02 Mar 2009, Posted by admin , 19 Comments

TinyURL Does Not Pass Value

Update: After almost 4 months, the TinyURL link has now started to pass anchor text value. As it’s unlikely that it would take Google 4 months to follow the TinyURL link, this suggests that Google has now started to follow TinyURL links and pass value as it should. HOORAY. I still think it’s best to use your own URL shortener though.

A while ago Rand Fishkin asked if URL shorteners such as TinyURL passed value. I started testing it when I read his post, and the results are pretty clear that, TinyURL at least, does not. At least, not anchor text value.

I put a link up on a test site with a unique word linking to another page on a different domain. The same page had another unique word linking to a different page on a different domain, this time going through a TinyURL. The target page ranks for the non-TinyURL link as expected. The other target page (linked with TinyURL) does not rank for the unique word. I gave it about a month to make sure, because it does tend to take Google around a week or so to allow for 301 redirects to pass value, as they catch up.

Why should TinyURL pass value?

Because it’s a straight up 301 to the target page. It should work. In theory.

Why doesn’t TinyURL pass value?

It’s hard to say whether this is an algorithmic or whether Google has manually chosen to not follow TinyURL links. If it’s part of the algorithm, then it could be because TinyURL has tripped a filter with Google. It’s commonly accepted (although hard to definitely prove) that having thousands of different domains 301 to yours can result in your domain getting a penalty, or a ban. It’s hard to say if that’s true, but I’ve seen and heard about it from personal experience. It probably just receives a manual review, but if it’s spam then it’s going down. Either way, a filter gets tripped.

The nature of TinyURL may be causing the problem. The domain itself has millions of seperate pages that all 301 to different domains. To Google, that’s a massive signal. It’s huge. No other kind of site acts like that. This may trip a filter, causing the link value (sharklarking) passing through TinyURL links to be reduced to 0.

Why isn’t Google correcting this?

Because seriously, why aren’t they? I used TinyURL, instead of the other URL shorteners, because TinyURL is used by Twitter. I know Twitter nofollows links anyway, but Twitter feeds appear in loads of different places in dofollow form. This means that hundreds of thousands of links that people are talking about and, in effect, voting for aren’t being counted. Why not? I thought that was the point of the democratic web?

15 Feb 2009, Posted by admin , 0 Comments

Was Google Right To Penalize Google?

In case you didn’t know, Google Japan recently used a Japanese based Pay-Per-Post system by a company called CyberBuzz to help launch a new set of Google features.

It looks like Google Japan (and Cyberbuzz) aimed to get bloggers talking about Google’s features, to raise awareness and to get a bit of a discussion going. They wanted it to go viral. Techcrunch first broke the story and news spread around the SEO community rapidly. Matt Cutts then went on to announce Google Japan getting a penalty, getting it’s toolbar Page Rank knocked down from 9 to 5.

So, Google was seen as buying links and got a penalty for it. This is Google’s way of saying that even they aren’t immune to their own rules and regulations. It fits in well with their ‘Don’t Be Evil’ thing, at first glance.

But Google Japan weren’t doing it for the dofollow links – they were doing it to spread awareness. It’s advertising. It’s like buying banner ads. If the site doing it wasn’t Google, if they were just a regular company that doesn’t know anything about Google’s link penalty system (and it looks like Google Japan’s marketing department fits that bill) and they just want to advertise their new product, shouldn’t they be able to? If it’s this easy for Google to accidentally buy links and get a penalty, how easy is it going to be for other sites that don’t know anything about SEO?

09 Feb 2009, Posted by admin , 0 Comments

If You Were Google

If you were Google, what would you do to keep your index relevant, authoritative, fresh, spam-free and useful? How would you build your algorithm to make sure that only honest, natural and editorially given links were counted?

Would you follow Wikipedia links?

Would you allow sites to lose link authority by linking out, or would you reward them for it?

Would you allow directory links to count in your algorithm?

Would you allow blog comments to pass link authority, or would you try to discount them? If so, why? If not, why not?

Would you let links from press releases count? What about Twitter? What about Gtalk, and Gmail?

Would you allow internal footer links to pass value, or would you try to only count regular navigational links? Why would you discount footer links? Why would you not?

If a site had a huge number of inlinks from the same IP range, would you let all of those links count?

Would it matter to you if an inlink was surrounded by related text, or would you count links that were on their own?

Would you allow links from unrelated pages to pass value? Why? Why not?

Next time you find out about a technique working, or not working, ask yourself why Google would allow it to happen. Always, always think like Google.

27 Jan 2009, Posted by admin , 2 Comments

Hate Blackhat, Love Google

Blackhat vs Google

I’m just sayin’.