It's hard to find results that are not mass media or commerce, aside from sources like reddit, quora, and other forums. Perhaps non-commercial sites, despite producing good content, don't make it to top results. There must be others like me trying to make sense of the world and writing about it while attempting to use logic and the scientific method. Many blogs seem mostly attention seeking without contributing substance, which unfortunately seems to sell. With this hypothesis I started to look for alternative search engines, that I might find better results.
I found many articles, but they were not focused on independent search engines, with their own crawlers and indexing methods. If they were about alternatives, they were misinformed and non-comprehensive. Since I couldn't find the article I wanted, I started writing my own. I have combed through what's available and give a brief summary of what I found. I was also curious about how much alternative search engines are being used.
Update 20201027: Apparently, I am not alone in these thoughts. Borislav Apagiev writes about the decline of quality in search results. Note he is an entrepreneur: “Founded two search companies, always working on the next one”.
The most popular alternative to Google, is Bing. In this review, I will avoid counting those that re-brand someone else's search engine results. One example is Yahoo search, which uses the Bing search engine since 2009. I made exception with DuckDuckGo and SearcX since I thought they were notable, and as meta search engines they combine search results from as many engines as possible.
In my first pass, I ruled out search sites where other articles, including Wikipedia, state that the search site uses someone else's search engine. It's possible that any of those articles could be mistaken or even outdated. I verified infrequently.
In the second pass, I tested search sites that claimed to have hybrid results. I have yet to see an abundance of unique results, if any, for any engine that supposedly “enriches” results from bing and/or google. I wasn't trying to be thorough: I tried at least 2 different searches looking for unique results. Some results were exact copies, some offered the same results but the order was randomized. Or mixed and randomized from both Google and Bing
as in the case of DuckDuckGo's.
Update 2020: Sometime since I ran tests in 2019, DuckDuckGo stopped using Google search results, and aside from Bing results, had 5 unique results in the first 30.
I am still going through the third pass, which is to test each remaining engine for individuality of search results. Some are crossed out below. They are crossed out, if I did not find sufficient unique search results. Sufficient unique results means that the first twenty results were not listed in the first twenty results of bing and the first sixty results of google.
During the third pass, I found a directory of international search engines, which lead me to expand my results toward the bottom of the page. I like Seznam and use it regularly. As opposed to Baidu, English search returns mostly English results. When I use Baidu, I use a browser extension to translate the results into English.
Lesser known search engines providing unique results.
I'm only promising unique results. I'll leave it to you to decide if the results are good.:
Search engines that are mostly rebranded Google or Bing results, or are not working:
DuckDuckGo (bing, google)
Qwant (bing, google) just one unique at result #16
SpiderLine (bing) unique 1st-try #5 #14, unique 2nd-try #3 #8
Gibiru (google) unique 1st-try #12 #14 #18, unique 2nd-try #10 #11 #12 #13 #14, hides visited links color
Yacy (not working)
MeekD (bing, google)
Update: I found www.searchenginemap.com that covers where a lot of search engines get their results from.
The following are still to be tested (with searchenginemap.com indicator in parenthesis):
Using Alexa or Similarweb to determine the popularity of search engines makes for a very rough estimate. Another rough estimate, would be to visit the Alternativeto.net website. Another would be to look at the most popular search engine plugins at https://mycroftproject.com/dlstats.html. Lastly, there is the google trends site.
The charts below are from Google Trends, so please take this conflict-of-interest into account. If you browse to Google Trends, make sure and remember to select “Worldwide”, if that is what you seek, because it defaults to the country you are searching from. I use separate charts so that the relative scale of lesser known search engines can be viewed. If Google were placed in the chart with the lesser known search engines, they would all show as a flat line at the bottom of the graph, as is the case with Bing, the search engine that had the second highest peak in the timeline.
The only search engines that may get more traffic than google, in a specific region, are the localized search engines listed below the charts.
Localized search engines, with independent or semi-independent engines. What is amazing is that on google trends, the popularity of every single local engine is declining. They are ranked here by overall google trends traffic since 2004:
http://naver.com South Korea
http://www.daum.net South Korea
Same Localized search engines as above, by current google trends ranking:
https://www.seznam.cz Czechoslovakia 1
https://www.rambler.ru Russia 2
http://naver.com South Korea 5→3
https://www.search.ch Switzerland 4
http://www.daum.net South Korea 6→5
http://www.najdi.si Slovenian 3→6
http://so.com China 8→7
http://ant.com Bangladesh (low traffic; on par with Qwant)
http://leit.is Iceland (low traffic; on par with Qwant)
http://sogou.com China (low traffic; on par with Qwant)
Localized search engines with barely any traffic, like, ever:
Human edited directory:
Tor / Dark Web (tor web browser required to view results):
https://ahmia.fi aka http://msydqstlz2kzerdg.onion
not Evil http://hss3uro2hsxfogfq.onion
Also duckduckgo, searx, metager, yippy
Unlike duckduckgo, some search engines include *only* onion results.
In 60 minutes of searching and reading on the TOR network, results were really slow in loading. Each search engine had their own set of unique results. Some had rather distasteful results.
If you visit a search engine's website, click the down arrow in the web browser's search box. You may find the option to add the search engine's plugin to your web browser (tested on Pale Moon). Most search engine's homepages offer this feature, but a few don't. Update: For those that don't, on Pale Moon web browser, you can use the extension Add to Search Bar, by right clicking over the search field and selecting this extension's option.
You can also create your own search engine plugins here:
1) You can have a search bar set up with your search engine plugins using this extension on Pale Moon web browser:
Using Search Buttons Bar, you can search the same search query on a different engine with one click. It can look like this:
2) Context Search X https://addons.palemoon.org/addon/context-search-x
Or alternatively, I'm using Quick Context Search https://legacycollector.org/firefox-addons/502176/index.html
On Firefox: https://addons.mozilla.org/en-US/firefox/addon/contextual-search
If you highlight a word on the page, these extensions give you a righ-click context menu option to search with your list of search engine plugins.
If you have one of Greasemonkey, Tampermonkey, or Violentmonkey on your web browser, you can use the script Alternative Search Engines to quickly perform a search with the same query on any of the search engines: google, bing, yandex, and duckduckgo. It works by showing links to the other engines on any search results page, as you can see in the image below. I find this one useful on my mobile device, using a web browser that allows extensions, such as Kiwi, with the drawback that I have to request the desktop site in order to see the options.
Other user scripts include:
Since SearX is open source, you can use the software and have your own search engine server:
I decided to do something and create my own alternative search engine, of sorts. Search results from search engines, filtered based on the resource domains a website uses to construct a web page. The project is here: https://github.com/mekineer/single_domain_search
And Google said “let there be hyphens”, and there were hyphens. Dokuwiki defaults to using underscores as word separators. Wikipedia uses underscores also, but if you aren't Wikipedia, I'm going to have a hard time finding your site unless you go through the arduous process of mass-converting your site to hyphens. Maybe there's a good reason? Matt Cutts says it's because programmers use underscores in variable names, which are treated as one word.
Google does have a conflict of interest: money. So without any other bias, it will provide results that are backed by money. Google makes money on advertising. If a site isn't paying money for advertising, or involved in their advertising program, then it will rank lower in Google results.
Also, sites will rank higher if they can hire someone in SEO, to meet Google's artificial requirements for being a “good quality” site. While many of the requirements aren't necessarily bad, they have to be studied in order to be followed. Sites that don't follow the requirements will rank lower, even if they have content that may be of high quality.
So your top Google search results will come from sites that are trying to rank well on Google. While this includes producing good content, it is not the only factor.
The other conflict of interest is in politics, but I can't speak with any certainty other than repeating different news sources. The only certainty is the existence of bias, because humanity. Some are better at doublethink than others. I hope we can still all think the best of each other.