Forum Moderators: open
Checking different datacenters, I determined that I'm ranking slightly higher on www-va. Last week the same thing happened on a different datacenter ... maybe it was www-fi.
Since I'm fluctuating a little bit on some of the datacenters is it possible that I'm about to pass our competitor in the #2 spot? We've been working hard to gain new links over the past couple of months.
Is this a significant event? or am I just being hopeful?
If I am close to a "tipping point", what else could I do to help move up that one notch... more links?
Probably not. It just means that some searchers will see you in the #3 spot, whil others see you in the #2 spot. In the pre Dominic/Esmerelda era, the data across all datacenters was basically the same for the majority of the month. The only real fluctuation you saw was during the time period when a new database was being copied over to each datacenter.
That is no longer the case. Now, the datcenters rarely match. So if you see that you have moved up on one or two datacenters, you need to simply take pleasure in knowing that you are showing up in those positions for a portion of the total searches being conducted.
If you read much more into it, you will often end up dissapointed.
What I find interesting is that I'll jump to about 40th and then slowly slide back down again, over the course of several weeks. It's as if Google is favouring each of the sites in turn, (as it does when it puts a fresh tag on). There are no fresh tags around on that particular search, and I'm not sure what that means.
But to summarise, almost a regular jumping up and sliding down, like a wave breaking on a beach...
Anyone else see something similar?
DerekH
So yes, without a doubt an anomaly in one datacenter can foreshadow shakeup. However, the single difference between #2 and #3 is nothing. if the top twenty has three or four moves, then that might be worth noting.
Right now it looks like -in might be the next datacenter leader... and that is really too bad because it is totally anchor text spam.
In terms of sites ranking highly with few backlinks showing, that seems to be common now. A link is a link, so sites with PR1 backlinks can rank quite highly and have almost none of those links showing.
-in is very disturbing. Even the last holdout against Google's spammy anchor text addiction, page title, seems have been devalued.
But to summarise, almost a regular jumping up and sliding down, like a wave breaking on a beach...Anyone else see something similar?
Maybe yes. One key phrase has been rock solid for around a month - no changes in the top ten. However, on another phrase where I rank ~20th, I've seen jumps to ~17 followed by an irregular slide down to ~23rd. However, I recently acquired several new backlinks and this may be the cause. In a few days I'll be letting rip with a major site update that includes some key optimisation changes. It will be interesting to see the effects of on-site changes.
Kaled.
Yesterday (Nov-11) I was #2 on one data center, #3 on the others. I noticed last week that I was fluctuating between #2 and #3 on one or two data centers.
Today (Nov-12), I am #2 on all datacenters EXCEPT one. Sooo... could I assume that a week or two of higher than normal rankings on one or two datacenters could indicate "momentum"?
Anyone else noticed this?
There are sites that allow you to check several datacenters at the same time. You can easily find them using Google (hint: dance, tool)
Agreed, -in looks like it might be carrying an updated index.Only problem is my index page is missing from the SERPS!
Anyone else seeing this on -in? - I remember a thread a few weeks back where this old dom/esmerelda problem came back to light for a few peopl
As of this morning, -in is the only centre *not* carrying the updates I've recently done to my site. For me, it's the only out-of-date centre.
As time goes on, the evidence seems to be piling up that the new Google index is no longer accumulated at one centre and disseminated - there are a number of us who find that *our* best data centre for *our* pages is different to someone elses!
DerekH