Will Google Penalize Itself Again for Using Advertorials? | Hacker News Reader

Will Google Penalize Itself Again for Using Advertorials?

searchengineland.com

38 pointsfraqed20 comments

How does Google determine wether a link is paid or not? I stopped using nofollow at all on my personal site, because I figured if I like something enough to invite humans to go there, I might as well "vouch" for those links in the sight of search engines; and while I'm pretty sure a human would never flag the site, when reading about algorithms (in this case, without hearing how they actually work) I sometimes get a bit nervous, just because I'm doing my own thing without paying too much attention to what is "standard". Just take HN for example: why doesn't it pass link juice to the stories, wouldn't that help improve the web/search results? And if that got changed, would Google act all stupid, or simply enjoy and use the additional signal provided?

I believe that user-content sites use nofollow because it reduces the amount of spam (at least, that's the common wisdom), since without it, a spammer would be able to leverage the high rating of the sites (e.g. HN, Reddit, StackOverflow).

Server Fault had an interesting experiment. They dropped the "nofollow" on links from users that they considered trusted (based on age and other criteria). Then they found their rankings were dropping because those links were considered "spammy" by google.

Here's a post related to that by Jeff Atwood:

http://meta.stackoverflow.com/questions/51136/remove-nofollo...

6 months later they fixed the problem: http://meta.stackoverflow.com/questions/111279/remove-nofoll...

Thanks for the link. It seems to me that Google should be better able to deem links reputable than most sites. It seems a weak decision by Google to threaten sites into the safe action of just telling Google not to value any of their links (using nofollow) rather than Google determining value.

I realize lots of people try to trick Google. But I really think Google should be able to produce value by accessing links themselves. I can figure out lots of nofollow links that have value and really can't believe Google doesn't realize it is providing less useful search results because of a policy decision they made years ago to make their lives easier (telling sites they had to use nofollow and Google then wouldn't value the links).

I realize battling "search spam" is difficult. I think the nofollow strategy isn't a very good one. But I don't expect Google to improve until a competitor uses this weakness to gain significant market share (or maybe if google is thinking ahead and monitors search result value and notices a competitor is ahead - even before the users do they will react before market share declines). But no competitor has done so, so Google continues to punish sites and Google users with the Google nofollow policy.

I can understand it for something for a forum, or links posted in comments, but otherwise, I still wonder..

"a spammer would be able to leverage the high rating of the sites"

Sure; as would legitimately useful content. A spammer can already take advantage of eyeballs, so taking of nofollow doesn't mean you don't have to remove spam; and if you remove spam, what's the use of nofollow? As a safety net? Well, one could e.g. take nofollow off for links that have a lot of upvotes etc.. I'm just not convinced that even the harm by some random spammer getting some search engine love for a short amount of time outweighs the long term benefit of an "honest" graph of relevance and connectedness, if that's not throwing out the baby with the bathwater.

Sites like HN and reddit have a high rating partly because they have links people like, and do not consider spam, and nofollow might be presenting a completely broken picture to search engines of what is actually going on. That's fine by me, I'm not complaining, but I really do hope being a black hole of page rank doesn't become some kind of status quo, against which sites which do it "my way" would then stick out negatively.

> and if you remove spam, what's the use of nofollow?

For the period in between when spam is posted and when you remove it, during which there is an extremely high chance Google has already indexed it.

Don't submissions here get screened? I was under that impression. Also, this doesn't apply to stuff that has hundreds of votes.

But even ignoring all that, just as an Gedankenexperiment: Yeah, and? Stuff has to stick around for quite a while to rise, domains have to be known, etc. It's easy to get "into the index", just as easy to drop out of it, but when it comes to any real traction even a front page post for hours is just a blip.

I thought we were speaking in general not about HN specifically.

In any case, no they are not screened. Submit a story and witness it appear instantly at http://news.ycombinator.com/newest

I thought we were speaking in general not about HN specifically.

Not just HN, but I replied to this, so no: "(e.g. HN, Reddit, StackOverflow)"

HN has some lists of sites that will always be dead, and some hellbanned users who's posts will always be dead.

All posts go to (http://news.ycombinator.com/newest). It's a good idea to visit that page and upvote the great content, ignore the not so great content, and flag the really bad stuff. I think, but I could be wrong, that any user can flag an article. But maybe there's a karma threshold.

>if you remove spam, what's the use of nofollow?

To prevent a search engine from poking around links and executing actions meant from human beings.

I help someone manage a small online music stream with automatically queued user requests. The community around it is small and nice so there's no need for captchas or authentication. Without nofollow Google and friends would randomly request tracks for playback.

    To prevent a search engine from poking
    around links and executing actions
    meant from human beings.

That's not how search engines interpret nofollow. You should use robots.txt instead.

EDIT: for details, see the original blog post, which says "those links won't get any credit" not "those links will be ignored": http://googleblog.blogspot.com/2005/01/preventing-comment-sp...

Its how I understand no follow - Used in login and print page links and the like to stop the creation of massive numbers of duplicate pages.

It works for that, because it means "disregard this link for ranking". But it doesn't forbid bots from crawling those links anyway, which is what the other poster seems to think it does.

Well, according to the logs we don't get requests from Google or other kinds of bots so it works for our use case.

You can solve that problem by using POST rather than GET for the playback requests. (I.e. convert each playback request link to a <form> with hidden fields specifying the track info and a "request it" button, or just use ajax and "new XMLHttpRequest('POST')" or whatever is appropriate for the JavaScript library being used.)

Refactoring 4 years old PHP that spits out quirks mode table-layed HTML via echo... shudder.

I just do not understand this Do-follow and nofollow. I don't want to. I use captcha and give some credit to those who spam manually.

You wont like it when you get a UGC spam penalty though - had to sort one of those out for one of Reed Elsivers sites.

Reply on news.ycombinator.com