[thelist] Browser Stats

[aardvark]

> From: Michael Kimsal <michael at tapinternet.com>
> >
> >> i'm still having trouble considering 1 million people to be an
> >> insignificant number...
> >
>
> I don't know why.  Everything is relative.  When you are
> talking about hundreds of millions or billions of something, one
> million *IS* insignificant.

can i kill you?

seriously... one person is insignificant... out of billions, you are
nothing... a speck on a speck on a gnat's ass...

now, i suspect you don't much like that concept... and i suspect
anyone coming to a site who is trapped on an old setup, chooses to
use an alternative setup, or is disabled in some way will feel they
have been herded to the side... sort of a user-eugenics move...

> When you're 4 years old, 10 seems ancient.  When you're
> 80 you see that 10 is pretty insignificant.

when you're 4, you feel significant... when you're 80, you still feel
significant, but you're used to being shoved around...

i don't see how that corollary benefits anyone other than the
developer... it's a cop-out...

> >> 3% may be an insignificant fraction to you, but i do know many
> >> clients who consider 3% to be an unacceptable loss of viewership...
> >> imagine you've been hired to redesign a site and tell the client
> >> that they will lose 3% of their readers... or better yet, tell them
> >> that they will block out 1 million people... see how they dig
> >> that...
> >
> Yes, if you walk up to someone and say
> "you're going to lose 1 million potential customers because of this
> particular code" and leave it at that, yes, some might freak out.
>
> Follow it up with "but by doing this, we're making your business
> more attractive and accessible to 279 million people" and the
> discussion would be over.

but you're not making it more accessible to anyone -- you're keeping
it just as accessible (conceptually) to the 279 million, and now
permanently shutting out a million that otherwise could get in...
it's a net loss of bodies -- potential customers...

if you lose current customers, then the swing can be double that...

as for attractive, who's to decide?  you?  i see a *lot* of sites
that block me out on alternative browsers or older browsers that are
decidedly *not* more attractive... really, a colored scrollbar
doesn't do it for me...

> Take it further:
> 2 million potential customers won't ever do business with you
> because of some article in 'Christianity Today' which linked
> your product with Satan.
>
> 4 million people won't do business with you because your
> spokeswoman just announced she's got a live-in lesbian lover.
>
> 1.5 million people won't do business with you because you
> do business with a company in South Africa
>
> 3 million people won't do business with you because the CEO
> was too chummy with Clinton in 2000 and they think
> your company is a bunch of left-wing nuts.
>
> All those numbers seem big, but they're not that big, and they
> only represent potential, not hard numbers.

no, they can represent true losses if they consisted of any current
customers... regardless, your examples are incongruent because they
are decisions of policy and direction... a web site is not a
statement of policy unless you want it to be, and making it such is
kind of silly...

go print newspaper ads in green and red with the exact same
brightness, because that way you can make a statement about color-
blind users... go remove the ramps from your building... go repaint
the lines in the parking lot to limit all space to only SUVs...

*that's* what's going on... decisions are made to block people for no
reason other than developer ego...

> If, as I said before, you ran a '1992-era mac hardware' website,
> yes, requiring IE6 and Flash might be stupid.  But you wouldn't
> be alienating 3% of your visitors - you'd probably be alienating
> 70-80% of your potential customers.  Even if the total potential
> customer base was 50,000, 70% is a big portion.

absolutely...

> Statistically 3% is nothing.  A blip on the radar.  That's why most
> poll reports say 'margin of error + or - 4 points", cause it's hard to
> measure things accurately at such small percentages.

most poll reports give a margin of error appropriate for the report,
not most give +/-4... but your argument of statistics, however
flawed, still ignores the fact of a million potential or existing
users shut out for no reason other than 'design'...

sure, if you can guarantee your loss of 1million potential users will
result in a gain of more than 1million potential users, then you have
an argument, one that gains weight as that gain climbs well above
1million... but until then, it's a loss, no matter how small you try
to make it sound...

> By all means, keep writing decent HTML which should work OK
> across older browsers.  But for goodness' sake don't go out of your
> way to support NS3 or something similarly ancient.  Those browsers
> deserve to be put to rest.

but their users don't...

--
Read the evolt.org case study
Usability: The Site Speaks for Itself
http://amazon.com/exec/obidos/ASIN/1904151035/evoltorg02-20
ISBN: 1904151035