Link Exchanges And Google's Opinion, Are they worth doing anymore? Options

[Jonesy]

I've been reading a lot recently about Google's attitude to paid links and link exchanges: Google's opinion 1, Google's opinion 2

I am therefore debating whether or not to start my own link exchange program. I'm sure this must have been asked before, but has anyone got any positive proof of results on the SERPS from a link exchange program? By that I mean your site rose up the rankings through exchanging links. And can it be proven that this wasn't natural through age of site and other factors?

Or has anyone got proof of negative effects, barring the obvious penalties from being associated with link farms and black hat techniques etc.

I just can't make my mind up on this one without evidence!

This post has been edited by Jonesy: Jan 7 2008, 03:24 AM

[Autocrat]

I'm thinking it would be a little awkward... how do they identify a paid link, or a link swap?

What if the link is solely on a page for interesting links... along with 40+ others.... or offered as a directory of similar topics etc.???
Does this mean that Google is actually killing of what makes the web the web... by making links to other sites a no no ?


So I'm more than interested to know if this gets you penalised (either as having a link on another site, or from displaying links to other sites).

[Jonesy]

All I know is, following the last "Google Shuffle", there were some serious PR drops!

http://andybeard.eu/2007/10/pagerank-update.html

Google can easily spot a reciprocal link, but at what point does it become "excessive" and get penalised. I'm thinking if it's done on a small scale and the links help the user then it should be OK. It's the element of doubt that gets me though!

[fisicx]

From the google guidelines:

Don't participate in link schemes designed to increase your site's ranking or PageRank.

Which I think answers the question.

[Autocrat]

The question is, how do they make such a judgement?

Surely they would have to verify whether the two sites have common factors, shared interests etc.

Further, they would have to check the age of the sites and the links ... and whether they occured with a certain time of each other.

Then there is hte whole thing of being on the same server/ip etc... how to distinguish between the self-boosting group sites, and those that are legitimately linking to each other whilst on the same server?


It still strikes me that Google is crossing more than a few lines, and not really caring that they may be causing a fair bit of harm as well as some good.

[Jonesy]

OK then, let's take my own fledgling site as an example... We sell silver rings, so what's wrong with me putting together a page called "The Ring Box recommends" and linking to related sites such as jewellery cleaners, maybe perfume sites or other sites that are relevant to my main target audience (women)? Surely done on a small scale this is acceptable? Surely the sites linking back to me are also providing a benefit for their customers as they might be interested in what we have to offer?

It would seem logical that somewhere there is a threshold where Google decides whether or not your gaming the system.

I think I'm with Autocrat on this one?! I think the links should be sought out carefully and be relevant to the customer, and kept to a logical minimum.

This is very relevant, especially all the user comments: http://www.mattcutts.com/blog/seo-mistakes...xchange-emails/

This post has been edited by Jonesy: Jan 7 2008, 06:49 AM

[fisicx]

No problem with linking to related sites but just providing a list of links is of little value to your visitors and the search engines.

Far better to write a description of the target site. If you are recommending the site then tell us why. For example:

Silver rings, if left in a jewelry box, will become tarnished of time. The best method of restoring the surface is to use a polishing cloth. There are many types available but my favourite is the XXXX cloth available from most jewelers or online from xxx.com.

By writing a few words, the visitor understands the reason for the link and allows you the opportunity to throw in a few keywords.

If the target site decides to reciprocate by adding a link to your site then no harm is done. But you need to make the links organic, as if they occur naturally within the site. And if your aim is to provide useful information to your visitors then this will happen. But a page with 50 links to other sites that themselves have a page with the same 50 links will do nothing to assist your credibility.

It follows therefore that a link to a perfume site from a jewelry site is not organic and will no doubt be ignored by the indexing robots.

As with most things, a sensible and moderated approach will help rather than hinder. But as google says, don't do it to try and help your ranking, do it for your visitors.

[Jonesy]

Linking out is no problem, obviously, and I totally see where you're coming from and that's ideal, but... the chance of working with other site owners to get that kind of integrated link is nigh on impossible (from them to us I mean).

How many small eCommerce sites do you think can get that kind of link, embedded on the other site with all the relevant keywords?! I'm sure within the blogging community it's a lot easier.

What I'd probably end up with is a nice, organic link from my site to the other, and a simple link from a links page coming back from their site!

[fisicx]

And that's the rub. First thing you need is a site that is informative, interesting and somewhere people will want to visit. That will give you a fighting chance of building a few inbound links.

The big problem with ecommerce sites is that there are so many of them all competing for the same market. Unless you offer something different that makes you stand out from the crowd than now matter how many links you get you will always be catching up with long established sites.

If you work you way through the many threads on this and other forums you will see the same message: good SEO takes time, if you rush in with a quick solution then don't expect long term success.

Which brings us back to the start: make sure your site offers something special, run a low cost PPC campaign, write articles, start a blog, join in womens forums offering jewelry advice - make yourself known and in 6/12 months you should start to see some results. Anything sooner is a bonus.

[Autocrat]

Okay...

So basically, G sztarted rating sites, and one of the key approaches to this was how popular a site was based on links to it.
Then they introduced the idea of penalising some sites for taking advantage of this scoring method.
If they deem you as having Paid for Links, or Swapping Links, then you may suffer for it... whether you actually did it or not.
The only way to resolvce it is to appeal to them and attempt to convince them that it is an unjustified penalistation and htat it is a perfectly natural and legitimate set of links.

Other than that, we may end up having to set most links with the nofollow attribute jsut incase big G decides to stuff us over.
Now I know that is an extreme view... but what are the alternatives.
Even having a nice set of "organic" links could still get you into trouble, as it to my mind it would seem more like a bio was planned, and that some effort went into the link.
Link directories on the site would have to be avoided or use the nofollow, as a page full of external links would surely get attention! (Even though it is almsot as traditional as a HomePage).


Yet again G causes alot of issues without realy explaining anything.

[Jonesy]

fisicx - I know you're right. I understand it's got to be a long term strategy, but hardly anyone in eCommerce goes about it this way, which is worrying I suppose. I just needed to clarify myself which way to go as I want to do things "right", which in SEO terms 'aint easy.

Autocrat - your final quote I also agree with!

[bwelford]

The problem for Google and for us all is that payment comes in many forms. Google wants 'paid links' labeled so that their spiders can easily identify which should be ignored. However payment can be just that somewhere else you may write favorably about a website that has given you a link. It doesn't need to be cash.

Some cases are clear cut but many more are difficult to rule on. The simplicity that Google would like to have to make their algorithms work well just doesn't exist. It's a scale with all shades of gray. Luckily to be able to give 'relevant' results, Google can only take action on those at the extreme black end of the scale.

[Autocrat]

I suppose the only thing to do is suck it and see.

If you are linking to sites due to your opinion, then go for it.
If G decides to slap you down for it... contact them and ask them to sort it.
They are pretty quick in many cases (sometimes they may take a Month or two), but usually they seem to handle such an issue sharpish.

So I say go and try, and see what happens.

[Angela Charles]

You've all handled the "technical" discussion on this well, but I just have to add my two cents on the common sense of it all (or lack thereof).

Presumably, the complex algorithms are intended to deliver the best possible results without manipulation by sites that want top position on a specific keyword. However, I'd argue that Google has gotten far away from their goal of quality content. Too many babies are being thrown out with the bathwater. If a webmaster has worked hard to craft content that is targeted to the intended audience and then also seeks out as many links as he/she can to promote the site Internet-wide, why should they be penalized?

I've done many searches where the returned sites are pure junk ... is it because the really relevant sites were all dinged?

This seems to have more to do with Google promoting Adwords ... if you can't have any assurance that your search engine optimization strategy will position your site for success, then you'll probably resort to buying the position.

[projectphp]

Presumably, the complex algorithms are intended to deliver the best possible results without manipulation

No, that is the falacy of the extreme. The Japanese have it right here: they have words for "best" and "never", absolutes that should have meaning, but the rarely, if ever, use them. Why? Because never should have power. "I will never smoke again" should be a promise kept. Unfortuantely, English as a language has evolved to accept a very liberal intepretation of absolutes, and "best" rarely ever means the best bar none.

The sum total of many arguments in English is the use of a single word, like "wihtout", to create an reductio absurdum argument, i.e. to point out that the idea is absurd, and therefore wrong. Often, as in this case, if we change the sentence away from absolutes, e.g. to read "...the complex algorithms are intended to deliver the best possible results without undue and damaging manipulation...", the paradigm, and the expectation of perfection, evaporates, and the argument falls apart.

We are left, instead, to argue what "undue manuipulation" is, and that is something we can quantify and test for, and that is comparative. As an example, hidden text. If two sites out of all teh sites on teh internet use hidden text, that really isn't undue manipulation. But once the "tactic" becomes widespread, it does likely become an issue.

Similarly, Google can comapre two algorithms to each other, not to some arbitrary and impossible perfect algo, and choose to implement the better of the two.

Ultimately, that is what progress is: an incremental and measurable improvement over time, moving towards, but never reaching, perfection.

However, I'd argue that Google has gotten far away from their goal of quality content.

Oh, a straw man argument! Where did Google ever list this as a "goal"?

This seems to have more to do with Google promoting Adwords

What a leap! From the argument that "many searches" on Google return "pure junk" to a motive is a bit much for me, especially when "many searches" is such a nebulous concept. Is many 3? 50? 5,000? Are these searches common? Popular? Obscure? Should a search performed 100,000 times a day be deemed mopre important, in deciding upon algorithms, to one performed twice a month? These are teh issue shtat need to be addressed before leaping to conclusions, IMHO.

Just my $0.02, but I think too often people finda bad result, and want to claim something. People forget how terrible the Search world was back 5 years, and want to compare Google to perfection and, whgen they fall short, claim it is intentional ineptitude. I think that is terribly unfair. When you have thousands of people motivated to try to ruin your algo, motivated by $$, to think that the odd poor SERP is ineptitude on Google's part, rather than skill on the part of some $$ hungry SEO, is the wrong way round.

But back to paid links, people have the hierachy wrong. it goes olike this:
Links that add no value to the algorithm.
Of which paid links are a subset, and so are link exchanges.

If you change the focus from what constitutes "payment" to which links add value to an algorithm that uses them, then a lot of this hoohaa is kinda irrelevant.

[Autocrat]

...Angela Charles...
Not sure how you claim that some of that was missed, as I'm pretty sure we covered it above.
(I may be wrong, but I would say the above posts cover the trends, the technical and the general.)


...projectphp...
Damn fine catch and rebuttal.

The only issue I can say I have with it is that I believe Google has stated on several occassions that they intend to provide honest results, and that it is their objective (aim/goal?) to ensure that genuine sites of quality content are listed above those that are dubious.
(Maybe not those words, but that would seem to be the general sort of view.)

And sometimes they fail.
It is not the failing that causes issues towards google, but the simple fact that they seem to make leaps and bounds and perform small miracles, and appear to damage others in the process.

It is the fear of receiving such damage that started this topic, and many others like it.
Maybe the fear is misplaced, an over-reaction, an excessive worry over something that will not happen... but in general most of us do not know. All we hear is the little horror stories of apparently good sites getting penalised. Whether these are "really" good sites or not doesn't stop the creation of the mythological search-engine-bogoey man.

Add to that the various attempts of G to make money, blatantly moving away from the original type of service they provided, (the one that they said they would not do???), by cluttering the results page with adverts.
By making large steps in monetisation direction etc.

To me, more than understandable, they are a business - yet I'm sure they originally said they would not be like the others, they would not be playing favourites, would not permit funny results and would do their best to provide quality results...
LArge statements to make... and ones that are obviously comming back to haunt them.


Yes we are harsh on google... but when you go around making such claims, and then do things differently, you have opend yourself up to such criticism. It is also due to the fact that they are the biggest (or one of) that they get such attention.
The more fame/acclaim, the more likely you are to get big rocks lobbed at you for the smalelst of tresspasses and failures - it goes with the territory.

[JohnMu]

So you are saying, in other words, that since Google cannot catch every single case where someone games the system, they should let everyone game the system?

Somehow I don't think that would benefit our users .

John

[Autocrat]

I don't think any one said anything of the sort.
If it has come across that way, I'm pretty certain that was not the intended implication.

In reverse though, should they be using nets with ever smaller gaps adn tighter cords, running the risk of catching the proverbial dolphins?


It is not an easy thing, and the real danger lies in the large lack of actual knowledge.
Most things to do with SE's in general are theories and insights... not well known, established by the source facts.
Add to that the various little horror stories, adn it is no wonder that thigns get a little tangled.


Do you know of any "tue" reference that G clearly states how it handles links on sites?
How they will perceive links from A to B, and B to A?
'Cause I don't, nor do many others.

Without that information, we are left to assume and base thoughts on little more than heresey... and considering in many cases it can affect a persons lively hood, that could easily be perceived as a little unfair.

[Adrian]

False positives are a problem, and I'm sure Google don't want to penalise people doing good, honest work. But with the best will in the world, it's going to happen, and there's probably a compromise being made, how many spammy things can they catch, against the number of legit sites that might get caught up in it.

I expect the single task of working out whether 2 sites that link together are doing it naturally, or for PR reasons is quite complex in itself. And I wouldn't expect Google to tell us how they go about it. You know people would be analysing it immediately and working out how to make their links look 'natural'.

There could be factors like how many links go between the 2 sites, the link text, where on the page the links are, whether they link to the same pages, the age of the 2 ways links (if both sites link to each other at a very similar time....). Any/all of those things could be part of the equation, there may be other data that Google has access to, or ways of interpreting that data that makes certain links look more natural or not.

When you're surveying the internet, with several billion links, and you're trying to work out if any individual link is 'paid', it's likely that not only will you miss some that are paid, but will mark some as paid that aren't.

[projectphp]

where someone games the system, they should let everyone game the system?

"Let" is the wrong word. "Not care until it is widespread" is more accurate.

Wallmart used to call the police for every shoplifting offence. Steal a $0.20 battery, call the police. The theory was it would reduce the shoplifiting issue. The ultimate outcome? The police stopped coming to arrest offenders, because, come on, what does a court case cost? Police time. Court time. Court buildings etc etc. For $0.20? Reckon that was the outcome Wallmart were after?

The issue of spam is not absolute. 0% spam is not the goal. Useful SERPs are. If you quantify that, and continue to measure it, you can start to move from X% useful towards 100% useful, whilst accepting that getting there is nigh on impossible, the minutiae that many SEO bloggers live off is kinda lame and missing the point.

As an analogy, if X will only wear a perfect hat, and Y is just happy to block the sun, X is probably more likely to get sun burnt, whilst Y looks spiffy, and avoids some sun, in first her paper hat, and then her Old Sombraro that has the holes in it

When you're surveying the internet, with several billion links, and you're trying to work out if any individual link is 'paid', it's likely that not only will you miss some that are paid, but will mark some as paid that aren't.

That is binary Adrian, and binary elements make for poor algo factors, because they aren't terribly useful for anything other than breaking tiebreakers.

Think not in terms not of paid/not paid, but trust. Rather than on/off, imagine a scale, from 0 to X, of the trustworthiness of a link. If a site sells links, and also has some that are "legitimate", the legitimate links likely suffer from a "lie down with dogs, get fleas" penalty.

The issue likely isn't paid vs non-paid, but the trust Google can put in a link as an honest, independant, trustworthy and algo useful assessment of the link as a vote, or some combination of those elements. If we use that as the criteria, rather than a semantic discussion of what "paid" is, we can start to debate the relative merits of links in a way thaty is more acurate, long term, to the SEs goals.

As an example of the proposed scale, links like my link in the footer of this forum are likely less objective than a review in a newspaper, but more trustworthy and algo useful than a paid link on a spam site.

Scale, not binary, is the key, and the grey, far from being such furtile ground for SEO bloggers looking to make a name from some semantic example, should be embraced.