Link exchanging. Is it honest? If it is honest, is it worthwhile? If it is worthwhile,
how easy is it?
I notice a number of sites still offer to exchange links. You probably all know why webmasters do this, but let me just run through it again:
- It is good for your readers. When the subject matter of one site overlaps with the subject matter of another, you may be doing readers a favour by alerting them to the related content. A direct hyperlink is a more focused lead than the scattergun of a search. I remember hobbyists starting 'web rings' years ago, back when Yahoo and AltaVista ruled the search world with their brute-force crawling, and Google was just a thought in a Ph D thesis.
- It is good for the site operator. The more sites that link to you, the more traffic you will get.
- It is good again, exponentially good, for the site operator. Because the number of sites that link to your site is also a big factor in the ranking calculations of the search engines.
The principal is this. You can measure the authority of, say, the BBC in the UK by the extent to which others use, respect, or quote them. It is an indication of the authority they have already established. Translate this into the online world, then, and you can measure the authority of a site by the number of other sites that refer to them. Therefore, the search engines rank sites that most people refer to at the top of their lists.
Now I would in no way wish to endorse anything that tried to fool a search engine, or mislead a reader in any way. And some of this is becoming myth. But provided the mutual links really do help the reader, it may still be worth joining the link exchange party.
We sat down the other day and looked at how to set up a link exchange page for one of our sites. The first move was to review how other people are doing it.
"Link to us, make the link a permanent one, then contact us to say where in your site you have linked, then we will link to you".
That is a common theme we have seen on many sites. We will only link to you if you first link to us. Show us the evidence.
Fair point.
Which got me thinking. I could automate that. I could write a 'sites that link to us' page that reads the sites off a database.
Step one: who is linking to us anyway?
One way to find out is in your server-side code. Every http request has a piece of information called the 'referrer'. This is the URL of the page that the the user jumped from to reach yours.
Another way is in your server logs. The site I was working with used the 'Webtrends' package. (Which I would recommend).Among other things, this product produces some very useful reports from your server logfiles. Which just happen to include the referrers. Webtrends for our client site showed hundreds of referrers over the course of a few days. Mainly, it must be said, various flavours of the search engines.
Another way is through the search engines themselves. Try the 'advanced search' option in Google, where they will explain to you how to look for sites that link to yours.
Step two: capture the links as they happen
Every time a page is called, you can inspect the 'referrer' information of the http request, and, if it is one you had not yet been aware of, you can store the information for future inclusion in your 'sites that link to us' page.
Step three: respect the wishes of the other site operator
Do they really want to be part of a link exchange scheme? What, for example, if the other party is simply a search engine themselves? You will hardly be doing your readers a service if you just list the search engines that have crawled your site. They probably only reached you because of some robot activity, not humans following a link. You will hardly help the "linking means authority" guideline if you generate an insane escalation of automated links to sites that themselves only linked to you through some automated process of their own.
The way to do this is to inspect the 'robots.txt' file that sites optionally maintain at their root level. It is primarily intended for directing search engines about whether or how you want them to crawl your site. But it will do just as well for our purposes: if a search engine do not want their own links to us registered, then they should say so in their own robots.txt file. If they have none, or if what it says welcomes you to scan their site, then take that as an invitation to include them in your 'sites that link to us' page.
Final step: put it all into code.
This pseudo-code should do the trick. Implement it for your home page, the pages that matter to you, or even as an include in every single page in your site:
pseudo-code for any web page on the site:
what was the page that referred to this one?
if none was mentioned, this page access was a direct one
lets just look at the domain, rather than the page, to keep things brief
lets check it was not a cross-reference from our own site
lets check it was not a domain we already know about
lets read their robots.txt file to see if they want us to record them
they have no such file? or they have one and it les us in?
lets record the other site in our database
pseudo-code for the 'sites that link to us' page:
mention a caveat, such as 'feel free to exchange links with us. Simply link to us from one of your pages and our web-scanner will be happy to return the compliment. There is no need for you to take any further action. Please allow at least 1 nano-second for us to scan the entire internet after you have have tested your link to us, before we update this page'
read the sites in your database and list them all.
Thats it. A little extra date stamping to ensure you purge old links and do not build up too long a list, and you have a production-ready industrial-strength 'sites that link to us' page. That takes no management. Content without effort. Visitors for free.
