Calendar
Network
Feeds
- 48
- Sphinn It!
Went Hot: March 14, 2008 - 7:05 am
Posted By: PPCblogger 209 days ago
Topic Type: News Story (Jump to http://www.davidnaylor.co.uk)
Category: Google SEO
a) Increased ranking on short tail and killed a lot of long tail traffic .. Hmmm not good
b) Didn’t effect ranking’s at all .. Ok so my internal page rank may have lifted a little here
c) Decreased rankings in general .. NOT COOL !
d) Slightly increased traffic over all. Way COOL !
11 Comments
Comments
Good quote from Matt Cutts too - thanks for publishing that Dave.
great post by daven :)
and that's just not the DaveN FanGirl Speaking!
Not buying all this nofollow fuss. Will test it on my own, like everyone else should and go from there. But I am not going to jump on the badnwagon and say it is the greatest thing since sliced bread when it clearly is not.
I've been using nofollow on internal links for quite a while now, and I've blogged and written about it in detail.Using nofollow won't add more incoming links, and it's not a magic silver bullet.
If you think there's going to be some kind of massive PageRank boost on your "money pages" that will send the rankings upward, you only need to do the math to see that just isn't so. Any stories about adding nofollow and getting an instant traffic boost are dubious, because it's not a controlled test.
What "sculpting" (I love it when SEOs invent new keywords to rank for) can help with is indexing, and we've seen this repeated many times. If you don't have a problem getting all of your (unique) pages indexed, and you aren't comfortable doing the math, then steer clear of this and just go get more links.
If you are having trouble getting pages indexed, and it's not something stupid like duplicate or empty pages (write some words, it's not that hard), then nofollow can help.
Agreed incrediblehelp.
While I think there is some relative benefit to sculpting in certain areas, I'm already seeing SEOs suggesting it as a solution to bad site structure which is just a step in the wrong direction.
MarketingGuy, you nailed the #1 problem - trying to use this as a band-aid on bad site design.
There are two sides to site structure though. While it might make sense to have certain links on every page from a usability perspective, that can lead to a structure that doesn't work well with PageRank. In such cases, the defect is with PageRank, not with the site. We can change our sites, but we can't change Google.
I don't see using it to combat bad site structure as a problem. There are times when a client can't/won't fix their site structure. In those cases nofollow may be nothing more than a bandaid but that's much better than the alternative of doing nothing.
It is a problem if the site has bad internal linking structure. If you have 100k pages needing to rank for 100k terms then they need to be interlinked properly. Hording link juice through nofollow is no solution to that problem. It might result in some ranking improvements, but on that scale the knock on effect will hurt the site more than help it.
The issue is that if the site has a fragmented link juice distribution, then nofollow will just make that issue worse, while proper site structure would resolve the problem.
I think it's worth noting though that sculpting strategies can have varying level of effectiveness depending on the site size. A 200 page site nofollowing links to 10% of its pages is very different to a 500k page site doing the same. The former will result in a fairly smooth transition - the latter will result in an almost unpredictible manner.
If rankings are so important to clients that they want to go to the level of sculpting link juice then they should be given a kick in the backside to fix the underlying problems first. Bandaid solutions are insanely stupid for critical areas of business - the first time your arm gets wet the solution falls right off.
If you need to solve a problem - then solve it. A bandaid just leaves the risk of the problem occuring again. Why spend time sculping link juice in this way if search engines can just flick a switch and remove the value you get from it?
Fads come and go but I'll happily stake my entire business on the benefits of good site structure - i.e. the ability to weather any algo update.
@ marketingguy
Just to be clear I didn't suggest that a bad internal linking structure is not a problem, I just said I don't see using nofollow to combat that problem as a problem.
I fully agree that in a perfect world clients would never balk at any suggestions and never claim to have resource issues or any other obstacles that interfere with "doing it right". That's not reality though. There are dozens of reasons (valid or not) why clients may not want or simply can't implement suggestions to really fix the problem. When faced with one of those cases where you have suggested A, B & C to really fix the problem and they have all been shot down for one reason or another (valid or not) do you offer option D (nofollow) which could help minimize any potential problems or do you do nothing? I tend to explain that option D is not the preferred method, it’s not a real fix and the potential for problems still exists, but, it’s better than taking no action at all and is simple to put in place. Option D is never the first presented but it’s always in my back pocket in case it’s needed.
If I’m 100 miles away from the hospital and I cut my finger and need stitches I’ll cover the cut with a band aid until I get there. No point in bleeding all over the place until the real fix is in place:-)
That Matt Cutts' quote is a gem, and its also something I've been telling clients for some time now. Dazzlin Donna also said exactly that in her latest guest post over on seo-chicks.
Hey Jeremy - wasn't suggesting that you said that - I just waffled on a little more than usual because this situation came up recently in talks with a client. :)
I agree if A, B or C aren't doable than taking D (nofollow) for positive benefit is an option. The problem I see occurring more than anything though is that A, B and C are just too much hard work to implement so D is taken as the easy way out, which is the wrong approach and it isn't without risk either.