Alexa is BUNK

Dont worry about Alexa.... it is bunk.
One thing you all have to realize about Alexa is that its information comes from the use of its toolbar and not from some magical place.
If people (with the tool bar) dont visit your site, this does not mean that your site doesnt receive hits.

We have a very advanced Web Metrixs tool called Web Side Story installed on our web site and I use its information against Alexa to see what is what.

Alexa showed that our site dipped from 5000 to 8000 in a day. But Web Side Story showed traffic had stayed the same and even increased a little over the same time as Alexa said we were crashing and burning.

Short of it is Alexa is a tool thats accuracy is only legit if everyone on the INternet used it.
And guess what, they dont.

Get a good Web Metrix tool like Web Side Story, Traffic or even Web Trends and forget about toys like Alexa

True. But we are finding that it is very bad PR and misrepresents true stats. That's not good. The bulk of our 7000 members are teens and for some reason they keep fixating on Alexa Stats compared to our chief rival. It's just frustrating.

Its true that while Alexa may not be a accurate representative of what your site traffic demographics are, it does not prevent prospective advertisers from using it as a tool to decide -where- they should advertise.

I don't know what other advertisers look at... but I'm sure most recognise what Alexa can and can't do... dropping from 5000 to 8000 wouldn't mean very much (unless it was the three month average that gradually dropped) - 8000 is still a very popular site and more than likely meets any claims of traffic that they tell you when you query them about advertising.

If you're under say 15000, then you've got a great site, under 70,000 and it's still really good, between 70,000 - 100,000 and it's ok. Over 100,000 and the Alexa ranking deterioates too much because one or two users can distort the results too easily.

That's the way I look at it anyway... I'm sure there are people out there that put too much face in the actual value though.

lol. we went from 303,659 to 91.696 in one week. go figure!

I've told this some times: install the Alexa toolbar on your computer and yu've a top 100.000 guaranteed.

We did that. That's my point. Alexa is crap. You can really skew it.

ok. Crazieness. We are now at 47,330 . Go figure. And in their directory are listed as #1 for our topic. wow!

Its funny, its there, its probably the most accurate ranking of sites that exists, it does it the only way it can, by counting actual users, can be manipulated to hell, but so can anything else.

The fact is that if you look at the American sites in the top (I have no idea what those Korean sites are about other than floating ads) listings they do in fact reflect what I would consider the top top sites.

The fact is that being in the top 10,000 is a great thing to tell a potential advertiser. Being in the 50 in a major catagory is another great thing to tell an advertiser.

They like to exchange high numbers for low numbers, so live with it, love it, and use it.

Until somebody finds a way to spy on every user in world to see where they are going, seeing, doing, we are gonna just have to live with a voluntary "random" sample.

I can say with out a doubt that I have ran experiments on sites of all sorts of ranks. I think someone mentioned here that its like a slope that just keeps getting steeper and steeper and steeper.

Its easy to hit 11,000. It is not easy to crack 10,000 and its damn hard to move up in that area with out a continued growth in traffic.

Is it even worth taking the time to skew it though?

For example, is there any purpose to a better ranking? If I'm 700,000 and can get down to 100,000 by skewing, would it be worth it other than just to say I'm 100,000? Could it increase traffic?

Forget Alexa .. think Google!