| 3:10 am on Jun 26, 2011 (gmt 0)|
I dont think it should ever be eliminated as a factor. The thing is, years ago it was the biggest factor by far, probably like 90%+ and therefore made it easy to game google.
I will never make as much today from google as I did 5-10 years ago. I am thankful I took advantage of google and was able to build up a huge nest egg. With that said, I am now creating better websites, creating REAL businesses that are useful to the people who find these websites. So google is making the web better in that sense.
Backlinks should be 1 of MANY factors in determining how a site ranks. There will always be ways to game the system, its about making it as hard as possible so that quality sites get priority over MFA/crap sites that have no real value.
| 3:16 am on Jun 26, 2011 (gmt 0)|
I know this is the Google forum however Bing says it is THE dominant ranking factor inside their webmaster tools accounts.
Links aren't going anywhere any time soon.
| 4:42 pm on Jun 26, 2011 (gmt 0)|
They're most likely just going to change how they view them, as it was stated by the spam team member on hacker news. To me, this means over anchor text optimisation is going to cause REAL problems in the near future.
Links are never going to not be a factor, so I don't know what you were smoking to come up with that :)
| 4:48 pm on Jun 26, 2011 (gmt 0)|
No way, they're still going strong. As linkbuilder says, links stay, they just measure different attributes about links. If you remove SE's, links are still the roadmap of the internet.
Twitter might slowly measure in, but that's just another source of link data. And I doubt the SE's are going to tie their measurement of success on data from one single website, it's not comprehensive and too risky for them.
I don't see anything looming that's going to give as strong a signal of authority as a link from something like a strong .gov.
| 10:07 pm on Jun 26, 2011 (gmt 0)|
linkbuildr - That's like saying the Earth is flat. Never say never. It does not always have to be a link that references your website or brand, it might evolve to the point where your name gets mentioned in a title or a headline for example. I don't pretend I know where its going but I think we are certainly heading away from heavy link based scoring as we know it today.
|brotherhood of LAN|
| 10:13 pm on Jun 26, 2011 (gmt 0)|
IMO links will still be the SE's 'currency' for a long long time, can't see a better replacement.
How and why they measure them will likely progress leaps and bounds mind you.
| 10:36 pm on Jun 26, 2011 (gmt 0)|
It's not just about links; "citations" (linkless mentions) are also important.
| 10:51 pm on Jun 26, 2011 (gmt 0)|
|...can't see a better replacement |
I think that's a key statement. When a better replacement comes along I think links will devalue further. I think they will always play some part even if it's an ever dimishing role but any SEO factor that can be gamed has to be a target for the SE's to shoot at.
Actually there is a replacement - it's just not practical. A manual review :-)
| 11:01 pm on Jun 26, 2011 (gmt 0)|
why is manual review for all impractical ?
| 11:21 pm on Jun 26, 2011 (gmt 0)|
Manual review just won't scale, not even to the current size of the web, to say nothing of the future. And it won't supply ranking parameters for all possible queries, either, or account for the daily, hourly churn of URLs online.
Web search is not even close to the same as a directory, like Yahoo was at the start.
| 11:26 pm on Jun 26, 2011 (gmt 0)|
|It's not just about links; "citations" (linkless mentions) are also important. |
You think this is already in play? I'm wondering what a mention in a NYT headline might be worth.
| 11:56 pm on Jun 26, 2011 (gmt 0)|
Links will always remain an important factor for ranking by search engines
| 10:05 am on Jun 27, 2011 (gmt 0)|
Links are still the king for old testament SEO but don't think for a second that the new ranking currencies like social can't be gamed just as easily.
| 11:00 am on Jun 27, 2011 (gmt 0)|
My scrapebox is hammering it out as we speak :)
| 12:24 pm on Jun 27, 2011 (gmt 0)|
|Links are still the king for old testament SEO but don't think for a second that the new ranking currencies like social can't be gamed just as easily. |
Sites that are ranking that have no backlinks but are kicking butt with social media? Like, none.
The social media currency is worthless to businesses looking to rank. It's probably pretty good spending money for SEO companies looking to sell services to clueless companies though (I'm doing some of that myself).
| 4:50 pm on Jun 27, 2011 (gmt 0)|
Google spokespeople have said that the algorithm only counts natural backlinks that are created by other people without any involvement on your part. This would mean that links you buy or build yourself may not help your rankings. In other words, the algorithm may discount certain types of links, especially types created by common link-building methods.
| 6:11 pm on Jun 27, 2011 (gmt 0)|
|Google spokespeople have said that the algorithm only counts natural backlinks that are created by other people without any involvement on your part. |
With all due respect, that is, as far as I can tell, one of the LEAST accurate statements google has made about the way their own algorithm works.
If I were allowed to post links in this forum, I could show google how about 8 of the top 10 sites returned in the SERPs for a two-word phrase rely on unnatural link building. Heck, since the other two sites in the top 10 are an ehow and a wikipedia page, we could say that they rank by unnatural INTERNAL link building.
I could also show google a few sites with great natural links that are slowly but consistently losing ground in the SERPs - and these are competitors' sites, not mine, so I am not whining about some perceived injustice against my site.
Also, we have the Sears incident that pretty much disproves google's statement - although in the case of Sears they were at least getting in content links, as opposed to blog comments and directory listings and forum profiles, etc - at least as far as we know.
| 6:53 pm on Jun 27, 2011 (gmt 0)|
Its a minefield out there linking wise now.
I urge everyone to do a google trends on "backlinks" and look at what that implies.
I also urge everyone to EXPERIMENT - because I have listened to too much dishonesty and inaccurate information in the past - and in actual fact - theres plenty of evidence out there which is crying out right in front of everyones face which basically says this....
Its fine - go spam the hell out of google with spam links of every single kind you can get your hands on. The more the merrier. Its their fault - let them deal with it.
| 6:55 pm on Jun 27, 2011 (gmt 0)|
Of course - do this on domains which earn a fast buck and then get binned - you wouldnt want them changing their algo in the furture and finding your pension site in the trash can.
Having said that - they will probably trash your pension site anyway in the long run.
| 11:05 pm on Jun 27, 2011 (gmt 0)|
|Also, we have the Sears incident... |
Arrgghhh... meant to write JC Penny. Oh well, same thing...
| 11:52 pm on Jun 27, 2011 (gmt 0)|
|Back links as ranking factor must be on the way out sooner or later. Its a messy "game able" system and a long way from perfect |
Threads about the death of links always fail to suggest viable replacements that are not just as easily gameable.
| 1:17 pm on Jun 28, 2011 (gmt 0)|
If links are a proxy for quality and relevance, then conceivably, machine intelligence will eventually be able to score a site for quality and relevance without the crutch of looking to other sites for "votes".
Isn't this the shift we are seeing in recent algo updates? It may not be on the immediate horizon, but eventually an algorithm will be able to judge the importance of a site without trusting other sites to do the job for it.
| 2:36 pm on Jun 28, 2011 (gmt 0)|
visitor interaction will be the future of "sustained" search rankings. It's track-able and not as easily gamed.
sites will be categorized, basically informational or transactional (ecom).
1. informational site visitor interaction.
a. visitor stays on site (site is destination type, good match for query)
b. visitor returns to search results and selects different result (site is ok, maybe not the best match for query)
c. visitor continues on to another site via link (site type is doorway, likely that the next site visited via link is best match for query)
2. transactional site visitor interaction measurement:
a. conversion rate
b. there is no b, don't want to tell google your conversion rate they will happily introduce you to adwords.
this is of course overly simplified but I can picture something to this effect on a whiteboard somewhere.
| 4:24 pm on Jun 28, 2011 (gmt 0)|
|If links are a proxy for quality and relevance, then conceivably, machine intelligence will eventually be able to score a site for quality and relevance without the crutch of looking to other sites for "votes". |
Then you should try to get AS MANY LINKS AS POSSIBLE NOW - or start to make your web site really crappy!
Because in my niche, the crappy sites have all the links. So google is going to look at those sites and say, "These sites have lots of links, therefore they are quality sites, therefor sites that are similar to these sites must be good and will henceforth rank highly in google."
visitor interaction will be the future of "sustained" search rankings. It's track-able and not as easily gamed.
It's only somewhat trackable and it is easily gamed.
Even with google analytics, there are problems tracking how much time someone spent on a page. I believe if someone goes to your page and reads it for five minutes, then picks up the phone and orders something from you, but closes the browser, then it is considered a bounce, and the five minutes they spent on your one page doesn't get added to the time on site.
And not every site had google analytics on it. I am sure that if bing ever had an analytics service, many people would switch just to keep their stats hidden from google.
As for not being gamed... well, people still game adwords with click fraud. Also, if there is software that can go to hundreds of thousands of websites and automatically create forum profiles, directory listings, blog comments, what makes you think that there isn't software out there that will rotate through proxy servers and spoof user behavior?
I am sure I could easily spoof a high bounce rate for my competitors and a low bounce rate / high time-on-site for my own pages.
| 4:34 pm on Jun 28, 2011 (gmt 0)|
|what makes you think that there isn't software out there that will rotate through proxy servers and spoof user behavior? |
They're called botnets and they're available for rent right now.
Of course that's illegal. To do it legally you'd have to hire thousands of people on individual computers to take some action related to your website.
Not hard to do. Have a windows program that calls home for instructions. The instructions tell the program to send out http requests behind the scenes, i.e. query Google for a search term. Scan the page for a specific result. query that page with an http request so now Google sees that a visitor landed on the page. etc.
Sell it to college students for a once a month $50 gift certificate to their local pub. All they have to do is run a program 24.7 that does little more than hit some websites in the background. There you go, instant legal botnet you can sell to SEO companies. It'll be the new face of paid blogging :).
Meet the new firefox plugin that mimics visitor behavior - mozmimic.
| 5:49 pm on Jun 28, 2011 (gmt 0)|
like I said, not AS easily gamed (as link building) and with user behavior becoming more track-able everyday (think Chrome, google toolbars, google accounts, being perpetually logged in (facetwittspace...), and supercookies) it's already easier to tell the difference in a botnet and a real user than it is to tell the difference in a paid and non paid link.
| 6:08 pm on Jun 28, 2011 (gmt 0)|
Speculation though. I suspect the SE's are finding there's a lot of noise in the signal. And they can't count on the signal - way too easily biased.
Links as something that can be measured aren't going away AND they're a broad overview of the web. Measurements of user behavior can go away. Chrome can disappear from use. cookies can't be trusted. Much of this is transitory in nature - they jump over to using the Google PR toolbar as a reliable indicator and people upgrade to a browser that doesn't have it, they're toast. Plus, if they used this stuff in any noticeable way I'm pretty sure it's way easier to game than links.
Same problem with using social media. Google's not stupid enough to substantially change their entire algo to depend on data from a single website like twitter, particularly when they don't own twitter. If twitter feeds start actually counting substantially in the algo, and MS buys twitter, how's that working for Google's search results?
| 6:58 pm on Jun 28, 2011 (gmt 0)|
Here's my speculation. I think there's already a section of scoring in the algorithm we could call "engagement".
We've already been talking about many of the factors that could go into that score. I could be a complex combination of social media, contact information of various types, newsletters, RSS feeds... on and on. All the ways that a website actually gets involved with their visitors.
The final, aggregated engagement score would be combined with relevance, PR and "quality" scores. A site would not need to show ALL the signs - a well read newsletter but no RSS feed? You're still showing engagement.
| 8:54 pm on Jun 28, 2011 (gmt 0)|
that theory works great if you comparing NYT with Washington post over a month or so. You can argue that G can compare your shoe store with other shoe stores but they are so many in betweens for this to be unreliable. That said, I fear that Google has already tried to something like in Panda.