So what's a relevant link?

[ron_ron]

Let say for example you are in the air ticket reservations business. More often than not, someone who needs an air ticket also needs a hotel. So in the real world, an air ticket reservation site has a bona fide reason for exchanging links with a hotel booking site. But how would G look at it?

If your only concern is G, the Lord of the Internet, should a hotel site only link with other hotel sites in order to have focused keywords? Or, is G smart enough to figure out that air tickets and hotel reservations are related?

Lets look at it another way. Let say you have a hotel booking site that specializes in hotels in the Toronto area. Assume that you are trying to rank for Toronto Hotels. Does that mean that you would be better off only exchanging links with hotel sites whose websites had 'Toronto Hotels' as their main keyword density?

Now you could concentrate on that, trying to get as focused as possible. But you would probably have to settle for some pretty mediocre sites in order to do so. If you expand your universe to include site that have keyword rich text like New York Hotels, Paris Hotels, etc you could have a better chance of getting better quality links. And you could get many more links too. You could even expand your universe to the entire travel industry spectrum. You could have stuff like travel to China, safaris, sightseeing in Rome, etc. But the keywords aren't as focused. So where do you draw the line?

Of course you could have those sites linking to you use anchor tags like Toronto Hotels and Hotels in Toronto, etc. but the pages that your inbound links would be on would usually not have that sort of keywords as it main keywords.

DISCLAIMER: I am not in the hotel business. :)

[piatkow]

Work on outbounds that will add value for your visitors and inbounds where your site will add value to the other site. For keywords concentrate on your own content.

[julinho]

is G smart enough to figure out that air tickets and hotel reservations are related?

I think Google have many sources of data to make this correlation.

First, the web structure. There are probably many hotel pages linking to air tickets pages, which creates a pattern (and analysing backlinks to these pages, the pattern becomes much more clear).

Next, everytime people make a search, they tell Google that there is a relation. Several people search for [tickets to Paris] and then [hotels in Paris] in a short time frame; Google are certainly taking notes.

And then, user behavior tells Google even more. People visit a hotels site, click to a air ticket site, and then browse several pages, stay long, bookmark, etc.

Now, give this load of information, accumulated over years, to the Army of statisticians which Google certainly have on staff, and they will probably find patterns to tell the relevant links from the irrelevant ones.

Compare this to, e.g., footer links in a jokes site pointing to a mortgage site.
Few links exist on the web between jokes and mortgage sites (most are paid ones); few people click on such links; those who click, probably won't stay long; and searches for [jokes] and [mortgages] are in different circles.

Google can certainly tell the difference between these cases.

[cnvi]

Don't overthink it.. any link that benefits your end user's experience is a relevant link. Don't worry about having a few irrelevant links as most sites have them. Just make sure to avoid schemes that generate a high volume of irrelevant links in a short period of time and you will be fine. Ask yourself, "does this site I am about to link with benefit my end user?" If the answer is yes, get the link always regardless of pagerank or other metrics. If you aren't sure, skip it and move on to the next opportunity.

[joost]

I absolutely agree with cnvi. Let me elaborate a little:

First: what Google finds relevant is not the main question. The main question is relevance for the intelligent human being. For a visitor looking to book a hotel in Toronto, links to Paris reservation sites are completely irrelevant, let alone “travel to China, safaris, sightseeing in Rome, etc”. If Google finds relevance now because both "Toronto Hotels" and "Paris Hotels" have the word "Hotels" in them, it should (and probably will) in some future find a way to correct that error, and all your efforts are in vain.

Second: the word “Toronto” is just as relevant to Google as the word “Hotel”. And “Toronto” sure is far more relevant to your visitors. So why not exchange links with Toronto related sites. Make a directory about businesses in Toronto, restaurants, bars, museums, anything of interest to your visitor. These businesses would be all to happy exchanging links with you, because your hotel guests may want to visit their bar (etc).

Third: Sure Google is “Lord of the Internet” now, but gathering quality links for long-term rankings is a time-consuming effort. Even Google can go down and loose to a party that serves more relevant search results. If your guideline has always been relevant for the visitor, you win, if your guideline has been what is relevant to Google, you loose.

[wheel]

cnvi is of course, correct.

There's two primary ways to define relevancy of a link pointing at your site. First (not neccessarily in this order) is the content of the page or site that is linking to you. If I had to guess, I'd say that page is the important one.

Secondly is their backlink profile. If their backlinks are relevant - however that's defined- then a link to you should also be relevant. I don't know if Google measures this.

In the end, this ties in with what appears to be an outdated concept of neighborhoods. Go get links from where your competitors get links. I no longer subscribe to this notion.

My backlink profile is highly relevant but does not share the same profile as my competitors. I try and get backlinks from high end authority sites in my niche, ones that others don't have links from. I also work on niches that are very relevant to mine and touch my niche - but again are not areas that anyone else in my niche seems to use. For example, there are technical experts who design the product I offer. I target those technical experts including their licensing/authorization/gov't orgs even though they're manufacturing and I'm retail. Like being a PC retailer and getting links from Intel. Can google figure this? I dunno, doesn't matter. If not now, they'll get there.

[ron_ron]

Thank you all for your input. If you don't mind, I would like to ask another question that I know must have been asked and discussed a million times. There was a time when a lot of people started saying that a reciprocal link was worthless or at most, not worth much if anything.

About three years ago I slapped together an average site and created a link directory. I hired someone to do the link building. In my opinion it was sloppy work. If I would have done the work myself I would have been much more selective. There were many junk sites and although it still had some relevancy it was very broad.

In disgust, I just sort of let it ride. I did not even bother going through the link pages and deleting the garbage. Then over a period of a year, two, three I noticed it was on first page results. This sort of led me to believe that exchanging links, even through link directories, is not as bad as many were saying.

But IMO, it is better not to have a linking structure that is 100% reciprocal. There seems to be some webmasters (probably not many on this board) that think for everything you give, you must receive something equally in return or you are selling yourself short. I don't think you would be draining your PR (as some may think) if you outbound 90 links of which only 30 are reciprocal. Naturally I would try to get some one-way in bounds to balance it off.

[joost]

Sure link exchange works. Maybe not in the way it used to, but still. Incoming links are incoming links, any which way you look at it.

Whether or not they will rank your site on the first page depends on what your competitors do. If they have thousands of relevant, high quality one-way inbound links, your link exchange directory approach may not be sufficient.

In your case, since your site has first page rankings, my advice is not to make radical changes. If it isn’t broken, don’t fix it, is my idea. Maybe clean up outgoing links to crappy sites, get a few new one-way inbound links, and maybe you can enjoy a few more years of first page ranking.

Maybe have a look at the inbound link patterns of your competitors. If those are way better then yours, prepare for future adjustments to prevent deteriorating rankings. But just don’t get carried away.

[cnvi]

There was a time when a lot of people started saying that a reciprocal link was worthless or at most, not worth much if anything.

Reciprocal links being worthless is an old SEO myth. The engineers at the search engines realize many sites will not give a link without getting a link back. If you do your research you will see that search engine representatives such as Matt Cutts at Google have never said "do not reciprocal link". They have only stated to avoid high volume irrelevant reciprocals.

Link exchange is alive and well especially among niche markets such as the wedding industry. It's perfectly fine for the wedding photographers to link exchange with the wedding florist. It's better if those links are locale specific.

The links have value. They are all spidered by the search engines although the engines might not show a linkback in the index. Just because its not in the index doesnt mean it doesnt have value. Also keep in mind that link exchanges will generate traffic from the links themsevles.. and link exchanges are a branding function - they help spread the name of your website among niche markets.

Think of link exchange :

1st as a link building function
2nd as a branding function
3rd and lastly as an SEO function. Don't make linking decisions primarily on how you think its going to affect your rankings.

There is alot of misinformation on the web regarding link exchange mostly due to fear uncertainty and doubt (FUD) from so called "experts" that would rather have you pay them big $$$ for "seo".

Link exchange is as old as the web and will ALWAYS exist. Just make sure to use editorial discretion on making these links and do it slowly in a natural voume. That means avoid high volume duples services. If you use software to manage linking and publishing of your links (which is ok), just make sure its EDITOR BASED so you maintain full control over who you link back to.

[ron_ron]

OK, I think I am getting a clearer picture of this now.

As to software tools to help manage linking, I do not like to use software that makes use of the engines out of fear. There is one tool that works as a robot (I guess it is better not to mention names here) and totally independent of the engines. You enter a few sites to visit and then it follows the outbound links all over the place.

I decided to give it a try as I was hoping it could zero in on relevant links better or faster than I could. After the tool found a bunch of possibilities I started viewing at possibility in its special browser. Then I got a virus and ended up having to restore my entire disk from backup. I was lucky I was making daily backups.

I normally do most of my work in Mac and have never had a virus in 15 years as a Mac user. I do have a windows machine in case I need to run windows-only software such as this tool. But I am kind of afraid to test this thing further.

What I would like is a software that finds pages with outbound links (potential link partner) and which looks at keyword density at the same time. If the software I mentioned above will not do that then I guess I would do just as well without it. I do not need a software to create a link directory. But I would not mind a tool that would help find me good prospects.

[EdwardAdventures]

the links from the site which has similar content and keywords are relevant links.

[walkman]

Look at at similar sites to get an idea.

[wheel]

What I would like is a software that finds pages with outbound links (potential link partner) and which looks at keyword density at the same time. If the software I mentioned above will not do that then I guess I would do just as well without it. I do not need a software to create a link directory. But I would not mind a tool that would help find me good prospects.

Really the only tools I use are Google to find sites, and Yahoo to check the backlinks of sites.

Pretty much anything else that's automated is going to be either pushing the boundaries of spammy, or is going to be very very high end and you'll need to know the right SEO company to get at the tool.

There's a variety of plugins you can use to automated the Google search or the Yahoo backlinks check, but in the end, that's about most of what you need.

[ron_ron]

Wheel, can you tell me more about what these plugins do and where I can get them? But do they help the engines to track you and build a profile on you? I am afraid to use even the G toolbar out of fear that Big Brother is watching how I go about getting links, etc.

[wheel]

I use a plugin called searchsite. There are others, probably better. I don't think they make any difference in Google's tracking ability.

The plugin doesn't automate anything other than things like doing whois lookups, the Yahoo backlink command, stuff like that.

[austtr]

Relevance is such a wide open term. It means different things to different people... and algos.

The human mind probably thinks it's relevant for that Toronto hotel site to link with Toronto apartments, Totonto car hire, Toronto tours, Toronto flights, Toronto cruises etc... because we understand they are all part of a broader "travel and tourism" marketplace and may add something to your viewers experience.

But there isn't any relevance associated with Toronto Funeral Homes, Toronto auto parts, Toronto body shops, Toronto lingerie shops etc. Why would the word "Toronto" get you any brownie points for links that obviously add nothing to the viewer experience? Ditto for those Paris hotel sites... Paris is not relevant to Toronto even if both sites are about hotels.

It would be a pretty poor state of affairs if the algo can't make those sorts of distinctions and I'm going with the arguments that says they can.

One other aspect of linking that does not get a lot of air-play in here is having lots of links vs having fewer links.

For years I have followed the mantra of "quality is best" and have collected a sold base of relevant, on-topic links with a good proportion having decent PR. My links total in the hundreds. The sites that put me in the shade have thousands of links, mostly PR0, many off-topic and with very heavy cross-linking between sister sites.

That is not a "woe is me" whinge, but rather the sort of thing that makes me think link quantity, irrespective of "relevance", plays a much bigger part in the very competitive niches than we openly acknowledge in here.

I'm not convinced that Google is as good as it thinks it is in determining link relevance, or quality, and when it has to rank two similar sites, the one with the most links seems to have a head start on the higher ranking.

[onlineleben]

I wouldn't worry too much about relevancy. keep it in the broader niche and you keep everyone happy (visitors and SE).
For the ticket site I would recommend to create a booking widget that other sites, preferably hotel sites, can install.
Hotel sites can prepopulate the above mentioned widget with their city name and if possible even with a city name of the visitor (geo targeting). By this the visitor who looks for a hotel room also can find a very relevant page about traveling to that city.
An other advantage: you only have to maintain one page with the widget and have not toplaster 100s of links on your site like 'Flights to Toronto from Rome' or 'Flights to Toronto from Madrid'

Keep it relevant for the visitor!

[namco]

I suppose you could use Google's Wonder Wheel for ideas. When I put in "Toronto Hotels" I got a bunch of other hotels and "sheraton toronto airport". So clicking on that I got a couple of other hotels, a conference center, sheraton downtown, sheraton tripadvisor and moxie's classic grill.

It's probably not the best suggestion but worth a look.

[joost]

I'm not convinced that Google is as good as it thinks it is in determining link relevance, or quality, and when it has to rank two similar sites, the one with the most links seems to have a head start on the higher ranking.

I would rephrase that as: not as good as it (or its competitors) once will be.

The human mind probably thinks it's relevant for that Toronto hotel site to link with Toronto apartments, Totonto car hire, Toronto tours, Toronto flights, Toronto cruises etc... because we understand they are all part of a broader "travel and tourism" marketplace and may add something to your viewers experience.

I am convinced search engines will understand that also one day.

For years I have followed the mantra of "quality is best" and have collected a sold base of relevant, on-topic links with a good proportion having decent PR. My links total in the hundreds. The sites that put me in the shade have thousands of links, mostly PR0, many off-topic and with very heavy cross-linking between sister sites.

Yours is the long-term strategy that will win in the end. It is a lot easier for you to collect some additional low-quality links to rank now, then it will be for your competitors to gain quality links once the algo gets better.

Also: the ranking power of low quality links seems to expire faster. One of my sites ranks 1 for its main keyword set, and top 5 for about 20 other important ones. This has not changed for almost 10 years. Every now and then I see a competitor hire an SEO firm, and lots of low-quality links push their site close to mine (or past it) in the rankings. And then I see them decent again after a few months, a year, without me doing extra link building.

[austtr]

Yours is the long-term strategy that will win in the end.

I've been reading that for a very long time and despite all the experts proclaiming this as a proven fact, I get the feeling that I will be very old, very grey, incontinent and dribble a lot before it actually comes to pass.

Even if links are seen to be of little value (or have little if any relevance) I suspect the accumulative effect of thousands of them may be pushing sites to higher rankings than sites which only focus on the owners interpretation of quality and relevance.

To return to my original observation... it's very, very hard to know what an algo deems to be relevant and if it applies the same logic all the time across all all data sets and languages.

[joost]

I get the feeling that I will be very old, very grey, incontinent and dribble a lot before it actually comes to pass.

I certainly hope not ;-)

My strategy is NOT “or low volume high quality links or high volume low quality links”, but (if need be) “and low volume high quality links and high volume low quality links”, with the high quality links as an assurance for future ranking. You forgot to quote the second part of that sentence:

It is a lot easier for you to collect some additional low-quality links to rank now, then it will be for your competitors to gain quality links once the algo gets better.

In other words: if your competitors outrank you through mass directory, rss, article, and social bookmarking submissions, blog comments, profile links, link wheels, or whatever mass link strategy popular at that particular time, go ahead and do that also, but just don’t forget your quality links.

With:

It is a lot easier for you to collect some additional low-quality links to rank now

I meant there are hundreds of cheap services you can outsource this type of link building to, and dozens of cheap software solutions.

Quality links are always a winning strategy, but everything is relative: in some fields you can outrank your competitors with quality links only, in others, more competitive ones, you just can’t. This also depends of the quality of your site: does is deserve to have the sufficient amount of quality links to outrank your competitors? If this is not so, it will never rank in very competitive fields if you don’t play the game just like your competitors do.

In competitive fields there’s a permanent race where certain webmasters keep inventing strategies to manipulate the search engines, and the SE ‘s eventually catch up and devaluate that strategy.

If you ONLY play it just like them (and forget to build your quality links) your ranking will tank just like theirs once the search engines catch up.

And then, of course, there are the other ranking factors, such as domain name (very often underestimated), on-page seo, the search engine’s preference for “brand”-sites, etcetera.

[vijayseo]

Links which are associated with your web site industrye.g if you have web site design then you should make back link from web site design industry website

[ron_ron]

What about this situation? A lot of times I find a site that I think is a good site, relevant, and not too many outbounds. But then I look at the page my link would go on and it has as its title junk like "Link Exchange." If it is an otherwise good match, should I go for it or stay away from it.

I know there are a lot of junk sites that have a link page or directory that is titled "Link Exchange" or Reciprocal Link Directory" or something like that. But there are a lot of good sites that do that too. Some of these webmasters may do this just because other sites do it and not even have in mind to "trick" search engines.

Would you link with them?

[wheel]

Nope.

A lot of these pages seem to be greybarred. Now that doesn't necessarily mean anything other than Google knows how to track these pages - they're aware of them and what's going on. Maybe it means nothing.