Re: AltaVista Testing Paid Search

From: Matthew da Silva <mdasilva_at_ibd.yamatake.co.jp>
Date: Wed, 12 May 1999 13:59:22 -0500 (CDT)

Thanks David Yancey for finally admitting that business
users are suffering information overload and are only given
two options: reliance on virtual catalogs such as have been
established by the VerticalNet team. Or, on plain hard
searchin'. I'm not convinced that vertical portals fulfill
a basic need for information transparency. The emergence of
paid search engines supports me here because they underline
a fundamental lack of substantial gains made by Internet
consumers in locating information quickly. The result?
Persistence of brands, and proliferation of banner ads
which only exacerbates the problem.

The issue of search engines must enter into any talk of
advertising on the Internet. This is why our current thread
is long and lasting, and also why the idea of a free
Internet is not synonymous with an 'ad-supported' Internet.

Thanks also for pointing me to the FAST search engine. That
is a good launch point on the road to achieving
transparency, but it still depends on the quality of content
for its usability. In future, I see a real need for search
engines that are specific to a discipline or a market
sector. The Notary Public paradigm proposed by Mr. Arnold
could be leveraged in execution here, since it is clear to
anyone using Yahoo's flawed lists that the knowledge owned
by a cataloguer limits his/her catalog's usability.

DAVID YANCEY WROTE:
> all kinds of folks, with all kinds of
> search needs, and with precious little discretionary time to
> puzzle through the mysteries of the web. This means, for
> example, that many "family" and business oriented tools and
> directories will continue to emerge, services that basically
> do an end run around the traditional surf/search process.

The issues here are twins: confidence and data volume. It is
necessary, for success by a tightly-focused, 'themed' search
engine, for its users to have confidence in all the
information being in the 'engine when *they* want to search.
For your 'engineers, many of their users know as much or
more about the topic. Never- theless, most users want to
get all the latest data without doing the work of updating
catalogs they now must own. Hence, a confidence factor is
paramount and hence the 'engine must be employing top people
willing to do very hard work. Paying for their services is
the issue (need ideas here). A notary system could roll,
but I suspect adequate commitment, finally, to ensure
continuity. Again, consumer confidence factors into the
equation. Should a researcher just stop doing the work,
what would happen to the Site? Not only would that Site
die quickly, but other Sites will be tainted by its
failure.

I'd like to know what are the "tools and services" that
David has in mind. The old saying about doing the job
yourself holds true, *if* you want it done properly.

A rarely-debated alternative is adoption, aggressively, of a
meta- language that would serve to classify accurately,
data. Under an ISO scheme, such a project should have the
credibility it needs for each industry to adopt a
'standard' terminology for parts and specifications naming.
Such a scheme would help the 'Net to become more useful for
the professions and for tradespeople, merchants, consumers
and other individuals who depend on their industry for
their livelihood. Hey CommerceNet: are you a subscriber to
the Online Ads list?

Matthew da Silva
Online Solutions, International Business
Yamatake Corporation, Tokyo
mdasilva_at_ibd.yamatake.co.jp


Received on Wed May 12 1999 - 14:22:27 CDT