Re: Why no "Banner Ad Eaters"?

["John Kelsey" <kelsey@plnet.net>]

-----BEGIN PGP SIGNED MESSAGE-----

[ To: cypherpunks ## Date: 01/25/98 ##
  Subject: Re: Why no "Banner Ad Eaters"? ]

>Date: Thu, 22 Jan 1998 23:53:19 -0800
>From: Bill Stewart <bill.stewart@pobox.com>
>Subject: Re: Why no "Banner Ad Eaters"?

>At 11:16 PM 1/22/98 -0800, Wei Dai wrote:
>>It seems to me that blocking ads is no different from
>>blocking porn. All of the technology being developed for the
>>latter purpose (PICS for example) will eventually be used
>>for the former.

>With both PICS ratings for web pages and the new TV ratings,
>somehow the ratings only apply to the program and not the
>ads. After all, if each TV commercial had to be separately
>rated, people would rapidly develop equipment to autoblock
>commercials, and that just wouldn't do.

Note that there's a difference in incentives, here, too.
Porn sites in most countries, including the US, have some
strong legal and social incentives to rate themselves
honestly, and probably have relatively little financial
incentive to rate themselves inaccurately.  (Think of the
hassles you get with things like disputed credit card
payments made by someone's 14-year-old kid.)  This makes the
blocking software's job a lot easier.

I just can't see what incentive advertisers have to
co-operate with rating systems of this kind.  How would it
improve your bottom line?  Advertisers are likely to get
paid either on the basis of number of people who see the ad,
or on the basis of number of people who click on the ad.  In
either case, letting your ad be casually filtered out is
just not going to make you any money.  About the only
incentive I can see for letting your ads be blocked is the
desire not to make too many ad recipients mad at the
advertiser.  (Presumably, this is the reason why spam is
almost never used by reputable companies--they don't want to
make too many potential customers angry.)  But this doesn't
seem to apply to webpage ads, which manage not to be quite
intrusive enough to enrage their targets.

>Unfortunately, you're probably right, though providers and
>advertisers who really want their messages to get through
>will find ways to do it. The current banners are nice,
>friendly implementations in that they're easy to identify
>and block; newer ones will just be sneakier.

I assume that, sooner or later, the advertisements will be
woven in so well that it's all-but-impossible to get rid of
them without also getting rid of the useful content you're
trying to read/see/use.

[Good comments deleted.]

>Alternatively,
>they may go to clickthrough payment models - the web page
>owner only gets paid when people click on the ad, though
>perhaps at a higher rate than current "impressions".

Maybe.  Either way, in the long run, ads that don't seem to
be generating sales aren't going to be renewed.  Being able
to count clicks gives you one metric for this; another is
completed sales from those clicks.  This defines why web
page owners that are making lots of ad revenue will have
lots of incentive to make people who use their services look
at and respond to their ads.  People will try various things
to make this happen.  If none of them work, then ad-supported
pages will cease to exist.

Many of the services now supported by ads have other good
revenue models.  Sites like Dejanews and Altavista have
enough name-recognition to do things like sell custom
searching or research from their Usenet and Web databases,
or provide statistical customer profiles that don't reveal
customer identities but are still of use to marketers.

>Bill
>Bill Stewart, bill.stewart@pobox.com

Note:  I read CP-LITE instead of the whole list.  Please CC
       me on replies.

- --John Kelsey, kelsey@counterpane.com / kelsey@plnet.net
NEW PGP print =  5D91 6F57 2646 83F9 6D7F 9C87 886D 88AF

-----BEGIN PGP SIGNATURE-----
Version: 2.6.2

iQCVAwUBNMwNTiZv+/Ry/LrBAQH6KgP/c0o0ZiraSW7PbmDK6YQh+D3wp48cIn3S
manFeca05yuoJDvs6ZKO85ycvVTvVZXBP8tvVDDAMD35CEfxGMNc/0bk1hqS/rx9
/gkHy/OYeKvcSscP9KwWTRtGu+DhHxjNqxOuFFgw6QBCxA5KDdUakF64POlYbi/2
+/kMlWS3on0=
=QsU7
-----END PGP SIGNATURE-----