Flash penetration stats for March 2009 now live, graphs updated

Adobe has updated the Flash version penetration statistics, showing Flash 10 with an average 74% penetration rate for March 2009.

This sounds fine – considering Flash Player 10 was released less than 6 months before – but when updating my graphics with a historic view, it’s clear the adoption rate has slowed down a bit. Previously, it was the player with the fastest adoption rate to date; now, it has a curve that brings it on par with how version 9.115 did (the second best adoption rate curve to date).

This is speculation on my part, but the relative slow down may probably be attributed to the lack of a killer application of Flash Player 10. If I remember correctly, developers and publishers had many reasons to adopt Flash 9 (AS3, AVM2 performance) and Flash 9.115 (h.264 and hardware-accelerated video playback) – that made several websites (like MySpace) require its installation, accelerating adoption considerably. Flash 10 had nothing of the sort – and, frankly, I don’t think I’ve seen any real work done for or requiring Flash 10. That’s not to say Flash 10 lacks features – hey, I love Vectors and the new 3D capabilities, and  I can’t wait until I’m able to use it in every project – but they’re mostly aimed at developers, and most of the new player features can be emulated in Flash 9 (even if it means lower performance), giving little reason for exclusive FP10 work right now.

Still, that it managed to hold such a quick adoption rate with no artificial stimulus is probably a testament to the overall acceptability of the plugin among users. Name recognition may be helping in making people update faster than they normally would because they know what to expect.

Changing the subject just a bit, other interesting ramification is that now we’re seeing the raise of automatic player version and stat tracking. Websites like RIA Stats and StatOwl also track plugin (and plugin version) penetration, and have lower penetration numbers for Flash 10 – 64% and 55%, respectively. Food for thought.

And I still wish Google would publish global stats from its Google Analytics service.

  • Hi Zeh, I’m not sure if there’s any change in acceleration… these four-times-a-year consumer audits are susceptible to which week in the quarter they’re conducted.

    But Millward-Brown and RIAStats.com measure different things. The first gets a representative consumer sample and then tests their capability. RIAStats starts with a number of websites, and then tests their visitors. The latter will include spiders and robots in addition to consumers. (StatOwl didn’t seem to list a methodology.)

    jd/adobe

  • Zeh

    @jd: I guess I knew that (about how riastats and statowl do things), but thanks for the clarification, should aid readers. Also, I hadn’t taken robots into account – makes sense.

  • Hi Zeh.

    Just a quick note: player 9 adoption was also fueled by better access control permissions, with a finer grained crossdomain and security model. This makes a big difference for big players (MySpace, YouTube), and if I recall correctly, they did have some input into the process.

    Cheers

  • RIAStats.com uses a javascript to detect flash, silverlight and java plugin versions/presence. Generally, web crawlers don’t execute javascript code, so therefore riastats would not be notified of their visit.