Tue 28 Oct 2008
Internal Links – Only The First Link Counts in Google?
Blurb by Shaun Anderson (Hobo)
I thought I would share the results of another simple test I did to see how Google treats internal links.
What does Google count, when it finds two links on the same page going to the same internal destination page.
I surmised:
- Google might count one link, the first it finds as it indexes a page
- Google might count them all (I think unlikely)
- Google might count perhaps 55 characters of ALL of the available links (could be useful)
OK – From this test, and the results on this site anyways, testing links internal to this site, it seems Google only counted the first link when it came to ranking the target page.
In much the same method as my recent seo test where I tested how many words you should put in a link, I relied on the “These terms only appear in links pointing to this page” (when you click on the cache) that Google helpfully shows when the word isn’t on the page.
Again, I pointed 2 everyday words at a page that don’t appear on the page or in links to the page, and searched for the page in Google using a term I knew it would rank high for (Shaun Anderson) and added my modifier keywords. I left it for quite some time, and checked every now and again the results.

Searching for “shaun anderson” + “Keyword 1″ returned the page (cache shown above).

Searching for the term “shaun anderson” + “keyword 2″ did not return the page at all, only the page with the actual link on it, further down the SERPS.


It’s not exactly Google terrorism to identify this, so here is the actual test page where you can see the simple test in action.
So today
 on this site
in internal links
, Google only counted the first link as far as anchor text transfer is concerned
How you can use to your advantage?
- Perhaps, you could place your navigation below your text
- This lets you vary the anchor text to important internal pages on your site, within the text content, instead of ramming down Google’s throat one anchor text link (usually high in the navigation)
- Varying anchor text naturally optimises to an extent the page for long tail ‘human’ searches you might overlook when writing the actual target page text
- Of course, I assume links within text surrounded by text are more important than links in navigation menus
- It makes use of your internal links to to rank a page for more terms, especially useful if you link to your important pages often, and don’t have a lot of incoming natural links to achieve a similar benefit

Credit – Graphic first sourced at Search Engine Land and created by Elliance, an eMarketing firm.
Works for me anyways, when I’m building new sites, especially useful on longtail searches, and there’s plenty of editorial content being added to the site for me to link to a few sales pages.
Note: I would think Google would analyse everything it finds, so it would find it easy to spot spammy techniques we’ve all seen on sites trying to force Google to take multiple link anchor text to one page.
What do you think?
If you like this test, you might like;
- Limit Anchor Text Links To 55 Characters In Length?
- Will Google Rank Pages Better With Valid Code?
- How Many Words Will Google Count In The Title Tag?
- A Google Friendly Website Navigation System
- Do It Yourself Search Engine Optimisation
Did you know when you link to a Hobo SEO post we have search engine friendly links back to your site if approved? Our comments are also search engine friendly you know (once you've commented on a few posts)! Do you need any more encouragement to get involved in the conversation ;)

Didn’t even think of this…..
Good work sir! I hate having to re-test basic principles stuff like this once every few months to make sure it still applies, so thanks for saving me a job
Thanks Daniel
I love testing things, especially on-site, though I am a firm believer Google changes the way it works, and wants to improve everything it does.
Its nice to find the limits where you can, and work out how you can use them to your advantage.
- Enjoying working for Big Dave?
Hi Shaun
“not exactly Google terrorism”
..remember one man’s terrorist is another man’s freedom fighter.
you’ll have seo science committee over here demanding control groups soon I expect but looks fair enough to me.
gee..! I never knew it before mate. you sure a seo scientist.
I think google takes the link that describes in a better way the webpage, for example if you have the same 2 links but one of them has an extra keyword
Keyword Shaun Anderson
Keyword SEO Shaun Anderson
google will take the 2nd even if it is down the other.
I am not sure you were entitled to check by the rankings, since G started giving less weight to pages that don’t have keywords in the body, but using the “These words only appear..” trick is probably right.
You might want to mention that you and your readers might want to use a good alt tag and title tag on the logo image, because that’s where all the pages link to the homepage and it probably is at the top of the code, anyway. If it’s not, good for them, but it’d help, anyway.
Or better yet, like you said, move the header to the bottom and link to the homepage with keyword text.
Also, you might want to mention that the first link test was first (that I know of) run by Michael VanDeMar (read the comments, too):
Cheers.
THanks for the comments Yura – I discuss Michael’s tests a bit more here in an article entitled Does Only The First Link Count
Google pays attention to the wording of the 1st link on page if link mentioned more than once?
Awesome fact to ponder. Thanks for sharing.
Wow if what u tested is true are we wasting our time making multiple links on a page!?
Interesting article Shaun. My question would be:
Does Google penalize you for having two links in the body text to the same page, or lessen the importance of the one link they do count?
Normally when I add two links on one page to the same destination its to help users navigate and not for SEO.
I don’t think Google would penalise you for that. It’s obvious and excessive manipulation Google would penalise – not a linking practice that could occur very naturally.
Thanks or the comments
…it’s good to see you put it into practice too!
Hey Shaun, your site is broken in FF 2.0 – I can’t read it.
Cheers Tad – I will have the boys have a look – must have screwed something up
I think another issue to consider is how many backlinks are considered by google when one website has multiple links to a particular website. This may even be links to different pages from different pages. Lets say we have a link back to our website from the footer of all pages on a a site we have designed. Surely it will not be counted as 50 back-links but only as one.
This gets more interesting ib the case or press relase or article websites where we may actuall y submit many postings over a period of time. If it will only ever be counted as a single link back, would’nt it make more sense to try and target as many websites as possible as opposed to focusing only on the major ones.
We’ve done some basic testing of this and the results we’ve seen do seem to corroborate your post – thanks for heads up!
Take the domain http://www.domain.com.
If I type link:www.domain.com into Google, why are sometimes 90% of the results internal links like http://www.domain.com/directory/page1.htm?
Surely this internal linking cannot have any impact in the rankings? Or does it?
My site is PHP based and many of the links are relative and coded without the full domain name, and dont seem to get counted!
Any tips or insight would be very very appreciated
Links on your site do help spread Pagerank and help your pages get into Google’s main index, so link to your important pages often. It helps.
Use Yahoo as well for link:www.domain.com searches – it’s better and more comprehensive.
More fuel for the fire Shaun, Cheers.
Quick question; Can Google read text contained in tables. I have some appropriate text that would make ideal anchor text links to another page.
hey Shaun,
I like your blog, and I like your style. I’d much rather read the advice of someone who can demonstrate that what they say may be true (with experiments) as opposed to most of the “experts” that just make pronouncements.
I read about this “first link” thing a few months ago, and know there’s a lot of disagreement. I did my own experiment, and agree with you.
Now if you have time, can you do an experiment to answer comment #17?
~ Steve, Pinnacle Trade Show Booths, aka purveyor of trade show booths
Shaun, I could understand the point of long tail keywords but could not really digest “# Perhaps, you could place your navigation below your text” – how that could be effective advantage?
Adding myself to receive follow up comment via email so I could get to know your thoughts on it.