Alt attributes: used or ignored or penalized by Google?

Question

I've never been a hard-core SEO person. Mainly just sticking
with good
markup and good content and passing the buck when clients
what true SEO
consulting (mainly telling them to hire a SEO consultant for
that).
I got into a discussion a while ago regarding the effects,
either pro or
con, of using alt attributes as your main headings.
I know that Header tags are relatively important for Search
Engine rankings.
You want decent keywords in your headers.
On many of the sites I build, I reserve the H1 Tag for the
site logo, which
is usually an image:
<h1><img src="WidgetsInc.gif" alt="Widgets Inc.:
Quality Widget Provider for
the Greater NorthWest" /></h1>
And when I do HTML production for other graphic designers,
they often want
at least the H2 tags to also be image-based.
IMO, I've always considered the above completely valid and a
good way to
handle things. Since alt attributes are truly supposed to be
content (if for
no other reason than it's proper accessibility), google
should read them and
treat it as it would text wrapped in the H1 tags.
The counter argument I heard was that google ignores alt
attributes because
they could be spoofed. The example being something like:
<h1><img src="CocaCola.gif" alt="CocaCola rules,
Pepsi sucks!" /></h1>
And, I thought, hmm...that's actually a valid point. What's
everyone's
thoughts on this? Any known data regarding this? Does Google
ignore alt
attributes?
If it does, would image replacement techniques be any more
valid?
Specifically the practice of using a background image to show
the image, and
moving the actual text off-screen using something like left:
-999. That
would produce the same visual result and be (almost) as
accessible. However,
I was under the impression that Google ignores negative
positioned items for
the same reason as the above argument regarding alt
attributes (they can be
spoofed with fake keywords).
Thoughts? Opinions? Suggestions? Conspiracy Theories?
-Darrel

Answer 1

Hi Darrel,
I actually have 6 identical pages in a test folder of mine
where I
eventually will know for sure... I think.
The 3 pages are done with different methods as in the one you
describe
(negative text indent) and one that uses an image with an alt
attribute
and one that uses Thierry's perfect IR. They have the exact
same short
content. The last 3 pages uses the above methods but without
any alt or
title attributes.
It's only a week since I uploaded the pages so it's too early
to say
which will work better if any difference at all. I see the
folder has
been visited by googles spiders but no results yet.
Not much help I know but I also didn't knew for sure... thats
why :)
darrel skrev:
> I've never been a hard-core SEO person. Mainly just
sticking with good
> markup and good content and passing the buck when
clients what true SEO
> consulting (mainly telling them to hire a SEO consultant
for that).
>
> I got into a discussion a while ago regarding the
effects, either pro or
> con, of using alt attributes as your main headings.
>
> I know that Header tags are relatively important for
Search Engine rankings.
> You want decent keywords in your headers.
>
> On many of the sites I build, I reserve the H1 Tag for
the site logo, which
> is usually an image:
>
> <h1><img src="WidgetsInc.gif" alt="Widgets
Inc.: Quality Widget Provider for
> the Greater NorthWest" /></h1>
>
> And when I do HTML production for other graphic
designers, they often want
> at least the H2 tags to also be image-based.
>
> IMO, I've always considered the above completely valid
and a good way to
> handle things. Since alt attributes are truly supposed
to be content (if for
> no other reason than it's proper accessibility), google
should read them and
> treat it as it would text wrapped in the H1 tags.
>
> The counter argument I heard was that google ignores alt
attributes because
> they could be spoofed. The example being something like:
>
> <h1><img src="CocaCola.gif" alt="CocaCola
rules, Pepsi sucks!" /></h1>
>
> And, I thought, hmm...that's actually a valid point.
What's everyone's
> thoughts on this? Any known data regarding this? Does
Google ignore alt
> attributes?
>
> If it does, would image replacement techniques be any
more valid?
> Specifically the practice of using a background image to
show the image, and
> moving the actual text off-screen using something like
left: -999. That
> would produce the same visual result and be (almost) as
accessible. However,
> I was under the impression that Google ignores negative
positioned items for
> the same reason as the above argument regarding alt
attributes (they can be
> spoofed with fake keywords).
>
> Thoughts? Opinions? Suggestions? Conspiracy Theories?
>
> -Darrel
>
>
Kim
http://www.geekministry.com