Forum Moderators: phranque
6 months ago we were ranked much higher than we are today and were getting 1/2 the traffic we see today. does anyone know why is this?
i find it rather odd that as the site continues to grow our rank with Alexa continues to drop.
most say Alexa rank is meaningless, but i'd still like to know what would cause this behavior.
Alexa should be abolished because it gives a bad impression that your site might not seem popular solely based on the toolbar usage.
i would rather not be listed at all if Alexa is going to make my site seem "less popular" when in reality it's not - just less people using their crap toolbar.
> I've been curious as to why my site in Alexa rank continues to decline,
If it was about 5k to begin with - then the figure was fairly inaccurate...
But why is this post in sep?
Most polls are accurate to +/- 3%. If you only poll a tiny percentage of the market then how can your results be anything close to accurate? Instead of +/- 3% it becomes +/- 50% or more.
The real answer is that Alexa doesn't have a sufficient poll size to report anything that is remotely accurate.
Right now my sites are largely up according to Alexa, LMAO.....they are really all down due to seasonal factors. Not only is it inaccurate it has no real comprehension of time.
Come the first week in Jan my sites will take a hit on Alexa, at exactly the same time they bounce back from the seasonal lows.
Simply put, it is a waste of time to most people!
>> i would rather not be listed at all
Then ban the bot.
I find it good to find out who your real competition is.
ie, you would be woried if a competitor with an alexa ranking of 9000, was slashing prices. You would not be worried if that same competitor had an alexa ranking of 1,000,000
Its good for separating the competition from the want-to-be's
Now a site could block Alexa, say 6 days out of 7, to lower its Alexa score, as a form of counter-intelligence, but i doubt that it has come down to this yet.
"It's not how big it is it's how you use it"
" Alexa rankings are a poll.
Most polls are accurate to +/- 3%. If you only poll a tiny percentage of the market then how can your results be anything close to accurate? Instead of +/- 3% it becomes +/- 50% or more. "
...If you only poll a tiny percentage of the market...
that tiny percentage is largely Asian
if your site is one that is heavily used at major corporations or university locations that really limit toolbar installations
by installing the Alexa toolbar and hitting my own site, I was able to bring it from about 110,000 to 45,000
I believe this is incorrect (and actually, the way you've expressed it is also incorrect). If you assume the population you've polled is a random sample, and the population fits the normal distribution - the ole bell curve - then you really only need to sample or poll about 1000 people to claim it's accurate to within 3% or so.
Where the stats quoted go astray is that in addition to the +/-, there should be another % as well. i.e. it should say the poll is accurate to within 3%, 95% of the time. the other 5% of the time you'd expect that the survey results are further out of line than a 3% margin.
In short, as long as you've actually selected a random cross section of the population (tough to do) and the other assumptions hold, then you can actually claim numbers like +/- 3% with a sample size of only about 1000. (that's also why you'll sometimes see "+/- 3%, 19 times out of 20" - or 95%of the time).
I guess that begs the question, is there any commercial value to being ranked in Alexa?
I don't know, is there?
Alexa is not a random sample, it is webmasters and Geeks. A tiny percentage of the true population.
Will these people not be more partial to certain types of sites. Yes, WebMasterWorld.com should do great on Alexa, it is a Geek site. But, that is not a random sample.
At the end of the day Alexa can predict popularity among Geek sites, but, not among the general population, which is the bigger number.
Alexa is a small % player, whatever it says doesn't much matter as Yahoo and Goggle both replicate the functions with a much larger poll.
So the real question is why do we need Alexa at all? Why not just use the numbers from Google, Yahoo and to some extent MSN?
There is a bias among those that install the Alexa toolbar (just like the Google toolbar), especially among Geeks. WebmasterWorld's actual rank will be overstated because of this bias. Two good examples might include myspace (currently 14) versus Wikipedia (around 35). Since Wikipedia has lots of Geeks writing articles, I suspect their actual traffic is overstated relative to myspace.
The other problem with Alexa is that it is subject to gaming because of the way the tool works. This gaming is often apparent by looking at a long term trend. I used to frequent a blackhat forum (I'm a convert now, even though the money was better back then) and that forum had a very high Alexa ranking even though it was relatively quiet there.
As a gross estimating tool, it is actually pretty interesting - you can pick a top 50,000 from a 1,000,000+ site.
So to answer your question - you've got more visitors without the toolbar installed than you did in the past (for whatever reason).
It just makes you wonder...how many of those 30,000 or less sites manipulate their own total to make advertisers believe they are worth more. Sad really. If you are selling advertising it makes you ponder the idea of installing the toolbar to make yourself look better...like a woman does when she applies make-up ( can't hurt to look your best, even though it's not what you really look like :) Not that I would
The long version:
I am changing my tune from saying it is completely worthless to saying it can be used as a general guideline and can be an indicator of certain things, so long it is used along with other tools and it is used with the knowledge that it can be influenced.
Were someone to use AR as a solitary guideline to determine a site's worth, I would say this person perhaps lacks certain knowledge or information and is possibly making an uninformed choice / decision.
Were someone to use AR along with several other tools provided by alexa and then several more tools provided by other parties, then it is likely Alexa can help in the determination of said value.
For example:
A site with an Alexa ranking of 10k means little in and of itself, very little.
A site with an AR of 10k, a listing in the Y! directory, a Gpr of 5, a load-time of 12 seconds @ 28.8kbps AND a linkpop of 5k+ is MORE likely to have something under the hood.
Ultimately nothing is guaranteed but as a rule, if ALL indications point towards the fact that a site is a little better than the rest, then it is likely the site IS more than just another...
1) Not many self respecting geeks I know would be caught dead with the alexa trashware installed on their PC's.
2) Alexa toolbar on works with Internet Exploiter 5.0 or greater and no self respecting geek would use THAT piece of trashware either ;-)
