9 ms·

For years now I've ran a programming site (stackabuse.com) and have closely followed the state of Google SERPs when it comes to programming content. A few thoug

For years now I've ran a programming site (stackabuse.com) and have closely followed the state of Google SERPs when it comes to programming content. A few thoughts/ramblings: - The search results for programming content has been very volatile the last year or so. Google has released a lot of core algorithm updates in the last year, which has caused a lot of high-quality sites to either lose traffic or stagnate. - These low-quality code snippet sites have always been around, but their traffic has exploded this year after the algorithm changes. Just look at traffic estimates for one of the worst offenders - they get an estimated 18M views each month now, which has grown almost 10x in 12 months. Compare that to SO, which has stayed flat or even dropped in the same time-frame - The new algorithm updates seem to actually hurt a lot of high-quality sites as it seemingly favors code snippets, exact-match phrases, and lots of internal linking. Great sites with well-written long-form content, like RealPython.com, don't get as much attention as they deserve, IMO. We try to publish useful content, but consistently have our traffic slashed by Google's updates, which end up favoring copy-pasted code from SO, GitHub, and even our own articles. - The programming content "industry" is highly fragmented (outside of SO) and difficult to monetize, which is why so many sites are covered in ads. Because of this, it's a land grab for traffic and increasing RPMs with more ads, hence these low-quality snippet sites. Admittedly, we monetize with ads but are actively trying to move away from it with paid content. It's a difficult task as it's hard to convince programmers to pay for anything, so the barrier to entry is high unless you monetize with ads. - I'll admit that this is likely a difficult problem because of how programmer's use Google. My guess is that because we often search for obscure errors/problems/code, their algorithm favors exact-match phrases to better find the solution. They might then give higher priority to pages that seem like they're dedicated to whatever you searched for (i.e. the low-quality snippet sites) over a GitHub repo that contains that snippet _and_ a bunch of other unrelated code. Just my two cents. Interested to hear your thoughts :)

omega3 5y ago
> The programming content "industry" is highly fragmented (outside of SO) and difficult to monetize I think part of the problem is that the content producers are trying to cater to everyone: google algo (by artifically inflating word count, using specific keywords), beginner programmers, advanced programmers, potentially paid users, ephemeral users that arrive on the site via a referral or via googling. In the end you end up catering to no one. As a side note, RealPython.com is going to go down even more if they're going to keep their "register to view" policy I've started to see recently.
mritchie712 5y ago
Who is the offender?
btown 5y ago
There's an interesting dilemma here - if the algorithm were to favor "pages with lots of content alongside the exact-match phrase you're looking for" then it would incentivize content farms that combine lots of StackOverflow answers together. And if you favor the opposite, where there's less content alongside the exact-match phrase, you incentivize content farms that strip away content from StackOverflow before displaying it. Ideally, of course, you'd rely on site-level reputation - is Google just having trouble recognizing that these are using stolen content?
- ScottWRobinson 5y ago
  > is Google just having trouble recognizing that these are using stolen content? It's very possible. In general Google will penalize you for duplicate content, but that might not apply to code snippets since code is often re-used within projects or between projects. The code snippet sites also typically have 2 or more snippets on the same page. When combined, it might then look unique to Google since their algorithm probably can't understand code as well as it understands natural text. Just a guess
  - rodw 5y ago
    Also, while I imagine Google probably has (or at least easily could have) code-specific heuristics at play, it seems like it may be harder to reliably apply duplicate content penalties to source code listings, especially short code snippets. Between the constrained syntax, literal keywords, standard APIs and common coding conventions it seems like even independently-authored source code listings will be superficially similar. Certainly the basic/naive markov-chain-style logic that works really well for detecting even small examples of plagiarism in natural language content isn't going to as effective on `public static void main(String[] args)` or whatever. Obviously there are strategies that could distinguish superficial/boilerplate stuff from truly duplicated code, but between the volume of actual (and often legitimate) duplication on the internet and the (maybe?) low ROI for duplicate content penalties for code relative to other value metrics/signals, maybe this just isn't terribly important as a quality factor?
  - 6510 5y ago
Terry_Roll 5y ago
I've seen programming newsgroups, those things from the 90's, with what can best be described as MITM attacks having taken place when coders have been looking for solutions to problems and the solutions have not been correct. Most newsgroups were never secure so vulnerable to MITM from day 1 and what is being reported today is just the latest variation in that attack process. I've also seen Bing & Google citing StackOverFlow and the replies in SO awarding or agreeing on a solution comes straight from this "text book" "The Gentleman's Guide To Forum Spies" https://cryptome.org/2012/07/gent-forum-spies.htm https://cryptome.org/2012/07/gent-forum-spies.htm Perhaps it would be useful to dig into a posters history on a site and then decide who you trust instead of just trusting a random on the internet? How many people have download code from SO into VS and found it doesnt even do what its purported to do? I've seen plenty of that. Resource Burning the population, in this case programmers, is a perfectly valid technique for a variety of reasons, but the main one being, you are stuck in front of computer and that means you cant get into mischief away from work. Religions have been using that technique for hundreds of years and colloquially its know as "The devil makes work for idle hands to do" or something to that effect. Choose carefully what you want to believe and trust.
- webmaven 5y ago
  > I've seen programming newsgroups, those things from the 90's, with what can best be described as MITM attacks having taken place when coders have been looking for solutions to problems and the solutions have not been correct. Most newsgroups were never secure so vulnerable to MITM from day 1 and what is being reported today is just the latest variation in that attack process. Well, that's also the side effect of taking Cunningham's Law to heart, which says "the best way to get the right answer on the Internet is not to ask a question, it's to post the wrong answer."
jrumbut 5y ago
Recently I've just punted and begun searching SO and Github directly. One thing Google has gotten really good at lately is knowing that when I search for "John Smith" I mean John Smith the baseball player not the video game character or vice-versa.
- dhimes 5y ago
  I've always just searched 'John Smith baseball' and that works well in DDG too.
- nunez 5y ago
  you can also add 'site:github.com' or 'site:stackoverflow.com' to your search
  - jrumbut 5y ago
    It's not as good, especially for Github. GH also breaks down the results into types which is very helpful when you only want code or are looking for documentation or discussion.
charlesdaniels 5y ago
I wonder if we'll see a comeback of hand-curated directories of content? I feel like the "awesome list" trend is maybe the start of something there. I would be willing to pay an annual fee to have access to well-curated search results with all the clickbait, blogspam, etc. filtered out. Until then, I recommend uBlacklist[0], which allows you to hide sites by domain in the search results page for common search engines. 0 - https://github.com/iorate/uBlacklist https://github.com/iorate/uBlacklist
- freediver 5y ago
  You can access one without paying a dime. http://teclis.com http://teclis.com Problem is people usually want one general search engine, not a collection of niche ones.
  - wruza 5y ago
    It’s not a directory. Hand-crafted descriptions instead of random citations and/or marketing from the site itself is what makes it a directory. This one is a search engine. Maybe it’s a good one for its purpose, but who knows, without an ability to navigate you can’t tell. Problem is people usually want one general search engine, not a collection of niche ones. In my opinion, the reason they want a general search engine is that they think in their box (search -> general search). What they really want is a way to discover things and quick summaries about them: “$section: $what $what_it_does $see_also”. Search engines abuse this necessity and suggest deceitfully summarized ads instead.
- notJim 5y ago
  > hide sites by domain This gives me the idea to build a search engine that only contains content from domains that have been vouched for. Basically, you'd have an upvote/downvote system for the domains, perhaps with some walls to make sure only trusted users can upvote/downvote. It seems like in practice, many people do this anyway. This could be the best of both worlds between directories and search engines.
  - ancientworldnow 5y ago
Nextgrid 5y ago
> It's a difficult task as it's hard to convince programmers to pay for anything I wonder if there's an opportunity for a paid bundle of programming-related sites. I indeed will not pay for a single site (nor for a news site for that matter), but a $10-20/month subscription that covers quality programming websites could be interesting.
andai 5y ago
>it's hard to convince programmers to pay for anything Offtopic but I'm curious why this is the case? Is the Free Software movement responsible for this mindset?
- TheDudeMan 5y ago
  It's hard to convince anyone to pay for anything.
- john-tells-all 5y ago
  I train programmers, and strongly recommend they buy books or do other types of money/time investment to make themselves better (and more highly-paid programmers). They won't do it. I've had multiple programmers literally shocked and avoid my outstretched book. Once I got a question, said "I just read the answer to this in this book right here"... and the programmer refused to read the book to answer his question. I don't get it. This, coupled with companies lack of investment in their expensive engineers, is mystifying. None of the above has anything to do with the FSF.
  - tobyhinloopen 5y ago
    Maybe books are too low density. Like the “quantity of information” per amount of words is lower than a blog post for example. I dunno, I’m not really a reading type but I do own programming-related books. It’s the only type of book I own. I learned a lot from books like The Clean Coder, The Pragmatic Programmer and some 90’s book about Design Patterns with c++ examples and I don’t even write c++
    - Retric 5y ago
      Good programming books are some of the highest information density of any writing you can find. College professors tend to prefer the lower density versions which might be biasing people.
      - OJFord 5y ago
new_guy 5y ago
> I'll admit that this is likely a difficult problem because of how programmer's use Google It's beyond simple for Google to fix. Just drop those sites from the search index. But Google won't do that because it's in their interests to send you to those shit holes because they're littered with Google ads.
donmcronald 5y ago
I guess that explains why I've been seeing so many results that are nothing more than scraped content lately :-(
userbinator 5y ago
as it seemingly favors code snippets, exact-match phrases If only... it seems like the search results have gotten far worse for exact-matching. I regularly search for part numbers and markings, somewhere where exact matches are pretty much the only thing I want, and I can clearly see the decline over the years as it starts including more and more results that don't even have the string I'm looking for.
- phil294 5y ago
  ...which even applies if you enclose the specific term with quotes. These used to help against unrelated results, but not so much anymore. I don't know why. Same thing with DDG and Bing.
- rodw 5y ago
  Funny, it was the "lots of internal linking" bit that felt wrong to me. Not that these low-quality sites don't do that, but I'm surprised to hear that the new algorithm rewards internal links. I'm certainly not a full-time SEO guy, but I happen to have or work with a few sites - some fairly established - that make extensive use of internal links for design/editorial reasons. As far as I can tell they are helpful for (a) user navigation and (b) getting search engines to discover/index the pages but I don't think I've seen any notable advantage in or even impact on ranking based on those internal links (whether in the body copy or in head/foot/side "navigation" areas). Searching just now I do see some other sources making a similar claim, so maybe I'm just out of the loop. But in my cursory scan I haven't found much detail or real evidence beyond "google says they matter" either. I mean, that's not why those internal links were created in the first place, but it sure would be nice to get a ranking boost out of them too. I wonder what I'm doing wrong :)
999900000999 5y ago
I pay for Interview Cake. If you could create a site with both interview prep and quality code tips I'd pay for it.
galaxyLogic 5y ago
I don't know what is going on with Google but I thought their main idea is to give points to sites which other sites link to. Are these code-sites linking to each other perhaps?
- maks25 5y ago
  They’re called PBNs or “private blog networks”. A bunch of low quality sites all backlinking to each other. I don’t like the SEO industry very much.
Siira 5y ago
Please consider releasing any paywalled content after some deadline, e.g., after a year. It makes content practically inaccessible to a lot of people, including non-adults and people in developing countries.

omega3 5y ago

> The programming content "industry" is highly fragmented (outside of SO) and difficult to monetize I think part of the problem is that the content producers are trying to cater to everyone: google algo (by artifically inflating word count, using specific keywords), beginner programmers, advanced programmers, potentially paid users, ephemeral users that arrive on the site via a referral or via googling. In the end you end up catering to no one. As a side note, RealPython.com is going to go down even more if they're going to keep their "register to view" policy I've started to see recently.

mritchie712 5y ago

Who is the offender?

btown 5y ago

There's an interesting dilemma here - if the algorithm were to favor "pages with lots of content alongside the exact-match phrase you're looking for" then it would incentivize content farms that combine lots of StackOverflow answers together. And if you favor the opposite, where there's less content alongside the exact-match phrase, you incentivize content farms that strip away content from StackOverflow before displaying it. Ideally, of course, you'd rely on site-level reputation - is Google just having trouble recognizing that these are using stolen content?

ScottWRobinson 5y ago
> is Google just having trouble recognizing that these are using stolen content? It's very possible. In general Google will penalize you for duplicate content, but that might not apply to code snippets since code is often re-used within projects or between projects. The code snippet sites also typically have 2 or more snippets on the same page. When combined, it might then look unique to Google since their algorithm probably can't understand code as well as it understands natural text. Just a guess
- rodw 5y ago
  Also, while I imagine Google probably has (or at least easily could have) code-specific heuristics at play, it seems like it may be harder to reliably apply duplicate content penalties to source code listings, especially short code snippets. Between the constrained syntax, literal keywords, standard APIs and common coding conventions it seems like even independently-authored source code listings will be superficially similar. Certainly the basic/naive markov-chain-style logic that works really well for detecting even small examples of plagiarism in natural language content isn't going to as effective on `public static void main(String[] args)` or whatever. Obviously there are strategies that could distinguish superficial/boilerplate stuff from truly duplicated code, but between the volume of actual (and often legitimate) duplication on the internet and the (maybe?) low ROI for duplicate content penalties for code relative to other value metrics/signals, maybe this just isn't terribly important as a quality factor?
- 6510 5y ago

Terry_Roll 5y ago

I've seen programming newsgroups, those things from the 90's, with what can best be described as MITM attacks having taken place when coders have been looking for solutions to problems and the solutions have not been correct. Most newsgroups were never secure so vulnerable to MITM from day 1 and what is being reported today is just the latest variation in that attack process. I've also seen Bing & Google citing StackOverFlow and the replies in SO awarding or agreeing on a solution comes straight from this "text book" "The Gentleman's Guide To Forum Spies" https://cryptome.org/2012/07/gent-forum-spies.htm https://cryptome.org/2012/07/gent-forum-spies.htm Perhaps it would be useful to dig into a posters history on a site and then decide who you trust instead of just trusting a random on the internet? How many people have download code from SO into VS and found it doesnt even do what its purported to do? I've seen plenty of that. Resource Burning the population, in this case programmers, is a perfectly valid technique for a variety of reasons, but the main one being, you are stuck in front of computer and that means you cant get into mischief away from work. Religions have been using that technique for hundreds of years and colloquially its know as "The devil makes work for idle hands to do" or something to that effect. Choose carefully what you want to believe and trust.

webmaven 5y ago
> I've seen programming newsgroups, those things from the 90's, with what can best be described as MITM attacks having taken place when coders have been looking for solutions to problems and the solutions have not been correct. Most newsgroups were never secure so vulnerable to MITM from day 1 and what is being reported today is just the latest variation in that attack process. Well, that's also the side effect of taking Cunningham's Law to heart, which says "the best way to get the right answer on the Internet is not to ask a question, it's to post the wrong answer."

jrumbut 5y ago

Recently I've just punted and begun searching SO and Github directly. One thing Google has gotten really good at lately is knowing that when I search for "John Smith" I mean John Smith the baseball player not the video game character or vice-versa.

dhimes 5y ago
I've always just searched 'John Smith baseball' and that works well in DDG too.
nunez 5y ago
you can also add 'site:github.com' or 'site:stackoverflow.com' to your search
- jrumbut 5y ago
  It's not as good, especially for Github. GH also breaks down the results into types which is very helpful when you only want code or are looking for documentation or discussion.

charlesdaniels 5y ago

I wonder if we'll see a comeback of hand-curated directories of content? I feel like the "awesome list" trend is maybe the start of something there. I would be willing to pay an annual fee to have access to well-curated search results with all the clickbait, blogspam, etc. filtered out. Until then, I recommend uBlacklist[0], which allows you to hide sites by domain in the search results page for common search engines. 0 - https://github.com/iorate/uBlacklist https://github.com/iorate/uBlacklist

freediver 5y ago
You can access one without paying a dime. http://teclis.com http://teclis.com Problem is people usually want one general search engine, not a collection of niche ones.
- wruza 5y ago
  It’s not a directory. Hand-crafted descriptions instead of random citations and/or marketing from the site itself is what makes it a directory. This one is a search engine. Maybe it’s a good one for its purpose, but who knows, without an ability to navigate you can’t tell. Problem is people usually want one general search engine, not a collection of niche ones. In my opinion, the reason they want a general search engine is that they think in their box (search -> general search). What they really want is a way to discover things and quick summaries about them: “$section: $what $what_it_does $see_also”. Search engines abuse this necessity and suggest deceitfully summarized ads instead.
notJim 5y ago
> hide sites by domain This gives me the idea to build a search engine that only contains content from domains that have been vouched for. Basically, you'd have an upvote/downvote system for the domains, perhaps with some walls to make sure only trusted users can upvote/downvote. It seems like in practice, many people do this anyway. This could be the best of both worlds between directories and search engines.
- ancientworldnow 5y ago

Nextgrid 5y ago

> It's a difficult task as it's hard to convince programmers to pay for anything I wonder if there's an opportunity for a paid bundle of programming-related sites. I indeed will not pay for a single site (nor for a news site for that matter), but a $10-20/month subscription that covers quality programming websites could be interesting.

andai 5y ago

>it's hard to convince programmers to pay for anything Offtopic but I'm curious why this is the case? Is the Free Software movement responsible for this mindset?

TheDudeMan 5y ago
It's hard to convince anyone to pay for anything.
john-tells-all 5y ago
I train programmers, and strongly recommend they buy books or do other types of money/time investment to make themselves better (and more highly-paid programmers). They won't do it. I've had multiple programmers literally shocked and avoid my outstretched book. Once I got a question, said "I just read the answer to this in this book right here"... and the programmer refused to read the book to answer his question. I don't get it. This, coupled with companies lack of investment in their expensive engineers, is mystifying. None of the above has anything to do with the FSF.
- tobyhinloopen 5y ago
  Maybe books are too low density. Like the “quantity of information” per amount of words is lower than a blog post for example. I dunno, I’m not really a reading type but I do own programming-related books. It’s the only type of book I own. I learned a lot from books like The Clean Coder, The Pragmatic Programmer and some 90’s book about Design Patterns with c++ examples and I don’t even write c++
  - Retric 5y ago
    Good programming books are some of the highest information density of any writing you can find. College professors tend to prefer the lower density versions which might be biasing people.
    - OJFord 5y ago

new_guy 5y ago

> I'll admit that this is likely a difficult problem because of how programmer's use Google It's beyond simple for Google to fix. Just drop those sites from the search index. But Google won't do that because it's in their interests to send you to those shit holes because they're littered with Google ads.

donmcronald 5y ago

I guess that explains why I've been seeing so many results that are nothing more than scraped content lately :-(

userbinator 5y ago

as it seemingly favors code snippets, exact-match phrases If only... it seems like the search results have gotten far worse for exact-matching. I regularly search for part numbers and markings, somewhere where exact matches are pretty much the only thing I want, and I can clearly see the decline over the years as it starts including more and more results that don't even have the string I'm looking for.

phil294 5y ago
...which even applies if you enclose the specific term with quotes. These used to help against unrelated results, but not so much anymore. I don't know why. Same thing with DDG and Bing.
rodw 5y ago
Funny, it was the "lots of internal linking" bit that felt wrong to me. Not that these low-quality sites don't do that, but I'm surprised to hear that the new algorithm rewards internal links. I'm certainly not a full-time SEO guy, but I happen to have or work with a few sites - some fairly established - that make extensive use of internal links for design/editorial reasons. As far as I can tell they are helpful for (a) user navigation and (b) getting search engines to discover/index the pages but I don't think I've seen any notable advantage in or even impact on ranking based on those internal links (whether in the body copy or in head/foot/side "navigation" areas). Searching just now I do see some other sources making a similar claim, so maybe I'm just out of the loop. But in my cursory scan I haven't found much detail or real evidence beyond "google says they matter" either. I mean, that's not why those internal links were created in the first place, but it sure would be nice to get a ranking boost out of them too. I wonder what I'm doing wrong :)

999900000999 5y ago

I pay for Interview Cake. If you could create a site with both interview prep and quality code tips I'd pay for it.

galaxyLogic 5y ago

I don't know what is going on with Google but I thought their main idea is to give points to sites which other sites link to. Are these code-sites linking to each other perhaps?

maks25 5y ago
They’re called PBNs or “private blog networks”. A bunch of low quality sites all backlinking to each other. I don’t like the SEO industry very much.

Siira 5y ago

Please consider releasing any paywalled content after some deadline, e.g., after a year. It makes content practically inaccessible to a lot of people, including non-adults and people in developing countries.