51 comments

[ 3.1 ms ] story [ 114 ms ] thread
Now that subreddits have gone private , its even more difficult to get good result. Adding "reddit" to the end of the query is of no help.
Is there a tool for moderatorsof private subreddits to scrape the sub's content and upload it elsewhere?
I've long believed that SEO is a scam... Also, wasn't google supposed to start punishing over-SEOed websites because of people google bombing "miserable failure"?
SEO works; you can verify it easily by searching for how to do any action on linux. Unless you include "stack", all the results will be SEO spam sites that copied their content from stackoverflow.

Google does punish SEO, but it has failed in adressing the recent trend of blogspam. I think ultimately unless they're whitelisting sites (like SO) there is no way to algorithmically differentiate between blogspam and genuinely useful tech blogs, because the difference lies in natural language and not features of the page. Also, since these blogspam sites use very detailed language, they rank higher despite the language being just disguised keywords.

Ultimately it is up to google to do better if they want to stay relevant. The recent influx of custom search providers is i think a sign that the giant is faltering. People are looking at smaller players and are deciding "google isn't that good to begin with, surely these guys can do better?".

> I think ultimately unless they're whitelisting sites (like SO) there is no way to algorithmically differentiate between blogspam and genuinely useful tech blogs

It has always been utterly mindblowing that apparently not a single person with power in the entire chain of command at Google has ever recognized that websites showing exact replicas of content from the very small list of extremely famous moderated content generators like Stack Overflow should just be blackholed. The number of legitimate Stack Overflows (the genre of legitimate content producers, not the specific domain) is very small. The list could be successfully curated manually by a troglodyte.

Like, FFS, give google engineers a button they can click to mark a site as crap. They use the internet too and there are literally tens of thousands of them googling to find things on stack overflow all day every day. All they need is a "this site is an SEO scam" button.

Not doing this speaks directly to either malice or incompetence (personally I think both), because it's not a super obscure problem that nobody inside google has ever encountered before.

Why should Google punish the websites that go through the effort of applying their guidelines and best practices?

I can give you an example, while doing technical SEO I coordinated a project to implement structured metadata to the pages (for example if the page had FAQs we would use the structured data for that).

We went through that effort for Google to change their mind a few months ago. They decided to cut down on displaying featured snippets on search results, and with that we got a hit on our traffic. After following Google's own guidelines, and recommendations.

Dealing with this crap isn't enough punishment?

> Dealing with this crap isn't enough punishment?

Nope, that's not even close to enough punishment.

Google inevitably has to change their guidelines. Black-hat SEO attends every conference you did, and work it to their advantage. Google keeps trying to get ahead of them, but they always slip behind.

Your problem isn't with Google. Your problem is with your black-hat compatriots. That sucks, because you probably can't fight them on your own. Google has to, and you are the inevitable casualty of that fight.

(comment deleted)
It always feels wrong that websites with hard navigation and long useless texts that have high average time wasted (spent) will rank higher than human friendly one.
Google is culpable here. With it's seemingly infinite resources they have curated a fragile web, while failing to derank obvious spam and cloned content. To me, this seems more like inaction and misaligned incentives, which makes sense when you consider that the SEO and Adtech industries are bedfellows.
Self inflicted and so, deserved.

They could stayed neutral arbiters of content but they decided to incentivise websites to host ads, then let them fight to drive traffic to the site.

They deserve whatever doo-doo is accruing to them.

Theyre so clueless that they never seriously fought for social and now most useful (aka human generated) content is hidden behind login screens on Discord and Insta and the like.

I’m calling it now - now that most content we want to engage with is behind login walls, Pagerank/Search is anachronism.

Haven't used Google Search much since ChatGPT launched. Couldn't care less about garbage text flooding the web.
"Couldn't care less about garbage text flooding the web."

Isn't that what GPT is trained on, though? And future iterations as well?

Generators like GPT are also a big and increasing source of this garbage.
After GPT chokes on its own garbage...what next?
Apparently this happened with Google Translate: people were using it as a simple way to internationalize web sites* and the translation training started ingesting a large amount of its own exhaust. They had to stop automatically reading the web.

* I know this isn't i18n, but the kind of people who would try this are clearly too stupid to care, or perhaps know.

Now it's all neatly summarized in a single paragraph!
This reminds me of something Matt Levine said recently. It went something like "modern finance will find a way to make a risk-free asset risky". To me it's similar to the idea that reputation has value and hence destroying it to turn a profit can make economic sense. Hence corporate raiders destroying 100 year old brands in a couple years by extracting all good brand value. And then you have Google doing some sort of ranking where being at the top has worth. So then whatever algorithms there are get gamed which eventually leads to a loss in useful signal. Ignoring any critiques of mistakes or shortsightness of Google, it seems to me that this is in some sense inevitable.

I saw a poster (here on Hacker News maybe?) make the point that the fact that anonymous reviews on e.g. Amazon are no longer trustworthy is the historical norm. The fact that we had ~20 years were we could maybe trust people without their reputation was the weird part not the fact that you no longer can trust them. I mean personally I find it all sad, but I think it makes a lot of sense. Our trust in anonymous people has value and hence exploiting it (and hence destroying that trust) makes sense.

I think all of this is some human version of social entropy at work. Maybe it's a historical ebb and flow over time, but it seems like a natural result of our human nature.

> The fact that we had ~20 years were we could maybe trust people without their reputation was the weird part not the fact that you no longer can trust them.

I disagree, they did have reputation. The fact the those anonymous people were online and posting reviews during that specific timeframe in that specific environment served as a filter.

Once everyone was online, the filter no longer filtered, and so the reputation caused by the filter disappeared too.

I think that's a fair description of things. Taken things with your vocabulary, the reputation that existed was exploited for profit until the reputation was not more.
I think of the explotation of reputation as a form of arbitrage resulting from social norms loosening in the 1960s. It's another expression of the postmodern idea "Well, our institutions do untrustworthy things, so let's tear them all down!" At some point we realize we need institutions again.
Yeah I don't mean to say that it's all hopeless or something. More that there are forces that build up and forces that tear down and they are all kind of part of the same thing. All of this has happened before. All of this will happen again.
20 years ago I would spend hours browsing the Amazon.com music section because there would be hundreds of real-people reviews of records.

Today those types of honest reviews are long gone. I have to use all my spidey senses to figure out if a review was paid for, or even written by a human.

I realize I'm inviting the ire of HN, but I'm an SEO and fully agree with the sentiment of the article. Namely a lot of content on the web is purely for Google.

What some people miss is this isn't just SEOs gaming the system. This is a system that Google explicitly asks for. I've watched John Mueller's (a Google Search Rep) AMA sessions where he answers questions on SEO and he's advocated for webmasters to add text onto product listing pages to "help users" get a clearer understanding of the page. Nope. It's because even though Google has some of the best technology in the world, it's search algo is still pretty dumb.

No person needs a few sentences of text on an ecommerce page about couches to understand the page is relevant to couches and sofas, but Google does because it only understands text. That's also the reason why any recipe you find on the internet is buried at the bottom of the page.

The pointless adding of content to the internet is annoying as hell, but given that ChatGPT is able to create fairly cogent answers for a majority of queries it's, I'm frankly surprised Google isn't better.

Lastly, this article really only references creating content for SEO which is not all of SEO. Many parts of SEO actually do make sense. For example, making sure your content is actually visible onsite, redirecting old urls to new urls, making sure to internally link to pages people actually care about, etc... These should be website best practices but a ton of websites don't do them.

Good SEO is hard. Creating changes that are good for Google + users takes a lot of thought, effort, and testing. Bad SEO (i.e generating a bunch of pages/useless content for Google) is very easy and you don't need to know anything about actual SEO to do it. For some reason anyone able to look up a keywords search volume and write a post is an SEO.

I think the downfall of Google (the search engine) started when Google decided to host/participate in SEO conferences instead of keeping the relationship more distant (focusing on showing users the sites they want rather than helping people get users).
Couldn’t agree more. I have dabbled in SEO and it feels like the cost of doing business.

Google also seems to have it wrong these days. Unless I am searching for a specific website name Google search has been useless for me. General questions are often poorly met with results. I don’t find DuckDuckGo all that better.

The value of ChatGPT as implemented in Bing is it gives you the answer directly and then sites sources.

For technical searches I use google for maybe 1 search out of a 100 when DDG doesn't show what I need. And tbh, often google can provide better results. For the other 99 searches DDG is more than adequate.

And now using chatGPT more and more instead of search I think think that there's a big upset in progress. Should be interesting to watch.

> The pointless adding of content to the internet is annoying as hell, but given that ChatGPT is able to create fairly cogent answers for a majority of queries it's, I'm frankly surprised Google isn't better.

Yeah, and its not even that hard. Google could "interrogate" images or even website screenshots for keywords with open source tools. And YouTube is a juggernaut of speech-to-text and video analysis.

I know nothing about Google's innards, but I would guess decision makers aired on the side caution when dealing with their golden goose, rather than focusing on image and context analysis.

Can you say more on recipes? I had always heard that it had to with copyright - that one needed the copy above to create a copyrightable work.
I made another comment above asking about recipes which should answer your Q. But you may be right, copyright could be another component that I'm not aware of. However, SEO is certainly a large part of the reason for long recipe posts as it's definitely going to help get more traffic on average (to our collective detriment).
>That's also the reason why any recipe you find on the internet is buried at the bottom of the page.

I thought this was to game the metrics Google uses to see if a click out was useful. If you see a recipe and then bounce in 5s Google treats that as bad. If you scroll down for 10s looking for the recipe and then bounce after 5 it's not as bad.

While Google does seem to use engagement metrics to help with ranks I don't think it's as simple as you laid it out (which is fair you were likely oversimplifying for brevity). Google themselves have said just because you bounce quickly doesn't mean it's a bad result, could mean you found your answer and now you have another question based off of that.

My understanding from case studies I've read and tests I've seen done is 1) engagement metrics seem to be most useful at the top 5 rank positions in the search results 2) it's really hard to measure when improved engagement metrics actually improve ranks.

On the flipside adding content to a page usually results in fairly quick change in ranks and is relatively easy to track. And it makes sense. A normal recipe is a list of ingredients and an order/method to cook them. 100-300 words tops in most cases. Not a lot of info for a bot to understand what the context of the recipe is around.

Now if you spend another 750 words writing about how to do it Google gets a lot more context and reinforcement that what you're talking about is actually relevant. Keyword stuffing isn't a thing in that you can literally say the same word over and over and get higher ranks, but if you can stuff a post with relevant keywords sprinkled all over the place that's good enough for Google to say 'oh now I get what this page is about'.

So there may be some slight impact on engagement metrics (which is debatable since a lot of people find all that long text super annoying and will bounce because of it), it's the extra text/keywords that Google understands and values.

If it were the text they would put it after the recipe instead of before. That the recipe comes last means they want you to scroll past the text.
They put it after so you scroll past the display ads which are also on the page. If you didn't have to scroll down past the ads those ads wouldn't be monetized
> Many parts of SEO actually do make sense. For example, making sure your content is actually visible onsite, redirecting old urls to new urls, making sure to internally link to pages people actually care about, etc...

As you say, those are part of making a good website, not SEO specifically. I think it's a bit disingenuous for the SEO world to take any credit for them.

For a small site maybe, but you underestimate the challenge of some of these things for enterprise sites. If a new directory or slug is added to a site, what team in an organization is going to care if all urls are redirected correctly? For a site with tens of millions of pages, who is going to make sure they're all working the way we want? Who makes sure robots.txt files take into account those changes? Are sitemaps updated for crawlers? Are all our analytics tools in line with the new url structure? etc..

Every large site I've worked on doesn't actively think about these problems, and it's the SEO team who has thought about these problems.

Generally people care about their slice of responsibility. At most large organizations with enterprise sites, many parts of SEO falls outside of those slices of responsibility so people don't think about them. This is especially true if it's hard to put dollar values on them.

If someone is running a complicated site with many stakeholders and doesn't have a single "master" team that is devoted to the site, that's an obvious institutional problem that is orthogonal to SEO.
What's the name of this "master" team at these big organizations that take care of this stuff?
> That's also the reason why any recipe you find on the internet is buried at the bottom of the page.

I mostly agree with what you said in your comment, but you're wrong here. Ad monetization is the reason for recipes being at the end: it's related, but not directly Google's fault.

Having the actual content at the bottom means the proprietor can serve more ads interspersed with filler content, and since it takes longer to read the page, most visitors will stay longer than a few seconds decreasing the bounce rate (which improves SEO, amd more importantly, means you get paid for the ads as the time spent will hit the time-threshold for an ad impression)

I believe you are right in that " Ad monetization is the reason for recipes being at the end" - and maybe partially, maybe it's the only reason from the publisher's viewpoint - sure, maybe..

but ".. you're wrong here" - not necessarily - both things can be true..

and consider this: Google is choosing to put a bunch of web sites in the top results that instead of giving you the answer you seek (the recipe) - the sites contain the recipe but make you scroll past a bunch of other info - and ads - before you get what you want.

I've seen interesting debates about 'this increases engagement on-page time which the google algo likes ' - and things from googlers that say that's not the thing.. I've seen other guesses as to why..

regardless of the why - google is choosing to reward these sites with top position - they are long winded and include lots of ads.. so both things are true.

and the cynic in me leans toward people's paychecks and stock values are based on a thing.. and the search engine providing sites at the top that increase ad views, and they make money as an ad company.. and the old adage of do the best for the user is not at the forefront.

I've seen it in many other search phrases - things mysteriously happen that include censoring, that also increase the ad spend needed to reach customers.

One can blame in on this or that or point to whatever - but the truth of the outcome is the truth of the outcome - must of what people are searching for is buried and there has been in increase in ad visibility -

maybe that's not a bad thing for some, but it is what it is, and it's also an opportunity ripe for disruption with the likes of chatgpt type things - which is good and bad imho.

Also an opportunity for a more fun search engine - let google be the wanna be high brow prude censor machine it has become, and I think it will be the solid choice as a replacement for the yellow pages for quite some time.. but the best way to search and find many things connected to web, it is not.

Google, it's the new yellow pages. Both useful tools for the masses, and both an expensive must have for many local brick and mortar businesses.

As an SEO I don't see how Google is going to be able to fix this. I'm in the beta of the new SGE (Search Generative Experience) and it gives great generative information, but will be FORCED to show the sources front and center unless they want their ad revenue to collapse.

Given this, a new race to the top will commence by SEOs to try and get on that "3 pack" snippet of information.

I see so much gaming of SEO. One of my pet peeves when searching for some long tail technical topic: "[Sequelize] some long error message" for example. What I often see is Github issues on the SERP but above that you have some random blog that just scrapes from Github issues and adds an extra sprinkle of SEO to get to the top.
[flagged]
What is the alternative? How do you possibly have some sort of searching and ranking system that cannot in some way be gamed like this? Actually serious, not rhetorical. I'm curious if there is some way to prevent this mess or if it is an inevitable end-state.
As long as the web is chained to the ad world, there is no alternative. The end-state is obvious and inevitable. Advertising will ruin the web, just as it has ruined just about everything else that has cozied too much to it.
Ironically, LLMs could be a game-changer in dealing with SEO noise, not at the search stage, but during the indexing of web pages. By understanding context and semantics, they could help create a more refined and relevant index, effectively cutting through the clutter of SEO-optimized content.
One word: LLM-optimized
That's two words.
Hyphenated compound words are treated as one word. The same is true for closed compound words (e.g. chairman is one word even though it's made by combining "chair" and "man").