. If I take a photo of a product, does that automatically mean my review of said product is going to be good? It certainly weeds out the most obvious content farms that aren't even touching a product, but besides that it's hard to say. And because sites are incentivized to slice and dice their content in as many ways as possible to rank highly for as many important search keywords as possible (the definition of which changes from month to month, of course), the overall effect if you're someone who's actually reading the site on a regular basis like it's the monthly issue of a magazine will be one of being spammed with the same content over and over in slightly different forms, even if that content was originally the result of a comprehensive first-hand review.

So is OutdoorGearLab a review content farm or not? Honestly, I'm not even sure where the line between "legitimate review site" and "review content farm" is anymore.
posted by chrominance at 3:54 PM on January 17 [11 favorites]


Sergey Brin, Larry Page, taken from The anatomy of a large-scale hypertextual Web search engine

Can't really blame them or the employees. If billions of dollars were thrown at you, or if you're a worker getting millions in stock options, what would you do?
posted by They sucked his brains out! at 4:16 PM on January 17


> serious question: if not search, what is the effective alternative for surfacing web content?

word of mouth.

nota bene: not kidding
posted by bombastic lowercase pronouncements at 4:45 PM on January 17 [6 favorites]


"effective alternative for surfacing web content?" - for my needs - welcome to metafilter.

which is probably the tech equivalent of "word of mouth"
posted by Barbara Spitzer at 4:53 PM on January 17 [5 favorites]


another way of saying word of mouth is "link aggregator"
posted by I-Write-Essays at 4:59 PM on January 17 [5 favorites]


Bad search is thieving. It is thieving the non-renewable, plentiful (everybody has it) but unknowable individual's amount of time. If I have to wade through crap search results, give mental resources to determine the validity, possible bad information outcomes and still have the poison of doubt about the information, then that is thieving my time. Considering the individual's loss of time and then the larger population base, we are talking about a lot of stolen time.

I exchange my privacy, to a certain extent, with the search engine in exchange for more effective use of my time because doing research the old way is time-consuming.

But this whole need to monetize every moment, interest, taste or curiosity is wearing. How many side hustles do you have? How is arbitraging your circumstances working out for you? Was that listlicle correct that whatever the well-reviewed Amazon product changed your life?

TL; DR

Time is too precious to waste on bad search.
posted by jadepearl at 5:06 PM on January 17 [9 favorites]


I tried the Duck Duck Go route, but I found that results were different, but not necessarily better. Google serves up a ton of garbage for sure, but DDG served up some... more questionable results? I felt good about using it, but for finding specific restaurants, phone numbers, simple stuff, DDG was more cumbersome. And that's most of my web searching. Looking up movies, finding contact info, etc.

Maybe if I was using the web more academically I'd have a different opinion. I ended up switching back to Google because it's the devil I know. I can see the bullshit links (mostly) and know how far to scroll (usually). Just an amateur user's experience (though I have been online since 1994).
posted by SoberHighland at 5:26 PM on January 17 [2 favorites]


It took me almost a quarter way through the new century to say this, but Bing is surprisingly not bad. I agree with you that DDG can be a bit mixed, but it is the privacy trade-off.
posted by jadepearl at 5:35 PM on January 17 [2 favorites]


So is OutdoorGearLab a review content farm or not? Honestly, I'm not even sure where the line between "legitimate review site" and "review content farm" is anymore.

It’s a continuum, I think, but I haven’t seen outright sponcon, either openly (better) or obscured (worse), and I’d say they’re solidly on the more-legit end of things.
posted by box at 5:36 PM on January 17 [1 favorite]


highly recommend Kagi to anyone considering trying it. I considered myself a heavy search user and was worried about running out of searches and having to pay extra for more. But DuckDuckGo results have become such hot garbage in like the past 3-5(?) years that I was desperate for something decent. I signed up for a big Kagi plan a while back and they more recently adjusted their pricing down a bit and moved my plan tier to unlimited. But even though I now have that level of doesn’t-matter-just-smash-that-search-button-again luxury my numbers are still way under the amount of searches I would get if I was paying for the cheaper plan (you can see your daily search counts on your account page). I recently enabled it at work where I’m searching constantly (programmer with bad memory) and I’m still way under. Plus they have a feature where you can add bangs at the beginning of your search to use a free engine for stuff that should be dead easy like word definitions, e.g. “!ddg SERP acronym”. So I could totally downgrade my plan but I honestly find it to be worth what I’m paying. Lots of sites in the results are still a disaster of ads when you click through- particularly recipe blogs- but that seems more reflective of quality of the large-websites internet as a whole at the moment. I don’t really know anymore how I could approach finding things on the internet without a search engine, but I’ve decided that I’m lazy and ultimately I’m happy to pay for a decent shortcut.
posted by crime online at 6:32 PM on January 17 [6 favorites]


Google Maps is still excellent for what I mostly use it for: finding the optimal route from point A to point B, factoring in departure and/or arrival time and transportation mode.
posted by Artifice_Eternity at 7:03 PM on January 17 [6 favorites]


Bing is surprisingly not bad

Bing has always been pretty good. It's a big part of why DuckDuckGo has been good too. But every time I tried to switch to it, it suffered from being different. Turns out I'm highly symbiotic with the Google UI. Also it was never really better than Google, just nearly as good.

But Bing is doing a lot of interesting stuff right now with AI integration. They're investing heavily and it's definitely worth a look.

(TBH, Bing's main purpose has been to be just plausible enough of a competitor that Google pays enormous amounts of money to maintain their monopoly. Mozilla exists almost entirely on Google's payments. Google pays Apple $18B a year.)
posted by Nelson at 7:20 PM on January 17 [1 favorite]


So among my other ghostwriting gigs, I write SEO content for a bunch of places.

I think the first thing I really have to say here is that if you are expecting to make a living off a website - and that goes for little junk sites to big old school publishers like the New York Times or whatever - there is going to be a need to drag eyes to your site. They're all content mills, it's just the bar for publishability is higher for sites with a reputation. Higher reputation sites also tend to have other avenues of income - for example I spend a lot of time reading University Extension Service blogs from the States, especially the excellent University of Florida Extension Service. They don't need to get those clickthroughs to stay afloat.

The minute advertising and human attention became the new currency of the web, any push for accuracy over entertainment was doomed. Doubly so once the volume got to a point where it required an algorithm to curate. There just isn't enough human attention out there to curate things any more. It's a huge problem for people trying to run businesses, too. Those shitty Maps results echo target mentioned up thread is bad for business owners as well as consumers. If you are some poor schmuck just trying to run a dumpling joint you may as well be doing it from under a bucket, as far as Google is concerned.

I'll note that there is currently a push away from LLM use on SEO sites because Google can spot the output of those models and downranks the results. There was a spate of job listings around September for content creators who could pump out hundreds of articles a week using AI; those have dropped off completely and we're back to trawling for English speakers who are willing to pump out drivel at like, 0.006c a word NO AI in big capitals three or four times through the listing. Most of the content that reads like it's written by an AI not because it's LLM stuff but because it's the result of some poor grad student in a former British colony sidehustling at three am or the like. People who don't have the greatest grasp of the language but who are living in places where $6 an article is a not unreasonable renumeration for an hour or two behind a keyboard.

SEO at the production end is wild. On one side you have Google, constantly tinkering with the algorithm in inscrutable ways. Changes are often made that severely impact site traffic but for no discernable reason, and often seem to prioritize Adsense customers ahead of other mills. Sometimes techniques that rank well will turn to shit overnight. There was a question/answer format that was popular a year or two ago that I thought was a pretty good approach (so if you searched for "how to repot african violets" the top search would have the question and answer in bolded text in the first 50 words; I approve of this much more than "do 1000 words of waffle with the actual information load at the bottom after fifteen ads) but it fell out of favour and it's all skyscraper list building now (Top 11 Tips on How To Repot your African Violet) which is less useful to a reader but seems to rank better - and drives an arms race for just how many tips you can produce. This one is utter horse-shit and I am pretty sure it's got a shelf life of a few months more at most but it's still being spruiked by SEO gurus. I often tell clients that SEO is basically witch-craft, and that you just have to throw things at the wall and see what sticks, and be prepared for it not to stick in a month.

On the other side you have the side owners, almost none of whom write their own content. The really dreadful mills are usually just envisioned as revenue drivers and are increasingly commercial. One of my first clients was an agronomist who set his site up to provide a revenue stream while he was finishing a master's degree and people like that are increasingly rare. Most will have another income stream or a whole fleet of shit sites that cover as many "underserved" niches as possible.

Then there's the gurus.

There's an entire industry around SEO that's chock full of snappy dressed motherfuckers shilling classes, software, assistance programs, the works, all around the idea that SEO is easy money and all you need is some cheap freelancers and their special magic brand of software and you'll be rolling in money in 18 months. I've dealt with a handful of their victims and they all have a very rosy idea of how easy it is to game the algorithm, especially if you use their Very Expensive But Totally Foolproof Optimisation Software. I've had freelancer licences for a few of them and they turn the whole thing into a writing-by-numbers exercise and the very best gamify the whole thing, with a menu of keywords at one side that change colour as you use them and a whole bunch of metrics at the top like a scorecard. They're fun to write with but who knows how accurate their base data is? I'm pretty sure most of them are making it the fuck up, because they also tell their clients that it can take up to 18 months for their sites to start ranking, by which time the guru has at least two years worth of software subscription money out them and can safely leg it. I spend a lot of time telling prospective clients just how full of shit a lot of these guys are (as gently and as patiently as I can, because no-one likes to think they've been taken for a ride), but it's like the old saying goes - in a gold rush, the person making the money is the guy selling shovels. The gurus are responsible for a lot of the fouling of the water, so to speak. When I started doing SEO most of the gigs were for small businesses trying to rank themselves up a bit and turn a bit of foot traffic or web-store sales, now its all spectacularly bad content mills.

From my end, I'm writing the best content I can in the frameworks I have, and its keeping the lights on and I can at least be confident that I'm raising the bar a bit for what SEO can be. I try to steer my clients towards producing high value content and of my two most recent clients the one that did is still afloat, and the one that did not has gone under so I must be doing something right. I do my research, I write well, and I don't do fluff or filler pieces if I can help it. I'm almost entirely writing plant care stuff at the moment, which is a hobby of mine, so I write what I'd want to read myself and I'm happy with the idea that some poor bastard who finds my work will at least have something served that's accurate, informative and a relic of the old web where hobbyists were the ones doing the bulk of the writing on any given subject. It's paying off for my clients too, because my work is actually good and not just filler between the ads, and they're seeing more modest but more consistent traffic as a result.

There's an irony there, too, I should point out, that when I was freelancing for print media oh twenty fucking years ago it wasn't much different. I was writing around ads a lot of the time for magazines. At the time men's mags paid the best and on time, and all the 'reputable' publications either had staff writers or paid their freelancers very little, so it was a hell of a lot more pronounced, but even the bigger publications had a low regard for the actual content that attracted their readers in the first place.

Anyway. I'd frankly be happy if SEO pissed off entirely. I can write better without it, and SEO produced content has a bland uniformity to it that's there to pacify whatever the site owners think Google wants at the expense of the reader's pleasure. I'd love to go back to producing content that is entirely focused on being educational, informative, and entertaining, but so long as ad revenue is keeping the Internet afloat that won't happen, and Google will continue to suck. It's both the seller and the shill, and has nothing at all to motivate any improvement.
posted by Jilder at 7:48 PM on January 17 [34 favorites]


I switched my search-engine to Wikipedia. Can be a little annoying when I forget to pick a 'real' search engine from the drop down but its actually not terrible for a lot of stuff and I enjoy it being completely wrong rather than getting SEO boosted content anyway.
posted by phigmov at 7:54 PM on January 17 [7 favorites]


I will add one more thing before I shut my piehole - there is only so many ways you can answer any given search query. Google churns. The SEO race means there's always someone new at the top of the page, always some new keyboard jockey at the desk flipping out whatever word count the need to meet their bills.

The minute Google starts accurately producing search results that churn will stagnate. There's only so many ways you can repot a plant, and so it'd be the same set of top ranking pages every time.

I can see those top twenty pages staying largely the same for many topics, and this is not as good an outcome as you'd think because I can see that being an avenue of homogenisation that would not be great for the Internet either.

I don't have a solution for it. I think we can start at least with Google lifting its game on all of its ranking metrics (at its purest, sites are ranked by Expertise, Experience, Authoritativeness, and Trustworthiness) because frankly it's getting its lunch eaten - the first page results almost never hit any of those benchmarks. They aren't even ranking how they claim they are and it shows, and it makes it easier for shitty gurus and bad mills to make their fast money before they burn out.
posted by Jilder at 7:55 PM on January 17 [8 favorites]


Any recommendations on maps? I was thinking that I just couldn't navigate, but maybe the problem isn't entirely me. Maybe I'll get back to paper maps.
posted by Nancy Lebovitz at 8:28 PM on January 17


there are too many products to review and not enough resources given to sites to properly review them all.

And because there are so many cheap and crappy products, everyone is constantly replacing them and in search of more reviews.
posted by smelendez at 9:58 PM on January 17 [2 favorites]


More like smug lowercase pronouncements, amiright. Dude, good for you that you have a search engines worth of personal contacts but, frankly, I don't appreciate your attitude in your comments. Does the irony - and, dare I say, intense privilege - of vouching for being able to ditch the electronic web in favor of the human one while on the former escape you?
posted by DeepSeaHaggis at 10:43 PM on January 17 [1 favorite]


seconding that Maps question... tired of Google Maps considering street names a tertiary priority to letting me know about businesses I don't care a whit about

It's gotten fucking unusable on mobile just with all the screen-real-estate-hogging popups and swipe bars and junk. Just show me the fucking map, would you? No.
posted by ctmf at 11:03 PM on January 17 [3 favorites]


My Google results improved immensely when I blocked results from the spruce.com and Pinterest, and I don’t understand why Google doesn’t just do the same. Delist the garbage domains already.

Unfortunately I have found DDG mostly useless, and often have to redo searches in Google to get relevant results.
posted by Just the one swan, actually at 11:12 PM on January 17


How can you block a specific domain?? Please tell me, that would be awesome!
posted by wenestvedt at 5:23 AM on January 18


Adding '-site:pinterest.com' is supposed to block anything on the 'pinterest.com' domain. In theory putting a dash/minus in front of any term will suppress it.
posted by Karmakaze at 5:49 AM on January 18 [2 favorites]


Filed under: Things You Didn't Need a Study To Know.
posted by tommasz at 6:08 AM on January 18


Seconding "word of mouth" as the best way to discover new high-quality web content, and the idea that manually-curated link collections such as Metafilter are a good proxy for word of mouth.

No, I don't have a search engine's worth of personal contacts. But I have dozens of contacts (as well as strangers-from-trustworthy places, like you guys) who use search engines differently than I do, and therefore find different things, and have some intuition about whether their results are what I am looking for or not.

Yes, involving weird of mouth is slower for specific searches. Yes, the algorithmic search is still a step in the process. Neither of those is a bad thing.
posted by fantabulous timewaster at 6:11 AM on January 18 [3 favorites]


I'm happy with Google Maps. But for alternatives in the US, Bing Maps is probably the closest alternative. From what I've seen Apple Maps has gotten a lot better to but it's only available if you own their computers. DuckDuckGo has maps which are said to be based on Apple's.

For mobile (only), Organic Maps is pretty great and showing maps. It's not as good at searching for names of places nor doing driving directions. It's an offline viewer for OpenStreetMap, probably the best general purpose version of OSM.
posted by Nelson at 6:34 AM on January 18 [1 favorite]


I mean, I disagree with bombastic lowercase pronouncements, but I kind of get their point? By browsing through sites to find information, there's a kind of web of trust you're navigating. Machines cannot fact check, so finding information without using an algorithm can potentially make it more reliable. It also means it takes a lot longer and makes lots of things, in the current web, outright unfindable, and also contributes to popular sites becoming more popular.

It occurs to me that Google, by making so much findable without having to use hyperlinks and ordinary surfing, created an early version of a similar kind of problem as we're all now facing with (here come those scare quotes I've been using again!) "AI." Here me out--

"AI" takes advantage of the good will of countless internet users adding their work to the digital commons for the advantage of large corporations. As their output becomes more prevalent on the web, it both makes people unwilling to share their work with others for free, and means that it sabotages itself, because as "AI-"generated images proliferate, it gets fed back into the "AIs," and makes their output worse.

Google took advantage of the good will of countless people linking webpages on their sites and blogs, for their own advantage. As its search became more ubiquitous, people just searched for what they wanted to find, becoming less likely to follow and make hyperlinks. Fewer links sabotaged Google, because PageRank relies on them to figure out which pages are most useful, while SEO linkfarms further diminished the notability, to the algorithm, of those links. That fed back into PageRank, making the results worse.

Granted the analogy is not perfectly parallel, and Google is trying to make information easier to find generally while "AI" just doesn't care. But I think it's helpful to recognize that both are trying to use algorithmic processes in a way that aggregates value from individual contributions and resells it, and in the process makes those contributions less likely to be made, creating a sea of lukewarm crap for us all to wade through.
posted by JHarris at 7:00 AM on January 18 [2 favorites]




Filed under: Things You Didn't Need a Study To Know.
But might find a study useful to convince the willfully ignorant.
posted by Karmakaze at 7:51 AM on January 18 [2 favorites]


Or to create enough bad PR for the search engines that they actually start to clean up their indexes.
posted by humbug at 10:01 AM on January 18 [1 favorite]


I switched to Kagi. Worth every penny/you get what you pay for. Might not be one for the privacy single-issue crowd since there's a payment involved.

Dunno why so many like DDG...maybe because it's 'free' and as far as free goes it's maybe kinda I dunno 'better' than other free search engines. Pretty low bar. I tried DDG as my default for a while, and it's pretty much the same crap; unsurprising since it's mostly Bing, and Bing is crap. DDG didn't show as much pure SEO garbage as Google, but it sure seemed like it wanted to front-load search results with someone *selling* something related to my search terms before giving me any information *about* my search terms. But it also got to around to the pure SEO garbage pretty quickly.
posted by kjs3 at 11:57 AM on January 18 [1 favorite]


The thing that always bothered me about DDG is that it just piggybacks on other search engines, so it will always be limited to cleaning up their mess. I want someone who does their own original search. Does Kagi do that, or does it also rely on others to do the crawling for them?
posted by I-Write-Essays at 12:16 PM on January 18 [1 favorite]


Regarding reviews specifically, I just signed up for Consumer Reports because I was so disillusioned with the "buyers guides" on sites like The Cut and The Strategist. I truly can't tell if they are actually reviewing the items or just summarizing product descriptions. But honestly, CR is full of a LOT of cruft too. I am considering buying a milk steamer/frother and searching on the CR site yielded a whole lot of junky "best coffee gifts for 2023" type articles on their site. I did eventually find one true review article that reviewed a few brands but it was only like 4-5 of them.
posted by misskaz at 12:16 PM on January 18 [2 favorites]


Search in the conventional sense of keyword searching IS a bad method of finding things. Ideally everything would be accurately tagged and we could do a tag search. Obviously that has problems.

However..... It may very well be that AI provides a possible way forward. I just tried and while OpenAI's current version of ChatGPT isn't optimized for it, here's the suggested tags it had for the King Arthur flour company's website with a muffin recipe:

Site: https://www.kingarthurbaking.com/recipes/basic-muffins-with-berry-and-oatmeal-versions-recipe

Tags from GPT: Baking, Muffins, Breakfast Recipes, Berry Muffins, Oatmeal Muffins, Easy Recipes, Home Baking, Sweet Treats, King Arthur Flour, DIY Baking

Not the best list perhaps, but it's not awful and I bet you could devleop an AI optimized for tagging content.

Then, of course, it'd be another arms race as people tried to game the tagging system, but at least it'd be a step forward.
posted by sotonohito at 12:27 PM on January 18


Put another way, DDG isn't a google and bing competitor, it is a google and bing reseller. They are still subject to the whims of google and microsoft policy, can't exist without them, and can not display results that google/bing have purged from their databases.
posted by I-Write-Essays at 12:35 PM on January 18 [4 favorites]


Neeva was the last serious effort at an independent search engine that I tried using. It shut down 6 months ago
posted by Nelson at 12:45 PM on January 18 [1 favorite]


For that matter, I wonder if you could get people to use a browser extension and tag pages as they encounter them for a non-profit independent tag search engine.

The problem there, of course, would be filtering signal from noise. But it's worked fairly well for Wikipedia so maybe?
posted by sotonohito at 12:51 PM on January 18 [3 favorites]


Ideally everything would be accurately tagged and we could do a tag search.

"Tags" end up being the naive answer to all sorts of data handling issues, from search to data protection/governance to IAM. I've done dozens of these projects. You can make it work, sometimes, if the tagging system is predefined, limited in scope and simple to apply. As soon as you start trying to find an accurate, unique and consistent tag for more and more fine divisions, much less trying to apply changes to the tagging system post hoc, it very very quickly becomes a nightmare of unsustainable complexity. But for search, there's no end to how finely you need to slice the tags, and so complexity ballons out of control almost immediately. And that's without considering bad actors in the system, and for search, they will be legion.

And no, AI is not the magical answer to this.

Obviously that has problems.

Unfortunately, not obviously, since people keep thinking this is the solution.
posted by kjs3 at 2:05 PM on January 18 [5 favorites]


The problem of tagging you describe sounds a lot like the problem of staying "on topic" in discord channels on a server. That's just not how conversations work, and it can be hard to figure out which channel to start a conversation in when the server has two dozen specialized channels and a hundred threads for even more specific topics. /derail
posted by I-Write-Essays at 2:19 PM on January 18 [1 favorite]


I've been using qwant for the past few weeks. It's been OK as a daily driver at work digging up linux sysadmin stuff, although I'm not excited that it's 20% owned by the Axel Spring Group, so I'm continuing to look for good options.

I'd looked at Kagi and decided to leave it alone after seeing their involvement with Brave.

And I just dug up searx.org while searcing out the kagi link. It's a search aggregator, but I just did my usual ego surf test with it and found it did a pretty good job, so I think I'm going to give it a chance.

Finally, I really like search.marginalia.nu as a complement to regular search engines for my non technical search errands. For example, it gave me a lot of really interesting results on 'garum' that I would not have otherwise seen using google, et al.
posted by ursus_comiter at 2:43 PM on January 18 [6 favorites]


I'm a huge fan of refseek for factually correct information.

Marginalia is fun but not a great tool for finding time sensitive or practical information. Still good for bumbling around or doing pleasure driven reading.

As far as word-of-mouth networks go, that's basically what sharing links on social media is this days. Put in what you want to get out, so if you see something neat in your wanderings share it. Most people probably do have a few hundred people who share links with them regularly, it's just everyone's walled in at Facebook or whatever and working from a phone makes it harder to share links cross platform.

Anyway, in an effort to balance my ledger I'm also trying to do more of my own writing over on Medium and putting lots of old school inline links, which used to be standard across the Internet back in the Day. Remember that? When you never published anything without a link or two in the body text? Good times.
posted by Jilder at 2:57 PM on January 18 [7 favorites]


Honestly, I'm not even sure where the line between "legitimate review site" and "review content farm" is anymore.

Guys do you remember how reviews used to operate? The author would either get the stuff free, cheap, or the company would buy them, and they would physically try them all themselves. So you knew at the very least a human had given their reasonably honest thoughts about the thing. The reviews would cover the top brands for the type of thing, and usually list their general price at time of publication.

That is…not how works now.

And that means I can’t trust places to make big purchases, because I have no idea which reviews are real and which are cannibal piecemeal.
posted by corb at 7:56 PM on January 18


From way upthread:

Can't really blame them or the employees. If billions of dollars were thrown at you, or if you're a worker getting millions in stock options, what would you do?

This is why Planet Sell-Out (Earth) sucks so much! It is NOT OK for things to be this way!

These rich fucks don't know the concept of "Enough", so they just grow without bound, sacrificing their core product (and everything else, including their values) on the altar of MORE. It's not enough for them to become billionaires and provide something incredibly useful and revolutionary to the world that provided great value to billions of people... NOOOoOoOoOOo they had to have *ALLLLL* the money so they fucking turned right for ADVERTISING and never changed course.

The fact that people selling out their values, their life's work, their morals, their ethics, their sense of making the world a better (or decent-er) place, is the NORM is a travesty. I will NEVER not be pissed off about it.

Bill Hicks had it right about advertising and marketing. It ruins everything.
posted by cats are weird at 9:29 PM on January 18 [7 favorites]


I'm enjoying Kagi, I switched to it yesterday and so far it's worked well and hasn't annoyed me once. Its AI enhancements are good but so is the basic search engine. I particularly like the lenses; you can tell it to give results from recipe sites, or academic sites, or "small web". Nice way to turn a knob to customize things.

They've made a mistake with the free trial though. It comes with 100 free searches. I don't mind the login, I don't mind that I'll have to pay. But limiting searches is a real mistake. I felt instant anxiety that I was wasting my precious allotment, making me not want to use the product. I've used 87 searches in under 24 hours. If I continue liking it I don't mind paying the $10/mo for unlimited but they should really give new users a week of (nearly) unlimited for evaluation. Making your customers afraid to use your product during a trial is not good.
posted by Nelson at 1:58 PM on January 19 [3 favorites]


Originally I had the limited-searches subscription. In the middle of trying to find a way to get decent N95 masks to my mom in Hungary before her intercontinental flight, Kagi cut me off, with an upsell offer. I instantly cancelled my subscription, complete with a very testy message.

Later they came out with a decently-priced unlimited "couples" plan and I mean come on, acknowledging that two-person households exist was alone enough reason to re-sub. Been a happy customer ever since.
posted by tigrrrlily at 3:08 PM on January 19


This is a fantastic topic that I would like to chime in. It seems to be like Google has switched their focus to AI race and forgot all together about their search engines.

The level of spam in Google is the highest it has been since forever. There have been cases where Google Spreadsheets are ranking on top of Google for random keywords (try searching for JFK death penalty and see the results for yourselves).

Not to mention how old expired domains are being used to manipulate the search engine ranking positions.

It's embarrassing for a company such as Google to lose at their own game and this is the best moment for one of the closest alternatives to jump and get that part of their big pie. Google is indeed just a directory of links nowadays with no grasp of common sense.

Even by searching for a specific term or set of keywords sometimes it gives you completely different results
posted by andrewmc at 5:56 PM on January 20 [4 favorites]


A point that comes to mind here is that in any advertising-supported search engine, the organic results are ultimately competing with the paid results. An advertising-supported search engine that doesn't degrade the quality of its organic results right up to the point that would cause people to abandon it is leaving money on the table.

I had high hopes for Kagi but they seem to have lost interest in being a search engine rather than an AI-powered answers/discovery engine. What I really need in a search engine is the ability to run precise queries against as much of the web as possible. If I wanted AI-powered idiotproofing I would have stuck with Google, and in fact that's what I've found myself going back to, as Kagi somehow manages to have even worse support for search operators than Google, even the latter's badly degraded current state.
posted by Not A Thing at 6:09 PM on January 20 [4 favorites]


OK, Qwant started giving me an obstructive nag screen about ad blocking, so I'm throwing it in the trash.
posted by ursus_comiter at 8:43 AM on January 24 [1 favorite]


UGH I hate those screens so much! And apparently they've been normalized enough that everyone feels they can do them for stupid reasons. NPR earlier today threw up such a screen specifically to say they didn't have a paywall. Thanks for interrupting my reading to tell me that! RRRRAGRGHRAGHGHRHG
posted by JHarris at 11:20 AM on January 24


Qwant didn't even have the weasel worded, "click here if you hate us and refuse to expose your system to the tide of malware that comes from ad services" link to bypass the nag screen. Just a block and no search results.

So yeah, right into the trash.
posted by ursus_comiter at 11:47 AM on January 24 [1 favorite]


Kagi: Celebrating our first 20,000 members. That's paying members, and pencils out to something like $2M a year in revenue (hard to know since they have different subscription plans.)

Also a puzzling footnote
why did we go through all this trouble and allocate nearly a third of our investor-raised funds to produce and freely distribute 20,000 t-shirts?
That links to Kagi's June 2023 announcement they'd raised $670,000 from 42 small investors. A non-traditional and quite small investment. STill puzzled why'd they'd spend it on t-shirts for existing customers. I don't want a t-shirt, I want a thriving search engine.

My switch to Kagi is still going well though. A few times now I've tried Google instead for queries and never gotten better results. I am finding Kagi's "Quick Answer" AI thing quite useful, saves me time scrolling through search results for some types of things.
posted by Nelson at 5:16 PM on January 24 [1 favorite]


« Older Also, he's a hugger   |   SCOTUS takes aim at the government's regulatory... Newer »


This thread has been archived and is closed to new comments