Subject: RE: AdsOff 



>The "ethics" discussion usually revolves around the ideas that "the ads pay
>for the content on many sites and if you don't see the ads, you're stealing
>the content" and "if the advertisers aren't supported, we won't be able to
>give you the content any more...because we have to get paid for what we do."


i've been through this debate several dozen times with the marketers who
work here at the newspaper.   by that definition, i'm a thief for
surfing with image autoloading turned off.   the argument fails for
several reasons.   


1:  it falls prey to a two-value fallacy.   it assumes that either
everyone gets the ads, or no on does.   if everyone sees them, the
business will survive/if no one sees them, the business will fail.  
that's a badly oversimplified model of actual usage.   in real life,
there will always be a percentage of users who see the ads, and a
percentage who don't.   reducing the issue to two values makes it
impossible to discuss the relative sizes of those two groups.


2:  the argument carries an assumption that each page request costs the
provider a certain amount.   there's a measure of truth in that, but the
finances are more complicated than the argument implies.   the actual
cost of a page request is a few seconds' worth of electricity and
bandwidth.   that's negligible compared to the waste inherent in normal
operations, because few sites run at even a significant fraction of
their theoretical maximum.

all i'm 'stealing' is a small fraction of what would otherwise be thrown
away.   that theft is mitigated by the fact that i'm not /using/ as much
of the provider's bandwidth as i would have done by loading the image
and ignoring it.


3:  the argument implicitly equates the right to draw profit with the
right to survive.   if enough users view banner ads to cover the
operating expenses for the site, survival is no longer an issue.   the
site can support itself indefinitely on those users.   any ad
impressions beyond that are pure profit.   

if you know the expenses and price structure for a site with ads, it's
possible to calculate the exact amount of ad load necessary to cover the
expenses for that site.   that amount is usually a vanishingly small
fraction of the theoretical load.   in the long run, the argument above
often boils down to a demand for arbitrarily high margins.


4:  there are many other financial models for online content, not to
mention other ways of putting ad content into pages.   banner ads just
happen to be easy, and inexpensive to implement.   people use them
because they work, but if they didn't, people would be using something
else.   this ties in with the issue of margins, because many of the
current alternatives are less effective than banner ads.   many people
who use the argument above will also refuse to consider any other
potential revenue stream which offers lower margins.


5:  i personally know of three models for pricing banner ads.. pay per
click-through, pay per impression, and pay for position.   the
click-through and per-impression pricing models are affected by users
who refuse to view the image, but the pay for position model isn't.  
it's also the simplest to implement, and the most common in deployment.

in the pay for position model, an advertiser pays for their ad to be
included at a given point in the provider's site.   the provider makes
no guarantee about the number of impressions or click-throughs that ad
will get, only that the image tag will be included in the given page at
the given location.

the distinction is important.   the provider makes no guarantee that the
image /will/ be seen, only that it /can/ be seen.  the revenue stream is
not threatened in any way by the users who refuse to view the ads.  
many advertisers will pay strictly for the potential of being seen, even
if they know that users might not load the image.

refusing to load images does changes the numbers for the click-through
and per-impression models.   providers compensate for that by raising
the unit price.


6:  breadth of market coverage has indirect value for the provider.  
the cost of denying access to pages is measured in loss of business
opportunities.   the value of granting access is therefore measured in
increased opportunities.. the sales call which is easier to close
because the client comes to your site regularly (with image-loading
off), the client who initiates contact because your site is popular,
etc.


George: if you're following the thread (and haven't fallen asleep yet),
what would be an off-the-cuff guess at the difference in value between
closing a sale in six calls, and closing it in five (or three)?   you're
far more likely to know the numbers than i am.


the value of the benefits has to be balanced against the loss in ad
sales to produce a complete picture.   i don't have numbers to support
the idea, but the losses are so small that i can't help feeling any
benefit at all would outweigh them.


7:  the whole mindset behind banner ads is based on assumptions about
the web which contradict its reality.   it's an attempt to push unwanted
information onto the user, in an environment where the user has control
over what they want to receive.   the fact that most people use browsers
which automatically load images is a convention, but isn't fundamental
to the nature of the web.   anyone who capitalizes on that convention
has to accept the risk that it might change.   the premise that site
providers have the right to demand ad views is an argument against the
user's right to select content for themselves.