| 11:17 pm on Mar 31, 2005 (gmt 0)|
I wonder how buying a banner link in a portal fits into 'purchasing links'.
That'll put the jitters up a lot of legitimate advertisrs.
...but then, just to play devil's advocate, isn't Google guilty of selling links?
| 11:18 pm on Mar 31, 2005 (gmt 0)|
I think it is the rate of aquisition and the mix -- the patent document is fascinating -- haven't had a chance to digest fully yet -- only scan
| 11:19 pm on Mar 31, 2005 (gmt 0)|
Do you think that there may be a difference (from Googles point of view) between more or less ordinary rate of growth of reciprical links and a
|A large spike in the quantity of back links... |
| 11:26 pm on Mar 31, 2005 (gmt 0)|
If someone spent a fair bit on budget online ...that might create a 'large spike'. ..His he spamming?
| 11:39 pm on Mar 31, 2005 (gmt 0)|
As eZeB said it seems to be in the mix.
Get oneway links from a bunch of non-SEO IPs around the web and you can get away with a bit more. And that mix includes not just the direction of links but the Pr ratio and the newness as well.
What's the right % recipes for a balanced link portfolio? I/we can only guess, but ask the folks applying for the pattent for the real answers.
[edited by: Tigrou at 11:51 pm (utc) on Mar. 31, 2005]
| 11:41 pm on Mar 31, 2005 (gmt 0)|
sorry I missed something there.
| 11:45 pm on Mar 31, 2005 (gmt 0)|
ken_b makes a good point. And it leads me to ask more questions.
- What is an ordinary rate of growth?
- Does Google consider reciprocal links an ordinary aspect of the development of links?
| 11:50 pm on Mar 31, 2005 (gmt 0)|
Does this mean that if you buy 'inward bound' links as opposed reciprocal links, it will make a difference?
If you spend a huge amount of money on buying banners in portals or directories that create a huge spike ..are you a spammer?
| 11:53 pm on Mar 31, 2005 (gmt 0)|
I think another important question is what is the result of building links too fast? Penalty? Slight drop in ranking? Large drop in ranking? Ratio of drop in ranking to how large the amount of links you got? Ban?
It also brings up the obvious - if large spikes in backlinks can actually hurt you, then you can harm your competitor.
If large spikes can't hurt you, then who cares?
| 11:55 pm on Mar 31, 2005 (gmt 0)|
I like that, build a dumping ground for all our competitors and create a 'large spike'.
Naughty ..but nice!
| 1:06 am on Apr 1, 2005 (gmt 0)|
In light of this it might be interesting to consider the percieved difference between a "limk exchange" and two sites that just happen to link to each other.
The difference could simply be measured in the time span between when the two links appear.
In a "link exchange" the two links are likely to appear with-in hours, or days, of each other.
In randomly occuring (in other words, not arranged) reciprical links there could months in between the time the first one appears and then the other.
In arranged exchanges the likelyhood of a spike appearing would to be a lot greater than with random recips.
| 2:02 am on Apr 1, 2005 (gmt 0)|
Martinibuster, I think you know where I stand on link exchanges but for those who don't ... I don't participate in link exchanges and don't believe in them at all.
I believe in putting your site out there and allowing your links to grow naturally. The more authoritative content you provide, the faster the links will accummulate.
I believe that Google keeps a very close eye on link exchanges and also that they compare similar sites (ie. your competitors) in regards to total link volume.
If you give it some thought, a site which has been around for ten years is very likely to have many more links than a site which has been around only three years ... but that doesn't mean they can't compete on the same level!
If you take the ten year old (authoritative) site and gauge the volume of link growth on an annual basis, you can find a mean average for "natural link growth" and you might even be able to identify and discount unatural growth spikes.
It is my belief that if the ten year old site has 1000 links and the three year old site has 300 links, they may be considered equal in quality and authoritative content if it is determined that 100 natural links per year in a given sector is the recognized "mean average" for "similar sites".
Don't know if you follow my logic ... but its what I truly believe and as a result, I just let things happen naturally. My site competes very well with just such an "authority" site which is much older than mine ... but to be honest, my site is much more up to date and has a greater amount of useful information.
I expect that by this time next year, my site should surpass the older and more established site in the SERPS for almost any search because I assume my site will get quite a few more natural links from a greater number of sites than the other guy will. His site is outdated and because of that, his site should be in the downhill curve for gaining new links while mine is going uphill.
|Is exchanging links something Google frowns on? |
While I have no proof, I would say my guess is yes!
| 4:15 am on Apr 1, 2005 (gmt 0)|
>> Is exchanging links something Google frowns on?
To ignore is ok .. but to frown on, is bad and promotes extreme antisocial behavour.
Exchanging links, ideas, contacts, favours, drinks are a part and parcel of forming networks, both online and offline.
Does this mean if I give MB a cigar at the next pubconf, in appreciation of his fine work and he buys me a drink the next time he sees me at the bar... someone else would have a frown on his face?
The problem starts when and if MB only buys drinks for people who give him cigars.
(Was going to use a political campaign contribution analogy ... but figured this one is more fun)
| 4:20 am on Apr 1, 2005 (gmt 0)|
|Does Google consider reciprocal links an ordinary aspect of the development of links? |
Yes...to an extent. All things in moderation. SOME reciprocal links is natural. 75% reciprocal links is not natural.
I think you'll be seeing "natural" tossed around as the newest buzzword being repeated infinitely by SEO's in the upcoming months.
'Natural' is being determined algorithmically though...which means there have to be threshholds. Perhaps threshholds based also on other interdependant criteria, but indeed absolute quantitive threshholds. The REAL fun is gonna come as the SE's get better and better at blending their English (semantics) and their Math (algorithms).
Are recip links inherently evil?
Nope. Happends a lot.
Could too man recip links work against your optimization efforts?
Yep. It's all about balance.
I owe ya a drink in N'awleans for a fine chuckle shri.
| 4:30 am on Apr 1, 2005 (gmt 0)|
|I think you'll be seeing "natural" tossed around |
Or maybe natural "looking"...
| 4:44 am on Apr 1, 2005 (gmt 0)|
Every beautiful woman knows that "natural looking" can be enhanced by some strategic artifice. But not toooo much! Also, underneath the artifice she needs a base of good health and good attitude if she wants to sustain her beauty.
Link development is kinda like that .... it is indeed all about balance.
|I think you'll be seeing "natural" tossed around as the newest buzzword |
Some of us have been singing that song for ages already.
| 5:31 am on Apr 1, 2005 (gmt 0)|
I owe ya a drink in N'awleans for a fine chuckle shri.
Cigars are on me.
| 5:43 am on Apr 1, 2005 (gmt 0)|
I stop answering daily emails for rcp links ,in fact what boost your page high ranking and placement in the results are one way links , ie referals from authority sites edu or gov and you can achieve that only if your content is unique.I have seen pages with tones of backlinks i mean thousands and they get maximum a PR 5 bad placement in the serps and pages with a few hundred a PR7 or 6 and placement at the top 10,it proofs that rcp links are not so important anymore ,that's what i believe.
| 6:22 am on Apr 1, 2005 (gmt 0)|
Stuntdubl has his finger on that one. I agree, "natural patterns" or as ken_b and buckworks elaborate, "natural looking," may be on people's minds going forward.
| 9:51 am on Apr 1, 2005 (gmt 0)|
Ok so far I agree with this idea of natural mix, and natural growth.
But were does that led us to?
I am seeing lots of ´directory´ sites popping up. There you get your relevant sub category (dmoz like) one way link for cheap (often free for the moment).
| 10:20 am on Apr 1, 2005 (gmt 0)|
|If someone spent a fair bit on budget online ...that might create a 'large spike'. ..His he spamming? |
Below is what happens :)
|I think another important question is what is the result of building links too fast? Penalty? Slight drop in ranking? Large drop in ranking? Ratio of drop in ranking to how large the amount of links you got? Ban? |
There is no penalty. There is no gain. Google automatically post dates all these links, aka link devaluation. The gain might only get factored in after your links come out of a link sandbox which is at this moment is an unknown.
The emergence of the link buying market has fuelled this. In discussion with some text link brokers, they warn that links will not give value for a certain period of time. I for one, never buy text links, I think its a rip off, and unless you know what you are doing, you will get very very little ROI!
Brett is going to release a new post :P
" A successful site in 36 months "
-12 months sandbox
-12 months link penalty
-welcome to ground zero.
| 1:23 pm on Apr 1, 2005 (gmt 0)|
|In a "link exchange" the two links are likely to appear with-in hours, or days, of each other. |
Given the delays in free link approval, the above could be just one of many possibilities.
I am hearing enough alarm bells from directory owners (not necessarily link directories) to think that directory links are deprecated until proven innocent. It should be easy for SEs to have automated signature recognition that would flag sites as directories, along with additional fields for manual tagging if noted as a link-selling site. There would be no need to check each new link against the global index to see if there is a reciprocal and it could affect many innocent sites, such as the ones mentioned in news stories in hundreds of cities. These would be picked up during a backlink update.
Only "respected" directories (long history, good reputation etc) might be exempt from such a filter.
Sites with a high directory:non-directory ratio could be regarded as candidates for filtering.
Similarly, an old site need not be more kosher than a new one. Some of the old but minor FFA sites might still be around. I haven't checked thoroughly, but the sites devoted to the recent tsunamis would be good ones to check because they would have a high proportion of reciprocal linking for the noblest of reasons.
| 12:27 am on Apr 2, 2005 (gmt 0)|
I thought that 100% recips may be a problem some time ago so when I put a link up and ask for a recip, if I dont get the recip, I leave the link up anyway. Also saves a load of time checking up on it all.
| 3:16 am on Apr 2, 2005 (gmt 0)|
As we know Yahoo keeps tracks of searches and publishes breakout or noteworthy increases and decreases weekly. We also know google tracks and publishes less data in a less transparent format in the zeitgeist.
If you wanted to detect unnatural link bloom, one way would be to look for an increase in links without an increase in search volume. The greater the difference the more spam like it looks. This would also help them know when to deliver fresher documents as opposed to stale ones.
<tin foil hat theory>You could break searches down into categories (think Dewey decimal like system) and establish more accurate baseline measurements for categories. Searches for people like Britney or Paris would naturally be spiky. Searches for euclidean geometry would be relatively flat</tin foil hat theory>
I think recips, or more specifically recip pages are going to be "so 5 minutes ago". For example if I put up some amazing content on my website and got a link from the New York Times, would I be devaluing it by putting up a link back to the Times website on my home page? Lets face it from an spiders point of view a recip page is pretty easy to pick out. The format and link to content ratio is probably fairly consistent across different sites on the web. However if the links were placed inside articles or even relevant blocks of copy it would look less like a reciprical link directory.
My two drachmas.
| 7:08 pm on Apr 10, 2005 (gmt 0)|
If someone wants to link to me in exchange for me linking to them because our traffic is complimentry and we can further each other's business interests, then I will do it everytime. And I could care less what the Google thought police think about it. I remember when reciprocal linking was done for marketing and not seo. Its an excellent business strategy.
And for the record, I SEO many sites that have aggressive reciprocal linking strategies and all enjoy top 10 listings for HIGHLY competitive keywords. Until someone can PROVE that reciprocal linking doesn't work, then I will continue to use it to get high rankings.
| 3:47 pm on Apr 11, 2005 (gmt 0)|
So what are some rough guidelines we can put together?
I have been sending out 10 link exchange requests for each one of my site's every day for a little while. What this usually equates to is 2-3 new link partners for every 10 sent out.
This site has been around since late February. What are your thoughts?
All the Best,
| 4:23 pm on Apr 11, 2005 (gmt 0)|
-some rough guidelines-
So now all you have to do is get your site listed in DMOZ, you get a very large spike of backlinks due to many others try to replicate the directory. Also get on the first page of MSN Search for your keyword, MSN Offers RSS feed for the top 10 sites returned. Get your site listed in Yahoo Directory and get another spike. So the competitor cant do anything bad to your site, but another search engine position can? The funny part is that all mention above newly created links will be a part of the page that has Ads by Goooooogle. Its like saying the only useful link to our searcher that contains good information is when we say it is, if you think otherwise pay us 35 cents per click and as long as you do it’ll stays useful.
| 6:53 pm on Apr 11, 2005 (gmt 0)|
Regarding a PEAK in backward links, I guess it is related to the datafeed sites which populated the results a couple of months ago.
If you build a site, lets sa with a datafeed from "the largest online bookstore", use mode rewrite and place the same page footer on each page of the site, you will end up with about 30.000 NEW pages from a NEW site spidered in less than a week.
So it seems to me that a high peak is this: several thousands of links from one site to another.
As far as I understand, this situatio is considered a peak.
Having said this, we can discuss whether you will receive a penalization or simply Google will ignore these links which is what I guess it happens.
| 7:28 pm on Apr 11, 2005 (gmt 0)|
It seems pretty simple to me.
Rate of acquisition, above a "natural" acquisition rate based on a bell curve would be a red flag. The acquisition rate could be tempered by spikes or growing traffic on that term. Its all speculation, but not that hard to comprehend. Jumping in links from 0 - 2000 in a week is not a natural thing, unless the content is so pandemically viral that it explodes.
Reciprocal links from "unrelated" topics are not as helpful as those from related topics. Too many from unrelated topics, become hurtful.
There are several "topical" linkfarms that still are very powerful at Y and MSN, but totally disregarded at G. Once again G is leading the pack in determining that not all links are created equal.
Lots of good arguments in this thread. I believe that at this stage, most of the Google algo is autogenerated by the algo itself - identifying and reporting statistical anomalies for human review, and incorporation into the algo.
As far as testing the algo's, the toolbar is a natural - on the fly.
IMHO, the goal of the algo is to build artificial intelligence that sees the value of a page the same as a human, its just that the computation is too large, and requires the algo to do most of the work itself, with feedback from the toolbar.
Content, old fashioned on page optimization, and and backlinks from sites that approve of your content is the way to go. The shortcuts (buying, linkfarms, random reciprocals) are all short term fixes, and just like every other "spammy' or blackhat tactic have short term effects, and will eventually be discounted, or penalized.
Long term and bulletproof, or short term and targetted for extinction.
IMO, SEO is more about compliance and good content building in service to the user than about discoving shortcuts and workarounds. Leave that to the fly-by-nighters.
| This 31 message thread spans 2 pages: 31 (  2 ) > > |