5 ms·

I mean you can't blame Google too much. In the early days, a large fraction of Internet users made websites and had hand-picked, curated links. This gave Goog

I mean you can't blame Google too much. In the early days, a large fraction of Internet users made websites and had hand-picked, curated links. This gave Google a fantastic ranking signal with a high signal-to-noise ratio. This is mostly gone now and honestly I'm surprised their search is as good as it is.

ipaddr 5y ago
You can blame google because they made changes that forced sites to adopt these practices. The hand-picked curated links websites still exist but are so far back in the index they will never show.
londons_explore 5y ago
Google still has archives of the internet from back in ~2005. They can still use the internet back then as a ranking signal for todays content. Ie. Imagine a now-dead blog which was very knowledgeable about types of violin and would have ranked very highly for "best type of violin string cleaner". Google can look at what content that blog had, and find a page on todays web with similar content saying the same kind of thing.
zozbot234 5y ago
The death of hand-curated general web directories like DMOZ has also deprived search engines of a hugely relevant "signal" for high-quality content. I'm not surprised that the SEO blackhats have basically won since then. (You can view the schema.org specification for machine-readable content description as an attempt by the big search engines to partially reverse this dynamic and give white-hat SEO the upper hand again. IMHO, independent website owners should enthusiastically adopt these detailed descriptions if they care to "save" the Web from the onslaught of blackhat SEO junk. But a worthwhile successor to DMOZ (probably based on federation) must also be a piece of that puzzle.)
- Dma54rhs 5y ago
  How does in your eyes schema help against spam? Product rating schema was instantly abused when it Google started using it which makes sense since you markup whatever and however the fuck you want.
  - zozbot234 5y ago
    If you markup stuff maliciously (ala the old meta keywords spam), you can be banned for it after a simple human check. OTOH, it's really hard to tell wrt. most low-quality SEO-format junk "is this malicious stuff that should be nuked from orbit, or just a clueless webmaster who doesn't know any better". It raises the stakes in a way that lets good content stand out if it chooses to.
uncomputation 5y ago
I think you have it slightly backwards. While not entirely Google’s fault, search engines motivated much of this “noise” increase. It’s not as though there were people who just made hand-curated, “high signal” websites and then all those people died. It’s even not that all those people switched to Wordpress or other CMS. It’s that the algorithms used by search engines directly incentivized a lot of the awful practices we now associate with the web. I think the first example of this, which is not nearly as apparent anymore, is it was more beneficial for a listicle to break up its elements into a separate page per element, requiring the user to click and reload an entire page to traverse the list, rather than just have the list on one page. This increased the number of back links if anyone wanted to link to multiple items on the list and allowed per-item SEO friendliness. Another more recent example is the infamous recipe blog. While some people genuinely like adding a backstory to their recipes, most add all of that fluff because Google penalizes short content (likely in an effort to reduce spam I would guess). This results in a weird lose-lose situation for everyone involved which has completely inundated the simple and extremely common search for a recipe. The only ones able to not do this are larger sites like Allrecipes which have enough reputation/clout in the eyes of search engines to avoid the spam classification.
- Lascaille 5y ago
  > larger sites like Allrecipes But the problem here really is the consumer. In the past, people would have bought a recipe book or subscribed to a cooking magazine and thought nothing abnormal about having to exchange money for data. Now the consumer expects quality data for free. Obviously something has to suffer.
brimble 5y ago
Google used to actively fight the spam. Some time around '08 or '09 it was like they very suddenly gave up and never seemed to give it a serious attempt again, as if they'd simply surrendered. Unfortunately, they also made searches much more "fuzzy" around that time or (IIRC) a couple years before, foiling manual attempts to avoid spam by using very specific language or unusual phrases.

ipaddr 5y ago

You can blame google because they made changes that forced sites to adopt these practices. The hand-picked curated links websites still exist but are so far back in the index they will never show.

londons_explore 5y ago

Google still has archives of the internet from back in ~2005. They can still use the internet back then as a ranking signal for todays content. Ie. Imagine a now-dead blog which was very knowledgeable about types of violin and would have ranked very highly for "best type of violin string cleaner". Google can look at what content that blog had, and find a page on todays web with similar content saying the same kind of thing.

zozbot234 5y ago

The death of hand-curated general web directories like DMOZ has also deprived search engines of a hugely relevant "signal" for high-quality content. I'm not surprised that the SEO blackhats have basically won since then. (You can view the schema.org specification for machine-readable content description as an attempt by the big search engines to partially reverse this dynamic and give white-hat SEO the upper hand again. IMHO, independent website owners should enthusiastically adopt these detailed descriptions if they care to "save" the Web from the onslaught of blackhat SEO junk. But a worthwhile successor to DMOZ (probably based on federation) must also be a piece of that puzzle.)

Dma54rhs 5y ago
How does in your eyes schema help against spam? Product rating schema was instantly abused when it Google started using it which makes sense since you markup whatever and however the fuck you want.
- zozbot234 5y ago
  If you markup stuff maliciously (ala the old meta keywords spam), you can be banned for it after a simple human check. OTOH, it's really hard to tell wrt. most low-quality SEO-format junk "is this malicious stuff that should be nuked from orbit, or just a clueless webmaster who doesn't know any better". It raises the stakes in a way that lets good content stand out if it chooses to.

uncomputation 5y ago

I think you have it slightly backwards. While not entirely Google’s fault, search engines motivated much of this “noise” increase. It’s not as though there were people who just made hand-curated, “high signal” websites and then all those people died. It’s even not that all those people switched to Wordpress or other CMS. It’s that the algorithms used by search engines directly incentivized a lot of the awful practices we now associate with the web. I think the first example of this, which is not nearly as apparent anymore, is it was more beneficial for a listicle to break up its elements into a separate page per element, requiring the user to click and reload an entire page to traverse the list, rather than just have the list on one page. This increased the number of back links if anyone wanted to link to multiple items on the list and allowed per-item SEO friendliness. Another more recent example is the infamous recipe blog. While some people genuinely like adding a backstory to their recipes, most add all of that fluff because Google penalizes short content (likely in an effort to reduce spam I would guess). This results in a weird lose-lose situation for everyone involved which has completely inundated the simple and extremely common search for a recipe. The only ones able to not do this are larger sites like Allrecipes which have enough reputation/clout in the eyes of search engines to avoid the spam classification.

Lascaille 5y ago
> larger sites like Allrecipes But the problem here really is the consumer. In the past, people would have bought a recipe book or subscribed to a cooking magazine and thought nothing abnormal about having to exchange money for data. Now the consumer expects quality data for free. Obviously something has to suffer.

brimble 5y ago

Google used to actively fight the spam. Some time around '08 or '09 it was like they very suddenly gave up and never seemed to give it a serious attempt again, as if they'd simply surrendered. Unfortunately, they also made searches much more "fuzzy" around that time or (IIRC) a couple years before, foiling manual attempts to avoid spam by using very specific language or unusual phrases.