JD on EP

First, the usual disclaimers:

• The Millward-Brown consumer survey does not accurately reflect *all* machines on the internet, just ones consumers can control. If your audience is in a library or a particular business or other situation where individuals cannot install, then your results may be different. (Although, in the past, when business users were separated out in such studies, the results did tend to be similar to consumer results.)
  
• The testing asks whether people can view a series of webpages in their current browser, without installation. It is a measure of actual current capability, and is not a JavaScript or server detection routine.
  
• The testing is of major versions in clientside capability, not minor versions, but you can interpolate the results against release dates to estimate.
  
• The sample size is calculated against entire regions, and the sample size would be too small to reliably estimate viewership capabilities for a single country within a region.
  
• The results are for a general consumer audience, and may or may not match the audience for any particular website on the web.
  
• I can't give good sourcing on the data today, and don't feel quite comfortable just using a comment from another staffer in a weblog, but I know that in the past it has taken a good while to confirm and then publish the results, and so such a gap isn't unexpected. The numbers are also in line with the previous two surveys and verbal info I've received on 5-10 million successful installations per day. You may wish to wait on talking to clients until the Adobe site goes live with the results, though.
  

The 80% level: In the past, clients would usually hold off on using new Player features until the audience was already up towards the 80% rate. I don't know why -- support costs, fear of losing visitors, whatever -- but 80% viewability was usually the tipping point for when client resistance would fade, and you could actually use new features.

But the past six months we've already seen an explosion in deployed Flex projects go into public use. After a decade of proven adoption rates, the approvals to start Flex projects came after only a few months of Player 9 distribution. Not only were these decisionmakers correct, but they got their 80% audience size early, within nine months. The gap between when distribution starts and decisionmakers agree has dropped to practically nothing.

Adobe Flash CS3: There's never before been a release of the visual authoring tool, where audience capability has already been above the 80% level. Put another way, we've never had a situation where you could use the full features of a new tool the moment you open the box. This time we do. And there are also the integration improvements with other key design tools, the smoothest realworld workflow ever for design professionals. And CS3 adoption is estimated to be particularly rapid, in large part due to the new Macintosh improvements.

I don't know what effect these reinforcing trends will have, but it's a significant change, and could have surprising ramifications over the next half year. The rest of the world will be creating SWF9, quick.

Future improvements: One other implication of so many decision-makers adopting Flex so early -- and having their expectations exceeded -- is that this will color perceptions for the next round of improvements. New Player features will results in more job approvals, sooner. (For instance, there's been no real talk about voice yet, but when there is, there will be widespread belief in its rapid deployment capability.) Not only will new technical capabilities be deployed to the world quickest atop Player, but the world is also increasingly of the belief that Flash will offer practical features, fastest. This gets projects approved, contracts signed. You can do fun stuff more quickly.

We're in new territory here. The old 80%-in-12-months maxim is gone. It was blown away by how Flash is now a core part of the web, used by the majority of major websites as a matter of course. That pulls more installations, faster and faster. And that pulls new capabilities clientside faster, too.

One thing's for sure... whatever new technology is introduced over the next year, it's going to be practical and realistic out in the real world, fast.....