Some more boring statistics

    Good time, dear.
    As I understand it, the man loves numbers - he's a techie for the most part. And often topics are published on the resource ( one , two, and even with the donkey’s death indicator No. 6 - three ) with optimistic browser statistics. It shows that the share of ie6 has fallen to its lowest level and continues to decline.
    This is a global trend. I decided to compare it with the regional one. Let's see what happened ...

    I work in a web studio (I will not name) in Zamkadia. Specifically, the city of Samara. Our customers are mostly from surrounding regions. I climbed into the analytics module from Google, took our three sites and tracked the share of the ie6 browser for 14 months. Why exactly him? Well, probably due to the fact that he gives the headache more than anyone else (although personally I am more “afraid” of the Opera). Then, for the same time period, I selected donkey statistics from three sites: liveinternet.ru (statistics for Russia), w3schools.com (without reference to the region), gs.statcounter.com (with the filter settings for Russia).
    The exact addresses of the statistics pages:
    liveinternet
    w3schools
    gs.statcounter

    Our sites (I will not name addresses (:) I conditionally divided by target audience to compare the share of the browser in the corporate sector and among ordinary people.
    The specifics are as follows. One site represents a furniture company focused on individuals and legal entities - its target audience - and those . and other
    second site - site of the company engaged in the rolled metal - its CA - legal entity.
    The third - the company manufactures plastic windows, site focuses more on individuals.
    // All images are clickable

    first chart - Three regional statistics site the arithmetic mean + VALUE. of each month the percentage of new visitors to not less than 75%, the share of Russian-speaking visitors is not less than 99%, the proportion of visitors from the Volga Federal District -. not less than 90%.
    image

    As you can see, for this browser at least 30%, and on average about 40%. The data is noisy, but there is a downward trend, if an approximation is made, it turns out that the drop is about 0.9% per month.
    All three graphs are weakly correlated with each other (55% data correlation), but the differences are not so significant as to suggest the dependence of the browser share on the target audience.

    The second schedule. Here the picture is different and much more optimistic. The difference between liveinternet and gs.statcounter.com is interesting - although both take statistics for Russia. Cross-correlation of data is about 98%. The last graph is the average of the previous ones. No comments. I would like to add a few things from myself.
    image





    image


    Schedule number 2 - there is a fall after November 2008. Just in the middle of November, it was announced that Firefox 2 browser support would be stopped ( news on the hub ), after which by January 2009 the share of FF 3 had grown by 3% according to w3schools.com). Coincidence? I hope not.
    After July 2009, the share of ie6 drops sharply on both charts ... At the end of June 2009 FF 3.5 is coming out - maybe a coincidence, because an increase in the proportion of FF at that time was not fixed.
    One more thing:). The linear approximation of the graph from our resources has the form y (x) = -0.009x + 0.504. If we solve this equation for y = 0, then x = 56, i.e. according to these statistics, our favorite browser will disappear by April 2013 :).
    Similar "very scientific calculations" for the global schedule give us a date ... December 2010.
    Well, let's wait)

    Also popular now: