Regarding Google's "total results" number -- that is, in the blue bar at top right of their SERPs, where it says something like "Results 1-10 of about 12,300,000 for [keyword]", the "total results" number is 12,300,000. This is the number of web pages we'd be competing against for results if we were to target that keyword organically.

So we look at this number a lot when determining which keywords to target. The number fluctuates some, which is to be expected as new sites are indexed and old sites are taken down.

Here's what's really puzzling, though -- running the same searches on the east and west coast can often display startlingly different "total results" numbers. Examples:

"halloween costumes" - 6.45M on east coast / 3.84M on west coast / difference of 68% (okay, but maybe that has something to do with the fact that it's a seasonal/timely search, so here are two more non-seasonal examples)

"disney vacation club" - 2.55M on east coast / 1.65M on west coast / difference of 55%

"pos equipment" - 2.62M on east coast / 1.47M on west coast / difference of 78%

We tested about 25 different keywords, and they aren't all this bad, but most have differences in the 30-48% range.

What gives? Is this yet another number from Google that I have to toss out and ignore because it's worthless? Any insight would be greatly appreciated.

>we look at this number a lot when determining which keywords to target

How do the number of results have anything to do with what keywords to target?

>running the same searches on the east and west coast can often display startlingly different "total results" numbers

that's mainly because you are not using the same Google data center and the results are not synchronized very well across hundreds of DCc.

Thanks, John. We look at the number of results to gauge competitiveness. Let's say we're deciding between two potential keywords that have roughly the same search volume (frequency of usage). If one pulls up 1 million results on Google and the second one pulls up 10 million results on Google, we probably want to target the first one because we will be battling fewer web pages for high rankings. It'll just be an easier fight to win.

I figured these discrepancies may have to do with different Google data centers...but still, to have discrepancies of that magnitude is really surprising. It makes me wonder just how different everything is between data centers.

Any additional insights are more than welcome -- thanks, everyone!

>If one pulls up 1 million results on Google and the second one pulls up 10 million results on Google, we probably want to target the first one because we will be battling fewer web pages

I don't think that is a valid way of thinking. The allinanchor and allintitle results will give you a fair idea,among other factors, but the number of competing pages has nothing to do with the difficulty of gaining top rankings.

I agree that the raw number of returned sites isn't the greatest measure of competitiveness, and certainly we do look at allintitle and allinanchor, as well as a live analysis of the sites currently ranked high. But the raw number of returned sites is one piece of the puzzle -- and it's the one that's most easily generated in an automated fashion.

If anyone knows of any software or automated tools that can generate allintitle and allinanchor for hundreds of keywords at a time, I'd love to hear about it! Then I could forget tracking returned sites altogether.

Of course, my basic problem may still remain -- I have not compared allintitle and allinanchor figures coast-to-coast yet...