MediaWhiz Blog: Archive

SERoundtable has posted on an SEO Chat thread, Main deciding factor [when choosing directories in which to list your site].

Some joke that the best directory is the free directory, or better yet there is no recip link required for inclusion. There is a good discussion about design and the visual appeal of a directory. While I do agree it is nice to see, it doesn’t necessary mean the directory is always the best.

I agree that it’s best to stay away from directories that require a reciprocal link. My earlier post on TrustRank stressed that low quality directory listings may not help your rankings (and could possibly even work against you). When in doubt: get listed in those places you would like to be even if Google didn’t exist. Yes, I toot that horn a lot, but I truly believe it!

This thread on SEO Chat has some interesting advice. Namely, that when you link out to an "un-themed site", you should use the rel="nofollow" attribute. I try not to get emotional about these things but this attitude really ticks me off.

When did we all become Greedy Linksters? I guess it started with PageRank and morphed into the mess we have today, with blog spamming, Arelis et al. Admittedly, I’m as greedy for links as anyone. Heck, we even made a blog about how to get as many/the best links as possible to your site (p.s., isn’t our blog great? you should link to it).

So your site is about blue widgets, and your hosting was graciously donated by HostingCo Inc. You, of course, want to link to them in your footer with "Hosting graciously donated by HostingCo Inc." The problem (apparently) is that HostingCo Inc. isn’t ranking for "blue widgets", so you’re afraid that by linking to them you will lose hub points with Hilltop.

Link to them. Trust me. The algo will get over it. Don’t be a Greedy Linkster! *steps off soapbox*

I don’t normally visit Jill Whalen’s High Rankings Forum, but there’s a fantastic thread up there called Out of Link Building Ideas.

Here’s what I’ve tried so far, can any one offer some suggestions? The
ones below have worked out somewhat ok (Marketleap 758, PR4) but
obviosly I’d like more. Yes I am keyword optimizing all these links and
will continue to use the techniques below but they’re getting less
efficient every day.

If you’re a serious white hat link builder, and I am, you know what it feels like when you’ve given it everything you’ve got. You’re in all the directories, you’ve gotten 12 articles syndicated, you’ve traded links with everyone possible ranking top 500 for your term (and a ton of people who aren’t), etc. etc. But you’re still ranked #7! How will you get to #1?

At this point, I usually do two things:
1) Rent a couple very powerful, on-topic links (prepay for six months — it won’t help unless it stays up for a while!)
2) Put my content creation in overdrive (e.g., pay an author to write 3 articles per day, instead of 1). The content will drive traffic and links (very slowly, but remember, we’ve already gotten all the "easy" links)

And if I was a different type of person I might even try
3) Think of a very creative way to make your site better or say something crazy. Aaron Wall is always tooting his horn on how creative / excellent / terrible ideas attract a lot of links.

PageRank isn’t dead yet–but it’s slowly dying. Although many people still use PR as a rough metric when weighing the value of the link, certain developments lead me to believe that its days as the primary metric of the value of a link are over.

1) Toolbar PageRank is now only updated every three months
2) Most SEOs agree that its impact on the SERPs is now extremely limited
2) PageRank attempts to measure the "absolute" link power of a page, but certain algorithmic components (Topic-Sensitive PageRank [PDF], Hilltop) only reward sites that are linked from their topical community

This all begs the question: if we aren’t considering PageRank, what metrics should we use to guage the value of a link?

McMohan asks, "’What if’ Google were to pull down that greenbar from its toolbar?" in this thread at WMW:

My list would include these, in order of importance -

a. Alexa Rank (With all its anomalies)
b. SERPs that the site/page enjoys for its target keyword/s.
c. Quality IBLs if any (such as dmoz)

Although Alexa data is indeed shaky, we have reason to believe Google will soon be using traffic and user behavior data as a way to boost/devalue links and pages, and Alexa is the best source of that data freely available. Of course the page’s position in the SERPs would be ultra important in light of Topic-Sensitive PageRank [PDF alert] or Hilltop. Finally, quality IBLs such as DMOZ or .gov’s would certainly help a site gain "trust" in the eyes of a search engine (TrustRank [PDF]).

The downside is that it will now take more time and effort to figure out which links are good, and which are bad. We may even see "Ask yourself, ‘Is this good for my visitors?’" pop up more often in link building forums. White hats rejoice

Question: What Should one do if DMOZ Link got rejected?

Answer: Complain about it.

Let the productive mud-slinging begin!

And P.S., I’ve never been offered a bribe as a DMOZ editor, although I’m certain that it does go on (probably on a limited basis, and only in "high margin" categories).

P.P.S, If you would like to get listed in dmoz.org/Games/Gambling/, send a $2,000 cheque to:

Andy Hagans
137 Fake St.
Anytown, State 12345

And don’t forget to double the amount if you want a keyword-rich title.

Disclaimer: That’s a joke, son, a joke…

Well I’ve been tooting my horn about the greatness of article syndication lately, but I’ve failed to highlight one aspect of it that may dampen it’s effectiveness: having the same article syndicated on many sites creates tons of duplicate content.

The implications of this are unclear though. Firstly, there’s a lot of confusion out there as to what a duplicate content penalty is, including if it even exists and how it works. We do know that when Google finds multiple copies of the same exact content, it tends to use the first copy of it (or, sometimes, the copy with the highest PageRank) when delivering results. This is known as the duplicate content penalty.

I am not however convinced that this is a penalty per se. I think of it more like a filter. The term "penalty" seems to imply that a site has done something naughty and needs to be punished in the SERPs, whereas a "filter" aids search quality by dampening the rankings for results that a searcher will probably not find useful.

Even if that issue was settled, it would be unclear if the fact that the pages were filtered/penalized for duplicate content would affect the value of a link from those pages. It is certainly logical to believe that link algorithms are run separate from other algorithms, with the results then aggregated into the overall search algorithm afterward. The short answer: I don’t know.

How this relates to article syndication: if every syndication instance after the first is truly penalized, then even if your article is syndicated 100 times, you will only gain the value from being linked from one of those instances (either the first instance, or the instance on a page with the highest PageRank). But both my instinct and anecdotal evidence tells me that this is not the case. Thoughts, anyone?

Gary Price of Search Engine Watch Blog has found yet another Stanford research paper [PDF].

In this paper we study how web pages can be interconnected in a spam
farm in order to optimize rankings. We also study alliances, that is,
interconnections of spam farms. Our results identify the optimal
structures and quantify the potential gains. In particular, we show
that alliances can be synergistic and improve the rankings of all
participants.

Seems like old news to me