How will Google interpret this link?

[realmaverick]

I've been trying to figure out a way, to not leak link juice via function buttons i.e download, rate etc etc.

One of my competitors, has all their "function buttons" written like this:

<a rel="{url:'http://www.domain.com/ajax/ajax-lightbox.php?id=9&amp;item=TGFtYm9yZ2hpbmk=-2970316',size:{x:578,y:408},classOverlay:'',classWindow:'',closable:'1',closeBtn:'1',ajaxOptions:{method:'post'}}" id="970316" href="#" class="parseLightbox button assigned"></a>

They're placing an ajax link, inside of the rel attribute. Obviously attempting to not loose large amounts of link juice, because these kind of button appear a lot on both our websites.

Will the above work though?

Some of the competitors don't show these links, as they force signup. something I want to avoid. I'd really like to find a solution.

I realise I'm probably trying to attempt the impossible :o/

[tedster]

If there's an <a> element, then it takes a share of link equity along with the rest of the links on the same page. There's no way around that anymore.

[deadsea]

tedster is wrong. My testing shows that Google currently ignores <a href="#" onclick="fn()"> links and <span onclick="fn()"> psuedo links. They DO NOT count towards the number of links in the page in the denominator of the percentage of pagerank passed to other links.

However, you have links embedded in the rel attribute. Google is certainly smart enough to find things that look like urls in javascript and follow them. I haven't tested urls in rel attributes, but it would not surprise me if Google can discover them as well. I have not tested whether Google drops pagerank on the floor for urls it discovers that are not in anchor hrefs. It is possible they drop pr on the floor for every url they discover, regardless of whether it is an anchor href link.

You currently can get link from your pages without dropping pr on the floor by employing techniques like you describe. I have used javascript to scramble the text in an onclick. Me tests shows that google did not pass pr across these and did not drop pr on the floor.

Google may however update their algorithms at any time. In addition, it looks spammy. You are trying to hide something from the algorithm. I'm not confident that such tricks would pass a manual review by a Google engineer.

[tedster]

I'm willing to be wrong - my whole life is a series of lessons learned from being wrong. Can you share any details about how you tested for the denominator?

[deadsea]

I create hundreds of pages. Put a few sentences of unique content on each. Arrange them such a way that different ones will get different amounts of pagerank based on their link structure. I let them soak for a month or so and measure their pagerank. The best way to do this is count the number of times gbot crawls each page. Higher pr pages get crawled more often.

[tedster]

Sounds like a good testing process - thanks for bringing us up to speed on this.

Google may however update their algorithms at any time.

How true that is - the whole area of JavaScript links has evolved quite a bit in the past couple of years. Used to be they just weren't processed at all. Then around 2009, simple script and form links started to get different treatment and be added to the webgraph.

Your testing makes me wonder if common Lightbox pop-ups are part of the reason for Google not dropping the floor PR - or if it's just their unwillingness (currently) to parse all kinds of JavaScript functions.

[bumpski]

With Google preview they're throwing a full blown browser in the mix. It's likely they will be doing more with those preview results than just showing previews. Evaluating Javascript and CSS and the many tricks they both facilitate.

I'll bet deep down Matt Cutts would love to get rid of PageRank! Perhaps it was Democratic in the early web, now it's Aristocratic!

[mirrornl]

Thanks Deadsea!

[Sgt_Kickaxe]

You see adsense on those previews so the crawler is executing javascript for sure. I don't see why links would be ignored when functions aren't.