Friday, September 28, 2007

Is the concept of a "backlink snapshot" with Google a thing of the past?

I haven't been playing the SEO game very long. My first foray into the art came several years ago when I bought an SEO for Dummies book and the SEM for Dummies book. (my take - SEO for dummies is good to learn some basics while the SEM for dummies was directions on how to use websites, something I already know, and was worthless).

As I delved deeper into the ins and outs of optimizing a website, one of the things that nearly all SEO "experts" talked about was the "snapshot" of backlinks that Google took every quarter. It was these backlinks that supposedly made up the backbone of the next quarter's Pagerank. I've seen a lot of folks posting on blogs and message boards recently about the "snapshot" taken at the beginning of July 2007, and that this was the "big one" in preparation for the Quarter 3 toolbar update (which hasn't occurred yet, just like I predicted). So its something that at least some folks still believe.

It's all hogwash. Hooey. Bunk. BS.

Google has become a much more fluid machine, and even if they don't yet have a real-time "snapshot" of each website, the backlinks, rankings, etc. are updated weekly for new, lower ranked sites, and as often as hourly for more highly ranked pages. Heck, even the Google Watchdog blog that doesn't yet have any Pagerank on the toolbar can have a new post indexed in several hours. Things don't work the way they once did, and the backlink "snapshot" is among the things that just doesn't fit with Google's updated modus operandi.

I've had at least 3 backlink updates in my webmaster tools over the past 6-7 weeks, and the webmaster tools are notoriously behind on the backlink list. If this were the case, why would Google used some arbitrary date months in the past to gauge the worth of a site? The answer is that they wouldn't. It may be entirely possible that when doing a toolbar PR update, they take a "snapshot" one day and begin the update over the next week. But, that would be for purposes of showing a public number that may rise or fall over weeks and months behind the scenes and in the hidden depths of Google's servers.

I can't ever guess what goes on in the giant borg brain over at Google for sure, but I'm almost positive that the idea of a backlink "snapshot" is already archaic in the fast moving and quickly developing world of search engines and SEO. I haven't yet learned enough about SEO to consider myself an expert, but I'm pretty good at deductive logic, and its a fairly simple deduction to make that Google has abandoned the quarterly backlink "snapshot". They've moved past that to better techniques that work more closely to real time. And while Google is evolving, a lot of so-called SEO "experts" are not keeping up. I guess (hope) Natural Selection will weed them out.

Google showing highly ranked, NOFOLLOW backlinks in the webmaster tools

I have many (several hundred) backlinks from PR8 and PR9 sites that are marked as NOFOLLOW. Most of these links are from posts in blogs, comments on articles, etc. I don't expect any PR juice to follow from the links, they're simply me adding to the conversation.

None of these links were showing up in my webmaster tools list of backlinks. I expected that - they are essentially ignored by the Googlebot other than to note the location of the link so that it can be indexed at a later date. The fact that the links didn't show in my list of backlinks didn't bother me.

However, the other day I was browsing through my list of backlinks (over 42,000 now), I started finding these weird backlinks from PR8 and PR9 sites. I made a check and the links are stilled marked as NOFOLLOW, but they now show up in my list of backlinks.

Has Google just now started including these links in the webmaster tools? Do they have some weight, even though they are marked as NOFOLLOW? Or did I just miss them when I was looking before?

Something to ponder...

Wednesday, September 26, 2007

Huge change in SERPs for me overnight

Unfortunately, the huge changes were all in Yahoo (which was acting funny today - see this thread at the Digital Point forums for some examples).

I've been optimizing three fairly competitive keywords for Google. I have them all up to page two in the Google SERPs, one of which is at #11, so I'm getting them close. Today for some reason, Yahoo had the website I've been working on ranking those as #1 for all three keywords. Until yesterday, they weren't even in the top 100!

What's funny is that I'm getting more traffic from Google for the keywords today, even with them ranked #1 on Yahoo. The ratio of Google referrals to Yahoo referrals is about 4:1 today. That should tell you something about the dominance of Google over Yahoo.

Tuesday, September 25, 2007

Some of my favorite optimization tools

Here's a list of some of my favorite optimization tools:

  • Ranking Check: - allows you to check the SERP values for 3 keywords. The check runs on a single domain. You can also store up to ten keywords for free that Ranking Check will track for free.
  • Dig Page Rank: - shows the actual Pagerank for a website from up to 700 DC's (data centers). Shows faked Pagerank also.
  • SEO Moz Page Strength tool: - shows an interesting set of data for a website including the Pagerank, link information, domain age, and a Page Strength value that they calculate for you. Very nice tool to see data in an aggregate form without having to go to many different sites. (If PR goes away, I see a tool like this becoming very popular)
  • Backlink Watch: - get a list of backlinks to your website.
I'll be back with more later.

Update to algo change prediction

Been out sick with the crud this week, but thought I post an update on my prediction from last week.

I made the prediction last week about a major algo change coming up from Google. It didn't happen this last weekend, but there were some small changes. (Read the original post to get the skinny)

The variance for each day since I made the prediction has been:

Saturday, 9/22: 8
Sunday, 9/23: 0 (this is the first time I've ever had a zero. The actual variance was 0.419...)
Monday, 9/24: 12
Tuesday, 9/25: 26

So something happened on Sunday (I assume it takes 24 hours to fully propagate). I actually saw a huge SERP change for an important keyword on Monday morning when I started compiling the stats (a keyword not used in my formula).

I'm still waiting for the "major" algo change that I predicted to occur, but like I told thegypsy over at Digital Point, I could be wrong. If I am, I'll eat a plate full of dog turds (provided the major algo change happens soon).