Putting Alexa Rank to the Test

Alexa rank. We all know it’s totally unreliable as an accurate ranking for websites, but for some reason, we still keep checking it. Well, at least I do.

Today, I was trying to explain to someone why Alexa rank is so unreliable. It didn’t seem like they were quite getting it, so I thought I’d try this.

I took the statcounter stats for 19 sites, and compared those numbers with the Alexa rank for each of those sites. The results were pretty much what I expected, and also what I’ve seen when comparing my own sites traffic stats with alexa ranks.

(Remember, with Alexa Rank, lower is better)

Site Pageloads (per month) Unique Visits
(per month)
Type 4,193 1,517 1,093,097 local/ travel 12,322 4,100 7,307,702 News Blog 14,469 9,663 652,472 Blog 21,378 11,595 1,168,943 Blog 59,097 22,482 391,457 Gaming 61,701 25,615 384,627 Books 56,453 27,987 701,743 News/ Politics blog 65,559 39,846 685,256 Football 150,243 50,665 9,921 SEO/ SEM 256,571 74,608 125,565 Photography/ Design 436,716 98,814 78,620 Music 712,886 144,355 130,692 Photography/ Design 450,307 159,205 33,404 Programming 639,658 202,965 50,306 Govt 649,052 405,382 25,312 tech blog, espanol 3,518,005 830,437 62,039 Mac 2,496,956 1,452,791 6,040 Computer science 13,403,525 1,713,931 52,704 Clip Art for Myspace 18,075,839 7,848,084 2,098 youtube downloader

One of the best examples of skewed Alexa data, happens to be this site. It’s ranked under 10,000, while sites that get more traffic are ranked over 100,000. This is because so many more SEO folks have the Alexa toolbar tracker installed in their browsers, in the form of firefox extensions and regular toolbars. Those people visit tech/ web/ marketing related sites (like way more than they visit other kinds of sites.

Don’t get me wrong here. I still think Alexa is a really useful service. Especially when comparing really similar sites. It’s just good to be aware of this flaw.

If you’re wondering how I got access to the stats for these 19 sites, see this post about how to view public stats for thousands of sites. Also, if anyone else wants to share any real stats along with your Alexa rank, please do.

16 Responses to “Putting Alexa Rank to the Test”

  1. Scott Seong Says:

    Please add this stats to your list. I just looked up the stats on Google Analytics.

    Page Views (per month): 13,805 (2.49/visit)
    Unique Visits: 5,555
    Alexa Ranking: 1,136,869
    Site Type: Web Hosting Resource

    This site used to have Alexa ranking of ~760,000 (w/ similar stats), but the rank dropped significantly (~1.5M) when we had a few occurances of downtime. It’s gradually coming back up.

  2. Digitaldir Says:

    Wow that’s real bad websitespam in your comment there… *sigh

  3. B Jones Says:

    Could be, but at least it’s on topic. :)

  4. Matt Ellsworth Says:

    I can say that its skewed toward sites that get a lot of internet saavy folks – simply because those are more likely to be running the toolbar. Others said that if you but the button on your site – then it would make it more accurate. However we ran it on the site for 1 month and the rank didn’t change for the better – it actually went from about 34k to 36k.

  5. Sergey Rusak Says:

    I heard that Alexa toolbar are huge in Korea. Many people use it like we use Google. It would be interesting to track some Korean websites.

  6. Sergey Rusak Says:

    Just ignore SEO sites and everything will be ok with Alexa.
    By the way, is it true that Alexa has agreement with other companies to track stat through their toolbars???

  7. Duncan12 Says:

    This was a very interesting and enlightening article. But I must admit, that reading through the table was not nearly as useful as plotting the data onto a chart… so that’s exactly what I did. I think this might merit another article. Take a look.

    I plotted the Unique Visits per Month (the blue area) and 1/Alexa Rank (the pink line), both on logarithmic scales. Is there a correlation?

    Clearly. In fact it is pretty remarkable. I wasn’t sure what to expect, but this is a very good result.

    But, surely the other free traffic estimation services must be better, right? After all, your article implies that Alexa isn’t that good. Let’s check.


    Ooooffff. Not good. 5 of the 19 sites don’t have any data at all. And the remaining sites… might as well be random. They don’t correlate at all. So much for “real science.”

    What about Quantcast?

    OMG, that is even worse than compete. They don’t have any data for half of the sites. And the remaining sites that they actually have tracked? Why bother?

    What I take away from this exercise is this: Doing traffic estimation is hard. Most services do a lousy job at it. But Alexa is head and shoulders above the rest.

  8. Misafir Says:

    Who cares alexa ? you and me is having 22,888 and having daily unique visitors of 40,000-45,000 only and yet still are we have good Alexa Traffic Rank?

  9. eric Says:

    Good info, excellent the sites comparison. It give me something to think about.
    Good posting

  10. randfish Says:

    The sad part is, Alexa’s no better when comparing Apples and Apples – – you can see there how even inside a very limited set of sites that should be pulling very similar types of traffic, Alexa isn’t even useful in predicting any correlation whatsoever.

  11. B Jones Says:

    Very interesting. Wow. I actually tied putting them in a table, and looking at Alexa rank vis average monthly visits, and really wasn’t able to. It was a lot better than the random sites I used, but overall, I really couldn’t do it.

    Here it is. (added July/Sept average for SEOlogs just for more data)

    URL Avg Visits Alexa Rank
    SEOmoz 208,853 1,071
    SEOBook 277,118 1,138
    Matt Cutts 127,693 1,336
    V7N Blog 71,733 2,055
    ShoeMoney 220,611 2,141
    HighRnks 64,085 2,375
    TL Brokers 11,438 6,239
    SE Journal 116,759 7,353
    BruceClay 4,194 7,824
    CopyBlog 79,890 9,715
    StuntDubl 201,010 12,565
    ProNetAdv 127,393 12,795
    TRankBlog 13,820 13,805
    LB Blog 13,667 13,857
    Jim Boykin 16,264 14,507 44,266 18,538
    MikeIndust 106,763 18,556
    DaveN 170,732 19,851
    Daggle 23,635 21,360
    SEO Scoop 57,107 29,900
    Oilman 43,669 39,125
    SEObySea 27,517 39,970
    ThomBindl 1,811 76,860
    SugarRae 1,526 103,208
    Link Spiel 1,027 140,069
    SearchRank 3,226 141,532

    Thanks for that.

  12. Best Says:

    I think Alexa is useful for sites you don’t know at all so at least you find out whether the site has some decent traffic or not but comparing sites with Alexa is quite useless

  13. John W. Furst Says:

    Hello! Yes the Alexa data is not representative, but still somewhat comparable across similar niches. It’s also likely that there’s a much higher percentage in toolbar installations in the USA as compared to Europe. I heard Asians are crazy about Alexa. Best use is probably to derive trends from those figures. My humbly 2 cents. –John

  14. Is alexa widgets or using redirection myth? Says:

    My website isn’t having good position in Alexa. I have downloaded alexa toolbar & widgets on the website but no result. Even I read on many sites that alexa widget, redirections are myth. Is that true? I need your feedbacks.

  15. B Jones Says:

    You need to get firefox, and install something like the SearchStatus plugin.

  16. Ania Says:

    Now your score is #21,348 it’s great and still “dropping”. Awesome ! :)
    You know, I have an opinion about what you write is what you get! If you can write something with enthusiasm, than you an achieve it. Like writing on Alexa Traffic Rank… Great work & there’s no problem that you did not tell us the 5 ways :))!

    I’ve started a blog on career, jobs, and so on 2 days ago, and now it has a rank #10,000,667.
    Just watch it how fast the rank is dropping!!!
    It’s amazing!

    If I reach #1million I let you know it ;)!