Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

More content by people, for people in Search

blog.google

Post reply on HN
Print view View original

211–220 of 320 posts

Re: More content by people, for people in Search

#211

Post by rchaud »

whage wrote: post #76
firasd wrote: post #32

I'm not sure this whole drive towards 'original content' makes sense. It encourages people to make long rambling pages that don't get to the point In fact it seems like Google worked better when wikipedia was almost always the first result on various topics What I mean by rambling is if you type something like "types of oranges" and suddenly land on a page where some dude is clearly just filling up paragraphs for goo…

This resonates with me so freaking much. I am disgusted by how every bit of information on webpages seem to be buried between paragraphs of empty EMPTY talk and even when they seem to get to the point, most of the time there is no real information there. Sorry, this had to come out. We really need better search engines.

This is true of most non-fiction literature as well (self-help and business books especially), so I'm not sure if Google can necessarily discourage people from adding unnecessary padding.

Google is already scraping bits of website content and showing it to the user as a "featured snippet". Nobody is going to write short pages if Google can already rob you of a click so easily.

Re: More content by people, for people in Search

#213

Post by bradhilton »

For medical advice I would see more serious articles/blog posts in search results and fewer listicles.

For meal recipes I would just like to see the recipe and not a blog post with dozens of ads.

For programming questions I don't want to see the sites that just scrape and repackage stackoverflow questions; for a company run by software developers they have to know about this problem lol.

Re: More content by people, for people in Search

#214

Post by rchaud »

medion wrote: post #138
firasd wrote: post #32

I'm not sure this whole drive towards 'original content' makes sense. It encourages people to make long rambling pages that don't get to the point In fact it seems like Google worked better when wikipedia was almost always the first result on various topics What I mean by rambling is if you type something like "types of oranges" and suddenly land on a page where some dude is clearly just filling up paragraphs for goo…

Encourages people to make long rambling pages which don’t get to the point? Ever seen YouTube, it’s exactly the same. 10-11+ min videos saying almost nothing because it’s the most lucrative length. The internet of today is just everyone gaming everything for a dollar.

In the case of Youtube, the video thumbnail itself is usually a strong enough signal as to whether it's worth watching.

Enormous swathes of the platform are filled with gossip channels (politics, crypto, stonks, celebrities, music) hosted by clout chasers desperate to be seen as authorities on that topic.

Of course there can be good content inside of these categories, but you can generally tell which channels are "optimized for engagement", i.e. run like a business, and those are more amateurish and perhaps more authentic.

Re: More content by people, for people in Search

#215

Post by snowwrestler »

firasd wrote: post #32

I'm not sure this whole drive towards 'original content' makes sense. It encourages people to make long rambling pages that don't get to the point In fact it seems like Google worked better when wikipedia was almost always the first result on various topics What I mean by rambling is if you type something like "types of oranges" and suddenly land on a page where some dude is clearly just filling up paragraphs for goo…

Those types of rambling posts are clearly one of the targets of this update. From the Search Central guidance, here’s a question for site operators to ask themselves to “Avoid creating content for search engines first”:

> Are you writing to a particular word count because you've heard or read that Google has a preferred word count? (No, we don't).

https://developers.google.com/search/blog/2022/08/helpful-co...

Re: More content by people, for people in Search

#216

Post by Spivak »

layer8 wrote: post #198
seanwilson wrote: post #174

> Next week, we'll launch the “helpful content update” to tackle content that seems to have been primarily created for ranking well in search engines rather than to help or inform people. This ranking update will help make sure that unoriginal, low quality content doesn’t rank highly in Search, and our testing has found it will especially improve results related to online education, as well as arts and entertainment,…

> What about letting users give a vote/rating or leaving a comment on if a page was helpful? Difficult for such a system not to be gamed, e.g. via botnets.

And difficult for the system to not be gamed by brigadiers.

Also voting is an extremely skewed signal, what people say they like and what they actually like are very different and with search you can actually measure that by seeing which results users actually go to and don’t bounce from.

Re: More content by people, for people in Search

#217

Post by rchaud »

bradhilton wrote: post #213

For medical advice I would see more serious articles/blog posts in search results and fewer listicles. For meal recipes I would just like to see the recipe and not a blog post with dozens of ads. For programming questions I don't want to see the sites that just scrape and repackage stackoverflow questions; for a company run by software developers they have to know about this problem lol.

> For meal recipes I would just like to see the recipe and not a blog post with dozens of ads.

Why not just buy a cookbook? It doesn't seem like the convenience of accessing recipes at the touch of a button is worth the tradeoff of ads.

Re: More content by people, for people in Search

#218

Post by injidup »

SCLeo wrote: post #200

I write novels as a hobby and I publish them on my website. I recently noticed that there are websites that blatantly copy and sell my novels without my permission and are ranked higher than mine. I think it is because Google identifies their pages as the source. If anyone knows how to deal with issues like this, please let me know.

With what license do you publish them?

Re: More content by people, for people in Search

#219

Post by rchaud »

corille wrote: post #209
dgudkov wrote: post #133

I don't think Google will ever be able to distinguish texts that are written for people from texts that are written for Google. It's effectively the Turing test, and Google won't pass it.

The way I read it, this is aimed at derivative texts that were scraped from other sites. If Google indexes those sites, they can recognize when the content of a newly indexed site is copied from elsewhere.

Most of the blogspam on the Internet today are website filled with product 'reviews' and affiliate links. These aren't derivative of anything, so they won't be downgraded. Neither will the hordes of blogs breathlessly reporting celebrity and business gossip. A search engine won't be able to differentiate between the quality of a TMZ vs say Variety.

Re: More content by people, for people in Search

#220

Post by layer8 »

Spivak wrote: post #216
layer8 wrote: post #198

Earlier quoted context omitted.

> What about letting users give a vote/rating or leaving a comment on if a page was helpful? Difficult for such a system not to be gamed, e.g. via botnets.

And difficult for the system to not be gamed by brigadiers. Also voting is an extremely skewed signal, what people say they like and what they actually like are very different and with search you can actually measure that by seeing which results users actually go to and don’t bounce from.

I wonder about the bouncing, because I often either (1) check out a couple more results, only to find that they aren’t better than the first acceptable one, or (2) open the first ten or so promising search result in new tabs in one go before looking at the actual pages. Google can only make limited inferences about result quality from that, meaning that its ratings become skewed toward less sophisticated users.
Post reply on HN