DuckDuckGo has 50-100 employees, according to CrunchBase, while Google has over 100,000 employees. The exact numbers are irrelevant, but Google has around three orders of magnitude more employees than DDG.
a) DDG needs about 1/1000th the revenue of Google to survive at its current scale.
b) Software quality does not scale linearly with manpower. A 50-100 person team can do /almost/ as well as a 10,000 person team in the same domain. You spend a lot to go from 90% to 99% or 99% to 99.9%.
c) Unsurprisingly, I find DDG's search results good enough. I rarely go to Google, and only when DDG fails me.
d) If DDG can't target advertising as well as Google, they'll still have enough revenue to survive at their current level of quality. Their CPM is likely lower than Google's with less data and optimization, but I'd guess at worst 2x lower, not 1000x lower.
I'm glad they're doing well. On paper, it seems like a sound business model, and it's a heck of a lot more customer-friendly.
Curiously, I'd say 90% of the time I use Google is for math. If I'm typing a complex math expression (especially with units), Google tends to do much better than DDG.
When I switched to DDG, I was happier with it than with Google. To me, DDG feels like Google felt ten years ago, before it was infested with "AI." Today, Google search results look more like the invasive shopping mall ads in Minority Report, tailored to what it thinks I'll click on, rather than relevant results based on my query. I only rarely (~once a month) have to use Google search to find something that DDG just cannot, and usually this is some highly technical topic. (I did it yesterday because I couldn't find information about a strange, fatal regex error in Firefox - seriously, one regex that Firefox can't understand prevents script execution, even when wrapped in exception handlers.)
I think Google has hit some sort of inflection point where their further attempts to improve search have made it a less useful experience and are mainly in service of their dual corporate mandate to put al ads in front of all eyeballs and control all access to all information.
Even DuckDuckGo is becoming more like Google in it's uselessness. I think the problem stems from the idea that new is always better, so all search platforms become infested with garbage sites that are automatically generated to capitalize on the current search patterns.
For instance, I was feeling nostalgic and wanted to look at the Space Jam website last night, but I couldn't remember the URL, so I searched for "space jam". The first page was all garbage results about a potential Space Jam 2. Okay, that's generating a lot of buzz, I get it. So I searched for "space jam website". The results were hundreds of identical culture blogs talking about the website. Nowhere was the link to the actual website, which turned out to be spacejam.com by the way (I thought it was warnerbros.com/archive or something...)
Same thing with other search terms. I've been playing a lot of Animal Crossing lately. I remember a time when you could search for a topic about a video game and find a website answering the question, or a forum post with relevant information. Now all you get are blogs posted by people who clearly don't have English as their first language, and likely have not played the game. It's keyword blogs and fandom.com/wikia all the way down,
>I think the problem stems from the idea that new is always better...
This is definitely one of the curses of the tech/software industry and we keep cycling through certain phases, it seems, building nearly unmanagable complexity in the process.
> When I switched to DDG, I was happier with it than with Google.
I regularly do !g, but I think the thing that is reducing that fastest is the declining quality of Google results. If I search for ‘x, y and z’, I then see a load of results with small text underneath saying ‘include z?’. It drives me demented, why wouldn’t I want to include my search terms?
> It drives me demented, why wouldn’t I want to include my search terms?
Well, I think the answer to your question is Google is trying to be friendly to the layman. It's not hard to think of examples where the user inputs a something into the search box that a CS major knows will not result in the implied wants of the user. Google tries to do a lot of human to machine translation work, and I think you're seeing the result of it. That being said, you can use quotations to explicitly include terms.
This works for the masses, but in my experience when machines force a use case, it causes more headache than not cause I'm not like the masses.
A barely related side rant: my microwave decided it wants to error out when it thinks you're trying to microwave air. i have a bowl in there. stop telling me to open the door to put something in. Whatever engineer at GE thought they would be smarter than the user, you're one of the many reasons your company is going out of business.
> I regularly do !g, but I think the thing that is reducing that fastest is the declining quality of Google results
I don't know how I feel about it, but I feel like this has happened in multiple places now: The dominant option is losing ground to competitors, not because they got better, but because the dominant option got worse. I've similarly argued that the Linux Desktop is now at least as good as Windows (and possibly MacOS), not because our UIs suddenly got more friendly (although they have improved), but because Microsoft shipped the Windows 8 ("who needs the start menu?") and 10 ("we'll merge the old and new control panels any day^wyear^wdecade now") interfaces, so the bar is so much lower. I don't know that this is progress, but it's great if you want FOSS to win. Likewise, it sucks that Google is getting worse, but hey, at least now there's less reason not to use DDG...
Not to delegitemize your use of `!g`, but I'm sure I'm not the only one who never needs to use that, and if that's true, I think we need to hear more from the people who are perfectly happy with DDG by itself.
Same. I've been using DDG regularly since.. I don't know, 2013 maybe. (iOS 8 added DDG as a choice for a default search engine; that was at the end of 2014, and I had already been using DDG for quite awhile when that happened.)
It's been a long time since I used !g for anything. I did for awhile as an alternate search, but eventually it stopped being helpful -- if I didn't find any good results in DDG, then the Google results didn't help either. At this point it's probably been a couple years since I did a !g for anything.
If you regularly use !g, then why don't you just use Google? Serious question. It seems like if you are resorting to Google enough, there's almost no point for you to be using DDG.
I'd say I need to use it for less than 5% of queries. And often those 5% aren't solved by Google either because they're just hard to search for stuff or my original query wasn't good enough.
I do regularly use other Google search services, though. In particular, Google Books.
I find it depends on how much I know what I'm looking for. If I know exactly what I'm trying to get to, google will usually get me there faster. If I don't, DDG is usually a good place to start, as it's less festooned with SEO'd garbage sites. And, worst case, there's always !g.
This is basically exactly my experience. Google's index is much more current and contains more obscure content. When I know exactly what I'm looking for but forgot the exact name or something like that, Google's fancy AI really helps. For more general queries, DDG is sometimes better.
but recently I noticed what seem to be blind spots on many (maybe less popular?) topics. e.g. I can't find how to setup/configure borgbackup on nixos via ddg for any of ~3-5 query variations I tried.
For non-logged-in google with uBlock the relevant documentation is the first hit.
I used to try DDG a few years ago but eventually switched back to Google as default. Now I have it again as default since a year. Not sure when or how it happened but since a few months, I almost exclusively use DDG, switching to Google maybe once a week. Indeed that's then usually for niche topics. On the other hand I think my search and surfing behaviour also changed. I'm starting to rely more on Bookmarks again; also I'm actually reading websites again and browse through them instead of just CTRL+F-skimming through them...
Per https://help.duckduckgo.com/results/sources/, "...DuckDuckGo gets its results from over four hundred sources. These include hundreds of vertical sources delivering niche Instant Answers, DuckDuckBot (our crawler) and crowd-sourced sites (like Wikipedia, stored in our answer indexes). We also of course have more traditional links in the search results, which we also source from multiple partners, though most commonly from Bing (and none from Google)."
So yes, while they do get results from Bing, it's a bit disingenuous to say that DuckDuckGo is just Bing.
And the "four hundred sources" link links to 400 special case replies. They are probably useful, but fire rarely. It's basically Bing, and that page is a bunch of spin.
The page I linked above doesn't mention DuckDuckHack. Are you quoting that from a different source? It does mention DuckDuckBot which is their own web crawler.
Same! I knew about duckduckhack, but that it had been put in "maintenance mode" a couple years ago, but wasn't aware that DDG had started operating their own crawler. It's exciting to see--I'm hopeful they'll be able to leverage it to improve their result quality and capabilities while staying privacy-friendly.
Not necessarily - I would imagine many searches directly using Bing would return some form of "personalization" extending past the search terms itself. Duckduckgo claims not to consider this AFAIK. Although maybe the same supposed benefits of a search based purely on query could be achieved by a Bing or Google search without cookies.
If you exclude Instant Answers and widgets then pretty much, but the important thing is that they don't send your PII to Bing. Not saying that sarcastically. For me it doesn't matter what the underlying data source is.
How is it false? It literally says all traditional links are mostly from Bing: "We also of course have more traditional links in the search results, which we also source from multiple partners, though most commonly from Bing (and none from Google)."
All the other 400 sources and their own crawler are just used for Instant Answers and widgets.
It's funny, the first time I tried to switch to DDG, I couldn't make it stick. I eventually realized it was because I was frustrated by the lack of sports reporting carousels, which is admittedly a very silly reason.
I simply changed my behavior to use a sports website directly for those queries, and DDG has been totally sufficient for everything else.
Did that a lot, but honestly, these days I find google unusable. The add section at the top seems to be steadily creeping down, replacing more and more actual search results; googles attempt to remove URLs from the web annoys me to no end because I often go by URI more than by title and the fact that DDG has the sites favicon next to the URI oh boy that's a useful feature. In fact, I'd say favicons are what I go for 80% of the time; and using google I just feel like I'm completely blind. The additional clutter (top stories, twitter links, etc.) only make it even harder to visually parse what I want.
And then there's DDGs option to quickly switch to results in germany and back to international, depending on what I need.
Interesting, I don't think I've seen an ad on Google (or anywhere, really) in years. I've been using uBlock Origin on desktop and mobile (Kiwi Browser on Android, as well as AdGuard).
I only have uMatrix, which blocks most annoying iframe ads, but doesn't target things like googles ad results which are part of the page itself. Googles ads are actually quite nice in that they aren't flashy images, but they do appear like normal search results. It's the same for DDG, but there the ads have a big "AD" in a box behind the title; meanwhile in google it's way harder to visually parse what you want to click on.
Not to splice hair, but Google isn't just "surviving", so the amount needed for DDG to survive is probably even smaller than that. They don't need to be as profitable as Google Search (which is much more profitable than the company as a whole) to do well.
Google is not only massively profitable, they’re also running a ton of loss leading services designed to feed more data into the machine, like Gmail, as well as a ton of other experimental ventures that DDG isn’t doing.
Google is also headquartered in some of the most expensive real estate in America, while DDG is in Pennsylvania. I’d bet that Google is paying a lot more in land cost per employee than DDG is.
I live near DDG HQ, it’s not even a whole building. It can’t be more than a handful of offices. DDG’s real estate expenses are essentially $0 compared to whatever Google is spending.
Not anymore. Ruth Porat has been Ruthlessly cutting all the "give back to mankind" projects, YouTube became profitable recently after TEN years of losses, and Google for work is now profitable which means Google drive and Google docs makes money, finally. when I worked on search for Google we had 700 people in a building that was probably 25000 square feet. It felt almost like a chicken farm. Not many people born in America worked there.
I think the numbers you put refer as Google as company (maybe more Alphabet) and not Google as search engine. I'm curious to know how many people actually works on Google Search Engine.
"Curiously, I'd say 90% of the time I use Google is for math. If I'm typing a complex math expression (especially with units), Google tends to do much better than DDG."
This is interesting because it highlights the changing expectation of web search engines from when they were just tools to find relevant webpages for a query.
Search engines have morphed into all-purpose utility and information tools. The bar for what makes a good search engine has been greatly moved forward. Organic web search results are now just one part of the measure of quality and relevance.
I never use WA, but I'm typically reaching for fairly basic math calculations that don't involve units. Last example was calculating an estimate cost of DynamoDB based on the peak RPS I'd found, which is just basic math with a lot of significant digits.
Their own index is not used for any organic search results, only Instant Answers and other widget stuff. Money quote from your page:
> We also of course have more traditional links in the search results, which we also source from multiple partners, though most commonly from Bing (and none from Google).
I agree that quote is there (and have read the surrounding context), but it doesn't seem to me that it actually says that DuckDuckBot results aren't used in any non-Instant Answer search result, which I think is a distinction. They're very clearly synthesizing results from multiple sources, including many from Bing, but neither that page nor the DuckDuckBot-specific page seem to directly exclude the use of their internally-crawled results in the non-Instant Answers search results. Did I perhaps miss something?
You're right but at the same time they say that they source their "traditional links in the search results" from "multiple partners", which to me seems to imply that it's from partners only and not their own crawler.
Try searching for something random like "woxioai" on DDG and you will see that the search results are identical to Bing (albeit a slightly different order but I presume they do some internal sorting at least), while the search results are completely different on Google.
Does the DDG bot actually do a lot of crawling? I've never seen anyone talk about it except that it exists, while for Bingbot, Googlebot and random chinese bots in the webdev forums I go to people complain all the time about excessive crawling.
Because Bing does what Google does, which is uses your search history to show you targeted ads. DuckDuckGo only shows keyword based ads, which aren't as effective but do not require building a persona of personal information to target.
It's totally a problem and disingenuous of the article to talk about the "Business Model" of DDG and not mention this.
I now realize that I have almost completely stopped using Google in it's most common setting. ie. just search for the thing you are looking for.
I use google almost exclusively to improve search on other websites or for special syntax based searches that are far easier to implement than good general purpose search.
Site:reddit.com , site:stackoverflow.com are my most common search queries. Otherwise, it use it to open Maps, YT, HN as a landing page.
Research and news both are usually acquired from a curated list of sources I visit directly, with no involvement from google.
If it is something specific then I will use "quotes" to find that exact thing.
Outside that, google scholar and maps are amazing, but I don't think DDG is trying to compete there.
The idea of searching for something with only a vague sense of what it is, has almost completely stopped. For generic searches the 1st page results of medium pages and single pager MSM articles are both useless and shallow. It isn't Google's fault that popularity and quality are orthogonal to each other.
If I genuinely cared about my privacy, I would seriously consider trying DDG out. However, I have surrendered all my data to corporate overlords and don't think there is much point in struggling anymore.
I think your assessment of the shallowness of 1st page Google search results nowadays is spot-on. Anecdotally, I’ve noticed this situation get worse over the past ~5 years, to the point where I have to skip a few pages of search results and really dig deep to find high-quality content and writing.
Unfortunately, SEO sites like Medium that “democratize” writing seem to be a big contributor to this problem. The lowest common denominator there is not very high, yet shallow articles seem to get tons of exposure to search engines. Not to say there’s nothing good on Medium, I follow the people whose writing I like, and I like them quite a lot. It’s just lost in the noise.
I’m curious if you’ve found alternative sources (except HN, of course). Is DDG any higher quality?
I feel like the ideal for me would be something balanced between a search engine and content aggregator - I would like to search for what I want and get high quality results, but avoid the echo-chamber effect emerging from the use of upvotes as a ranking method as much as possible.
> b) Software quality does not scale linearly with manpower.
I believe there's a tipping point after which software quality doesn't scale at all; in fact, it only starts getting worse. People confuse quantity (roughly: number of features) for quality and they think they're improving quality by throwing more and more developers at it.
Software works differently to many other fields in that all devs need to keep lots of complicated technical context in their heads, which doesn't translate well to how managers experience scalability, as that field is just vastly different. It's also hard to explain to a non-programmer (or even an unexperienced programmer) that removing code is good and sometimes condensing two features into a simpler, more generic and versatile feature is positive both for code quality and usability; then again, from a marketting perspective, two specific features are better than one generic feature. And then you have ways in which more developers do get things done better, but only if managed correctly (aka. if the tasks are split along the right lines)
If you don't want to use Google even for math, I would suggest WolframAlpha[1]. It can be used for much more complicated equations/unit conversions etc.
> c) Unsurprisingly, I find DDG's search results good enough. I rarely go to Google, and only when DDG fails me.
Same, my problem is that whenever I search for something that isn't english DDG can't find even the most basic stuff. Aside from that the migration has been pretty painless for me though so I have my hopes up.
That has been one reason why I prefer DDG to google since day 1; I can tell it to give me stuff from my region, or just international. Considering I use the internet in english 98% of the time but occasionally need something specific to Germany, that's very useful indeed :D
I find myself throwing !g when I want to find a programming specific thing. Not sure this is because Google's results for niche tech sites is better or because I am more familiar with how to craft these as Google queries having used it almost exclusively since 1998.
Works fine with Russian. Shows "In partnership with Yandex" badge (Yandex is a major Russian search/email/etc company) in the corner on some searches. Whatever "partnership" means...
Search is not ten thousand it is not even a thousand people. People don't realize that Google is actually double click add engine plus a relatively small search team; not 50 but less than 1000.
Oddly, I don't find WA to be great. WA can do complex math, but it's slow. To give a good example of what I've come to expect from all three, type "2 feet + 2 meters" into:
* WA: You'll get a correct result, but it takes 4 seconds.
Curiously, I find Google's to be the most helpful. I see the obviously incorrect result and instantly fix it. DDG gives me nothing. WA is too slow to e.g. play around with numbers.
My last computation was trying to figure out which bolt to get for a hole (with McMaster's selections for both bolts and, at some level, holes). I was converting hole sizes / bolt sizes between metric and imperial, and between fraction and decimals, playing around until I found a combo which worked. Once I had Google giving the right answers, I could play around until it worked. With WA, I would have had to wait 4 seconds each time.
By the time I'm doing anything beyond that, it's generally in numpy. There isn't really a spot for WA in between.
> Curiously, I'd say 90% of the time I use Google is for math. If I'm typing a complex math expression (especially with units), Google tends to do much better than DDG.
I used to use google for this too, until I discovered GNU units [1]. GNU units is vastly superior to Google for complex formula and runs quietly in my command line.
Some examples:
Q: What's 72°F in °C?
A: $ units "tempF(72)" "tempC"
22.2
Q: How many teaspoons are in a tablespoon?
A: $ units "1 tbsp" tsp
3
Q: What is the total energy content of 1 kg or uranium, burned completely in a breeder reactor?
A: $ units "1 kg /(238 g/mole) * 200 MeV * 6.022e23 / mole" "MW*days"
938 Megawatt-days
Q: How many femtobarns are in a square teraparsec?
A: $ units "teraparsec^2" "femtobarn"
9.5e99 (very close to a google)
Q: How many hectares are there in a baker's dozen of mm^2?
DDG is good enough for me usually. I very rarely find myself using !g to go to google. Usually, when I do it's to see where my own sites rank on google. I do find google better at local search, which makes sense since they know more about me. Like if I want to look up my pharmacy's phone number. It's always the first result in google but I get mixed results from DDG. Overall, Duck is pretty good though.
Software is only part of a major search engine. Another part, no less costly is hardware. You need a lot and lot of computers to scrap and index the web. To be more specific, Google had 2.5 million computers back in 2016.
Luckily, DDG doesn't need nor hardware nor much software, as they are mostly just intermediary between you and Bing.
> c) Unsurprisingly, I find DDG's search results good enough. I rarely go to Google, and only when DDG fails me.
I felt the same way for a while and even switched the default engine to DDG... but lately, I've stopped finding things and I was increasingly searching with "!g" to get something relevant, so I just switched back.
> c) Unsurprisingly, I find DDG's search results good enough. I rarely go to Google, and only when DDG fails me.
and this is where the !bang shortcuts are useful, and I like to think that DDG analyzes which requests were made with those bangs to hopefully refine and tune their systems to provide better results over time.
do you find DDG slower to load than Google? I have a slow mobile data connection, and noticed that DDG was much slower to load than Google. not sure if it is a raw point of presence thing, or if my phone (android chrome) is cheating somehow to load Google faster.
That's what I had heard before, but I looked it up and it's not quite accurate. They do have their own crawler, and they pull from "400 different sources", Bing being one of the major ones.
https://help.duckduckgo.com/duckduckgo-help-pages/results/so...
> "a) DDG needs about 1/1000th the revenue of Google to survive at its current scale."
I don't think evil or benevolence needs a threshold. Size of the company is absolutely irrelevant. Marketshare is absolutely irrelevant.
DDG's redeeming quality is that they've focused their entire identity on privacy. Should they fail at privacy, then a lot of their users will bail. On the other hand it's not like some open source app or distributed protocol that users can inspect. It's still proprietary technology using a centralized model and you have to trust their word for it.
And in fact, for me their small size is actually a red flag. This is because big companies are also big targets for law enforcement. The EU has been historically trigger happy when applying fines to big companies, since it provides a valuable source of income (I imagine). Google is in fact under a lot more scrutiny than DuckDuckGo or other smaller companies.
And you can see this for example when interacting with Google's platforms, Adwords / AdSense, versus other smaller players in the "real time bidding" space. Google exposes a lot less about the user's profile, they leak a lot less, leaving a lot less opportunity for building your own user profiles.
They have also been transparent with their users about what they collect, even before the GDPR forced companies in the EU to be transparent. You can also opt out of logging app activity and profiling. Or more recently you could opt for a deletion of your history every 3 months.
Of course, you could say that you don't trust Google to do that. That's fine. But why do you trust DuckDuckGo then? They survive, after all, with the same business model. They say they don't give out your searches and info to Bing Ads, but you basically have to trust their word for it.
And given Google's history, how seriously they've handled security, plus they've always projected a certain maturity in handling user data (compared with Facebook), it seems to me that we are in conspiracy theories land. I was a DDG user, but seeing anti-vaxxers in the news recently led me to the conclusion that on the topic of privacy, my views are not evidence based. And I'm going to guess that a lot of other privacy conscious people did fall into the same trap as I did.
Going forward, for me there's only one way to ensure privacy and that is via technology (end to end encryption, open source, etc). And if I have to trust service and app vendors, I first need to identify my threats. Whom am I afraid of? What am I protecting against? Yes, I still fear Google, but now that fear is very contextual and task driven.
---
> "c) Unsurprisingly, I find DDG's search results good enough. I rarely go to Google, and only when DDG fails me."
DDG fails horribly for me on two kinds of searches:
1. local searches — it's completely useless for local searches — imagine that I'm in Romania and if I want Romanian results for "restaurants", DDG gives me results in Spanish, from freaking Spain
2. software development searches — I would have expected DDG to be better, with their integration with GitHub and StackOverflow but isn't. It's actually really really bad, many times completely ignoring my words, giving me results for words I did not type, so there's this annoying back and forth where I try refining my searches with extra words and syntax, to force it to search what I want; this rarely happens with Google and when it does, it's because there are no results available
---
That said, I would like for DuckDuckGo to improve and I am glad to see them doing well, because I'm a firm believer in competition. I would not like for Google to remain the sole option available.
On the other hand I really hope they divorce themselves from Bing. If Bing goes offline, or if Microsoft closes their API, it might kill DuckDuckGo. They do have integrations with other sources and their own partial index, but AFAIK Bing is their main source, which makes them effectively a Bing shell.
This is also why your operating costs are misleading. Because it's not an apples to apples comparison. Maintaining your own web index, apparently, is not that cheap. If it were, they wouldn't be using Bing.
Google isn't standing still. They're getting worse over time. I've noticed a significant drop in their result quality even in the past 1-2 years. Meanwhile DDG seems to be slowly improving. They are now my default and I only go to Google if I can't find something with DDG.
I like DuckDuckGo, but it's much worse than Google at localisation. It's annoying to have to search, then repeat the search with "UK" on the end to make it realise I'm not interested in buying e.g. building supplies or coffee or beer from US-based businesses.
I get that it has no concept of who I am when I search, but some kind of localisation based on IP address could maybe help?
Also, it feels a little disingenuous to say Google is not a search company, it's an advertising company. DDG's revenue is also built on ads. Is it fair to say they are also not an advertising company, not a search company?
I hate Google's localisation. My Accept-Language is set to English, I search for stuff in English, and Google throws piles of Dutch search results at me simply because I happen to be in Flanders. /ncr used to do the trick, but that's still showing me all sorts of Dutch stuff I'm not interested in.
Right, but it's kinda annoying having to turn it on and off, as it remembers and I keep forgetting what I used last.
If I'm searching for programming stuff or similar, I want it off. If I'm searching for shops that sells bricks for my new wall, I want local results.
I've also noticed that even with it on, the "local" indexing is distinctly significantly worse than Google. That is, Google is much, much better at finding relevant pages which may contain variations on my search term when using my native language.
I guess that's just a question of resources, which is fair enough.
> If I'm searching for programming stuff or similar, I want it off. If I'm searching for shops that sells bricks for my new wall, I want local results.
In all fairness, how exactly do you expect DDG to know what you want from this, without doing some sketchy privacy invasive scanning of your search? If you really want a company to scan your every search and then tailor the results specifically, then by all means use Google, that's what they're good at!
> how exactly do you expect DDG to know what you want from this
The obvious heuristic is “does this search term frequently appear on local result sites?” For example, “brick suppliers” is a search term associated with websites that have a business address, whereas “python string length” mostly isn’t. The search engine already has to scan your search to know what to send you from the index; it doesn’t require anything more than that to be able to guess at the intent of specific searches.
I posit that whether a search should be "local" or "global" is much more a factor of what is being searched for than who is searching. If I search for a barber I want local results, and I expect that that would be true for most people. If I search for a weird error message I saw while using my computer I want global results, and again I expect that is true for most people.
If true, then DDG could apply the same heuristics about location-sensitivity for everyone's searches and improve the results for everyone, without any tracking of detailed individual preferences.
Well in my case, if I write Norwegian words, it's highly likely I want Norwegian results.
Also as mentioned, if I search for "thai restaurant" (which is written exactly the same in Norwegian as in English), the chances are very high I want a local thai restaurant in my city and not one on the other side of the planet.
Even if I enable location in DDG the search results includes a bunch of thai restaurants in Asia and the US. The map insert however is pretty spot on however, so all is not lost.
This is super unreliable for me when it comes to "shopping" type results. I'll often have to append "uk" to get it to actually return UK results, even then it can be hit and miss.
Isn't completely reliable for me in New Zealand - I do get slightly more NZ-related results with it turned on (it's pretty much useless with it off), but I still get quite a few results (some even with .co.uk and .ca TLDs in the results!) from all over the world.
i think localization is tricky for DDG for non-technical reasons. their value proposition is privacy, and computer privacy is poorly understood by most users. to many, "this thing knows where i live" would seem like a contradiction in purpose.
I dislike Google as a company but their search results are orders of magnitude better than the competition. DDG or Bing are, to me, very bad (aren't they almost the same anyway?)
It's unclear why that is, and why search is so hard, but obviously it is.
Do you have an example or generally what kind of searches you are referring to? My experience with ddg has been "good enough" at least. The few times I don't find anything usable on ddg I ususally end up not finding anything on Google either
Not the poster, but here is an example. Elsewhere in this thread I posted a fact that I knew, but wanted to double check. It was about the LexisNexis news database containing 35k sources.
Here are the queries for "lexis news 35000" on Google and DDG:
You have trained Google to recognise Lexis as an alias for LexisNexis. DDG does not offer that, so one has to spell it out explicitly. You have to decide whether it's acceptable for you to change habits back to the common ground for Web searching.
The value provided by magic intent recognition (as opposed to traditional searching for keywords) differs from person to person. On the whole, it's likely for the better and that's why that feature was implemented at Google, but I have also read countless complaints about it on HN because situationally it pollutes results and can't be turned off.
You have trained Google to recognise Lexis as an alias for LexisNexis.
I disagree completely. Google knows Lexis is an alias for LexisNexis, but I did not train it from an individual level filter bubble perspective.
Look at the below results. They are sourced from Google but proxied so that Google doesn't get any information on the end user. The firsts result should be one that provides the relevant article I was initially looking for. [0]
Just one final point. Half the results on the DDG page for initial "Lexis" search I ran are for the car company "Lexus". So let's not pretend that DDG is respecting user search queries.
>Nexis is an online service for your strategic news and business research. Search for content from 35,000 publications. Conduct accurate searches of the 35,000 international sources in the Nexis online service. Make informed business decisions after getting the latest news and market reports.
this was the 8th result for me in DDG, with no filters or customisation or anything
this was the 8th result for me in DDG, with no filters or customisation or anything
So that link was 9th for me in DDG, but that exactly proves the point.
Google returned a relevant result in the first position. In addition, look at the result snippet from DDG:
The world is moving fast - Nexis news alerts help you keep up by delivering current news and information related to companies, trends, and events that matter to you. Set a Nexis alert related to a particular topic and get updates online or via email.
I wouldn't know that page had the information I was looking for unless I visited the page and reviewed the content. That's a lot of effort, when Google has the information in the top result and the snippet clearly indicates it is on the page.
(Edit: I also just looked at the content on that page, and it doesn't provide the information I was looking for. It provides the number of sources across all Nexis' searches. I was only looking for information on the news library.)
Also, half of the Lexis results on the DDG SERP are about Lexus the car.
Like I said, I like the mission of DDG, but in this and many other examples the organic SERP relevance is worse for multiple reasons.
I will try to document it for the next time this comes up.
I notice that Google is better at naturally knowing what I am looking for, and that DDG will show dodgy websites more often.
If I am searching for a PDF or a user manual, DDG will give me a lot of sites that are almost certainly malware laced (pdf-manuals.xyz) or something vs. what I get from Google.
Their maturity filter doesn't work as well, either.
The offshoot from this as well is: go to someone elses PC and do a query in google. You might find the results are just nowhere near what you want or expect.
Or tell someone to google something while on a call and click the 1st or 2nd link. You might be lucky and they're the same as yours, but more than likely than not they won't be.
DDG on the other hand, so long as I make the same query within a given timeframe, I'm relatively certain to get the same results back, regardless of which browser profile or PC I'm on. That reliability is (at least to me) a feature, and a damned useful one at that.
I've been using it at my daily driver since 2010. At first, I stuck with it because it wasn't Google. I noticed right away that it would turn up different results from what I came to expect at Google. Then I noticed that, although they were different, they were just as effective and getting me the information I needed. For example, it would prominently place websites I may not have heard of and that didn't show up high in the Google results, but when I just went with it anyway I discovered that relevant information that solved my problem was there.
I think a lot of people try out DDG, see some sites they don't recognize and then dismiss it as producing bad results because they are not Google results. We basically trained ourselves to use Google as the benchmark for what a search engine should produce - not in general terms of reliability of information but exact results. Then something else doesn't produce those same results and it's crap. Or, for a particular query, the top Google result is the best and the second or third DDG result is the best and that makes it crap. Search has always been a little bit of an art where you have to learn how to do it effectively and how to quickly parse your results.
I use Google maybe a handful of times a year and I would rarely find better results over there. It got to the point where I really didn't like their result page at all compared to DDG. It seemed spammy and it was difficult to parse. For a long time, they were fiddling around with things like displaying the full URL or making ads look more like legitimate results. On the other hand, DDG was doing things like zero click info and instant answers. I came to prefer their results much better. I showed up for the privacy and after a little while I stayed because it worked better.
I have tried Google a few times recently and they are doing some cool stuff with their results page. I particularly like the "featured snippets" and "people also ask" sections. Featured snippets is similar to DDG's zero click stuff, but it works with more complex queries like "best travel trailers for first time owners". But for years, DDG was doing cool things with the results page while Google was just a list of pages with ads, videos or news featured prominently at the top.
I've been using it pretty much exclusively for a lot longer than 2 years. I think part of the trick to switching is that google has got us trained as to how to search for stuff in a way that gets the results we're looking for.
It took maybe a month to feel like I was searching effectively after switching, but I never found ddg's results to be appreciably less relevant at that point.
On the upside, since ddg is less popular than google, people have put less effort into getting crap pages higher in the search results, though I'll concede it's gotten worse over the last couple of years.
For me, the bangs are the killer app, combined with searching from the browser bar: I don't have to wait for e.g. google maps to load to do a maps search if I search !m pizza in Poughkeepsie.
Yep. I switched to DDG last year and haven’t looked back. Between Safari/Firefox, DDG, Fastmail, and Apple Photos I think I’ve moved myself almost entirely off of Google.
Recently set up a Pi-hole to shield me from the rest. What else am I missing? Oh, YouTube! Drat. I’m a sucker for hobby videos on there. I’ve seen a couple of alternative front ends but nothing that has really stuck so far.
I'm happy with it. I'm not sure why I resisted turning on iCloud Photos. It makes cross device management a non-issue. Now I just need to find the time to cull my 20k something photos...
The major sticking point remaining is a shared library. I have my photos. My wife has hers. How do we manage photos of our children without duplicating?
While I am supportive of their business model, and search functionality, I find it off-putting the way they speak of privacy but then use maps provided by a GAFAM which are ultimately profit oriented and can _not_ be trusted to safeguard privacy regardless of their marketing. OSM would be a great choice here. I'll even dare to speculate that they were financially incentivized to make their choice in the face of a superior and actually privacy-oriented choice.
> I find it off-putting the way they speak of privacy but then use maps provided by a GAFAM which are ultimately profit oriented and can _not_ be trusted
DuckDuckGo itself is profit oriented, as is ProtonMail and a ton of other privacy first companies. Those aren’t mutually exclusive and I don’t think most people operate with that level of distrust.
I think the problem people have with Google is that their business model itself (targeted advertising) isn’t currently compatible with privacy.
This is anecdotal, but I had a positive experience with the lack of privacy recently. I bought a baby stroller, after doing the typical research and finding a used model, I ended up buying a new one from some babyshop. All the pictures had shown the thing with the stroller “hat/head/sunscreen thing” (sorry I don’t know the English word for kalche), anyway, it wasn’t included and that was why the brand new stroller had been priced the same as the used ones.
But it’s bought and it’s nice, so whatever I hit the manufacturer website and find the correct product and google it, and get a flash sale from one of our most prominent baby stores. It was 350 Danish kr including delivery. Without closing the tap I check a few baby shops and a price checker website and see it’s actually 600-700 Danish krs everywhere, including on from the company the flash sale on google is form. So I buy it.
Apparently I hit the right combination of search history, and store advertising/inventory at exactly the right time.
Being curious I called my local baby store to ask why they could flash sale me at half price, and after a bit back and forth they apparently do this thing where they’ll catch you early with a cheap item and then when you come back they to buy it the next day it’ll be priced higher, except by then you’ve made up your mind to buy it and will pay the extra and I was just lucky having already made up my mind when I got it because the other store has been cheeky.
Not really related to DDG, but it’s the first time selling my privacy has paid off.
In the UX world they are usually called urgency or scarcity triggers [1][2]. They also play in to the social default bias [3] people tend to have, where when alone you tend to do things that you know others have done.
While we are at it - what is the business model for freeware web browsers such as Vivaldi, Firefox, etc... I know they have some deals with MS, Google and friends but how such deals are done? Is it from some personal contacts or do you have to apply officially? What are the particular numbers - is it something you have to negotiate? What is the range in this case? etc...
I'm not. Being a re-seller of search and ads is not an impressive narrative for a company that sells the illusion of a search alternative. DDG adds value in the privacy realm, and I'd rather have it exist than not, but its continued existence depends on the continued will of a search api provider to sell and in the western world there's basically only 2 in that game: Google and Bing.
So first off personal bias, whize.co is the search engine I and my co-founder have been working on, it was nice to see basically our intended business model entirely validated by this article. Aside from the work we've done to validate it of course.
We entirely agree with their stated aims here but believe we can produce more quality search results than many current engines, hopefully it works out.
cheerfully says: "Showing 0 - 0 out of 0 for: KubeConfigLoader" although in your defense, it was blazingly fast to return that result. The shape of your XHR response seems to indicate you have an ES instance with only one shard, so maybe the site is still running in demo mode or something?
As a meta observation, in the bottom corner of every DDG result that I've ever received is a "Send Feedback" button, allowing me to advise them of bad search results in a very low friction way, but I had to come here to report my experience with your site
Yup still Demo mode the index is more than a few months old at this point. We're close to launching a beta with a much larger unrestricted index and a re-crawl strategy that keeps things fresh.
We've also changed our ranking methodology. The alpha / demo there was focused on novelty so if you search say "machine learning" you'd get repos that were recently active, starred and (hopefully) relevant to the topic, worth noting we also hard down-ranked repos that had been around for a while to let the smaller ones show through.
Through user conversation we moved away from focusing on novelty and moved onto "information staleness" the idea that elements in search results do not decay in relevance as quickly as they should based on real world events that happen around them. Happy to talk more about it if you want.
It's a bit strange but for me Google's biggest pull is live scores.
I follow cricket, tennis and football. And all the dedicated websites are heavy and slow (cricinfo, espn etc). Google scores are quick to load and update even before the rest.
Posted the same elsewhere. I've been able to change my own behavior to look for sports reporting elsewhere. Agreed Google has done an excellent job of it.
I am going to build my own search engine but with a monthly subscription fee. No ads, no tracking and I don't keep logs of your activity. So across between a VPN and a Search engine!
Sure, but the comment I replied to did not talk about a "niche" search engine. If you find a specific area of search that there is no good search engine for, or if existing products are really bad, of course you may succeed in that space, probably because you're at that point targeting the enterprise market -- but expecting normal consumers to pay for a search engine is naive at best.
I would wager that Google's database of new stories is bigger. LexisNexis survives partly through institutional inertia, but also I expect because they've tailored their search product to their specific niche in a way that is much more useful than Google's general-purpose algorithms.
Google's news articles goes back way further than the 70's. They've been scanning newspapers and integrating digitization archives. This is why you can you can look at word usage trends centuries back. But it's not very well integrated into search.
I and/or the company I work for would. Google is awful for finding the kind of high quality, niche information I need in my job. It just returns lowest common denominator rubbish.
As I wrote in my other reply, there was no mention of a "niche" search engine in the comment I replied to. Yes, there are future business opportunities and past examples of search succeeding in areas where Google is lacking: but when people think of a general "search engine" they think of a competitor to Google for regular Internet users, and there is no way that would be a viable business model.
> I and/or the company I work for would. Google is awful for finding the kind of high quality, niche information I need in my job. It just returns lowest common denominator rubbish.
May I know the niche that you are referring to? Also, how do other search engines like DuckDuckGo performs against Google?
I'm a consultant/researcher advising charitable foundations. It means I need to research all sorts of different topics depending on what area the foundation is interested in. e.g. one project might be on humanitarian aid, another on educational technology, another on infectious disease. So I don't need a search engine in a particular niche, I need a search engine that is good at finding high quality information in multiple niches. By high quality information I mean a mixture of academic articles, expert blog posts, podcasts with experts, policy reports, books, expert tweets, quality journalism etc.
I think there are other jobs that probably face this kind of problem, for example: policymakers, management consultants, journalists, nonfiction writers, some kinds of investors. They all need to rapidly learn things in topics outside of their expertise.
I haven't found other search engines much better, although I haven't yet done a systematic test. One thing that puts me off them is it seems from their marketing copy that most alternative search engines like Duck Duck Go are focusing entirely on being a privacy-preserving alternative. I want one that is focused on quality (and customisability), not just privacy.
I tried to install Searx so that I could at least get results from a variety of engines. But I got stuck in some dependency yakshaving so gave up.
A few years ago, I investigated if there is a business opportunity in quality search.
The biggest hurdle I found no way around is the content. Not all but a lot of high-quality information is paid. You get access to it as a user but not if you want to index it as a platform. And you need a lot of different providers to have good coverage. That might get easier if you have lots of users, but you do not get users without content. The platform chicken-egg problem.
Profitable niches like Bloomberg’s business information definitively exist but this would not be the high-quality generic search engine you described.
That's really interesting to know, thanks. I think for my purposes a lot of high-quality information isn't paid - it's academic, policy, and nonprofit sources mainly. I feel like you could make a lot of improvement for my purposes just by:
- Downweighting commercial sources and upweighting academic, government, and nonprofit sources.
- Using some measure of quality, maybe even something simple like length + reading age?
- Building big whitelists of quality sites and blacklists of low-quality sites, as picked by human curators and users
Beyond this, perhaps users could plug in sources that are particularly useful for them, including ones they have subscriptions to. I've thought it would be handy to have search results include ebooks, papers, and notes on my computer, for example.
But maybe the userbase of people who need high quality, non-subscription info is too small to get this started.
Maybe I was biased towards paid content because I did an enterprise search project on scientific articles before.
> users could plug in sources that are particularly useful for them
That could work if you have a local copy of the data which can be indexed. E-Book DRM might be a problem.
> But maybe the userbase of people who need high quality, non-subscription info is too small to get this started.
That was my impression. I almost always got puzzled looks when I described the idea. And most of those people were “knowledge workers”. But if you are a specialist, you already know where to look for the information you need in your area of expertise.
I remember years ago there was a desktop application for Windows called Copernic Agent. It was a meta search tool that combined results from multiple search engines. It had free and paid versions. It's not exactly the same as paying for a search engine, but it's close enough. It looks like it's abandoned now and the company moved on to desktop and enterprise file searching.
Also, there are people who pay for API calls to a search engine, for example Bing Search. Again, it's not the same as users paying a monthly subscription for unlimited access, but someone is paying in the end.
A paid search engine could work once it proves itself. This means they need to put a significant amount of work and server resources into it first, get a critical mass of users and then they can make it paid.
> 1) Build a better/equal search engine to that of Google
For what it is worth Google has been busy lowering that bar for about 10 years:
- lower quality results
- dumber ads
- more nauseating practices generally
I'd like to no however that last summer the most brain dead ads disappeared and the last few weeks I've seen decent search results (2009 quality) twice.
When Google is good old Google from 2009 - where they return the exact weird quoted thing I seached for - then I actually consider going back.
Not that I'm going to be the average user here, but I'm desperate to pay for a search engine that works like I want.
I've lost so many hours on google that the cost would make itself up. Just yesterday stuck in another gRut, looking for "what is X" to receive nothing but "how to implement X!".
Valuable hours lost every week. I'd happily pay for a search engine that just got it right.
I've lost so many hours on google that the cost would make itself up. Just yesterday stuck in another gRut, looking for "what is X" to receive nothing but "how to implement X!".
Doesn't a monthly fee mean you'll require users to sign in? In the purest sense that's tracking, you're just promising to immediately forget all of the correlations you're making between users and queries. I think you'd need a privacy policy/payment contract that adds a liability to the business to maintain privacy for users to trust that you aren't and wont track them.
DDG is surprisingly good, but still not good enough. I have my desktop browser using it, but still haven't on mobile.
When searching for technical queries Google tends to give much more useful answers - useful is the key word here. To give an example from yesterday, the query 'elixir zlib'.
In DDG I get links to C files in Linux kernels on a server called elixir. I get a useful answer 5 results down (which is fine, it's there). In Google, the first result I get is the erlang zlib documentation (which is what you actually use in elixir).
Things like error messages, queries involving the name of a programming language etc. are often just not as good in DDG.
> Things like error messages, queries involving the name of a programming language etc. are often just not as good in DDG.
I don't have any problem with those types of searches on DDG. Even obscure error messages turn up most of the time, and when they don't turn up in DDG they don't turn up in Google either.
Even when I do searches for a relatively less popular language like Haxe, I don't have trouble finding what I'm looking for. I only remember a handful of times over the 3+ years I've been using DDG that Google returned an ancient IRC log referencing my issue while DDG didn't return anything, but just the same I've had Google return irrelevant results while DDG gave me exactly what I was looking for.
Since my anecdotes offsets your anecdotes, the only conclusion is that anecdotes are useless to compare these things :)
I think it would be cool if one started an online advertising company in 2020 that employs about 50% graphic designers and 50% other and that specialises on providing online JPEG or similar images for websites, and where each add is completely human generated / human curated.
Also, the sizes should be completely fixed like you can choose between predetermined pixel ratios. Like maybe 162x100 or 1600x30 or whatever.
With people angsty with social media, I think it could be quite refreshing to have a trip back in time.
I find it very interesting that this article isn't just all "We're more holier [SIC] than google", but instead also argues that google could just as well adopt this busyness model. It doesn't defend google, but it doesn't paint it as this intrinsically evil IT-Monster that must be defeated, as many privacy advocates tend to do.
If anything, I feel DDG should get a more serious name, and one that doesn't sound funny. Seriously, whenever I hear DuckDuckGo, I feel it sounds like some blog or some meme sharing site. This only hurts its brand value and make it seem trivial.
One huge thing stopping me switching over is not being able to see Google reviews for restaraunts, doctors, etc. Yelp is, in my opinion, much less trustworthy and it would be nice to be able to toggle this the same way Google maps can be a preference.
Some commentors are saying Google is worse. As someone who uses Google, what's been worse about it? With regards to ads, I don't see them as I use uBlock Origin on desktop and mobile (as well as other ad blockers and no-script extensions).
What's worse is that these people hate Google. Many do because they failed the interviews there and openly state this as the reason for their criticism.
I am also confused about this wording but I assume it refers to the duckduckgo-app on android devices. It would be blatantly anti-competitive if google would refuse the ddg app in their app store, solely because they compete with them in the search-engine market -- there are legal mechanisms against this behavior.
Agreed. And there are laws against this sort of thing, and the regulators are slowly but surely catching up to the fact that they will have to start actually enforcing them.
It's not always a bad thing. The duck.com domain used to redirect to Google for a long time (they got it in a random acquisition) but now it's been changed to DDG instead. I'm pretty sure that this kind of analysis played a role there. (A more diversified ecosystem in search also helps with fighting SEO spam, which is by far the biggest actual threat to Google Search volume.)
Less than ideal case-scenario, users are spreading their data between Google and Duck Duck, Bing etc. Not empowering a single agent. In the worst case Duck sell our data to Google. I find that unlikely.
a) DDG needs about 1/1000th the revenue of Google to survive at its current scale.
b) Software quality does not scale linearly with manpower. A 50-100 person team can do /almost/ as well as a 10,000 person team in the same domain. You spend a lot to go from 90% to 99% or 99% to 99.9%.
c) Unsurprisingly, I find DDG's search results good enough. I rarely go to Google, and only when DDG fails me.
d) If DDG can't target advertising as well as Google, they'll still have enough revenue to survive at their current level of quality. Their CPM is likely lower than Google's with less data and optimization, but I'd guess at worst 2x lower, not 1000x lower.
I'm glad they're doing well. On paper, it seems like a sound business model, and it's a heck of a lot more customer-friendly.
Curiously, I'd say 90% of the time I use Google is for math. If I'm typing a complex math expression (especially with units), Google tends to do much better than DDG.