Re: Geographic Search Advertising vs. Yellow Pages

From: David Yancey <dyancey_at_proactics.com>
Date: Tue 21 Oct 2003 08:54:26 -0500

Loren Baker asks how geographic search will affect the Yellow Pages. While
Google and Overture experiment, we just went live Monday with the first
*working* national search site that can accept local-area listings along
with conventional website descriptions, so forgive me, Loren, for having a
strong interest in your question.

The Internet Yellow Pages are married to offline data, are saddled with an
economically non-competitive real-world sales model, and have shown no
ability so far to migrate toward a more flexible search technology. Most of
the IYP are nothing but conventional print business directory listings made
accessible online.

So Google and/or Overture and MSN (someday) will have them for lunch. right?

Not so fast. First, the technology to effectively bridge these two very
different kinds of search problems is really hard to build. Our method was
years in the design and two more in the making.

The kind of crawler and parsing methods Google and other "true" search
engines use fail to determine in any *predictable and consistent* way the
geographic context for a website description, much less an individual
page. I could put all you readers to sleep with the details, but will
simply say that the only reliable mechanism that assures a level of
relevance that users will want to experience again is some form of
page/site descriptor scheme. Ours can handle the job, in both the geo
descriptor dimension as well as the page content dimension - no crawler can.

Perhaps some day crawlers will be useful in geo inference, but that is not
going to soon be the case IMO. Added to the technical and linguistic
complexities, a very great percentage of the total online content is simply
off-limits to Google and the others.

Overture has its own mountain to climb, technically, in effectively making
its listings geo-searchable. As do all the PPCs.

So, to defeat the IYP, Google et al will have to become adept at
machine-typification, in the context of a human-managed, quality-conscious
process. This is not an easy business to build or automate, much less run
profitably, and is not done quickly. Indeed, many of us on this list know
from personal experience that the "back-office" processes in Google and
Overture already leave a great deal to be desired.

And, Loren, this is at the level of a few hundred thousand clients. The
Yellow Pages represent *millions*.

Which should remind us of another reason the collapse of the Internet
Yellow Pages is not likely to happen so fast: these companies represent a
richly profitable revenue stream approximately *10 times* as large as
Google and Overture-Yahoo *combined*.

Please do not get me wrong: no one is more interested than our team in
creating a search experience for consumers and business people that can
effectively (that is, from the user's *and* the advertiser's perspective)
integrate local-area and "national" website descriptions. But the Google
approach will need a total re-build to accomplish this. Meanwhile Overture
will find very quickly that their famous competitive bidding model will
effectively price them out of all but the largest local companies'
promotional budgets.

And don't mistake me re: the ability of Google and the other big guys to
figure out how to beat the Yellow Pages and grab a major share of their
roughly $18 billion in US revenues. They clearly have the muscle, and that
big pot of directory gold, er, yellow, is awfully tempting.

But it is going to be a long war, not a short one. And just maybe, some
*new* smart search company will beat the leaders to the punch - again.

PS: be aware that this very week the Yellow Pages industry is holding a
"summit" on this very question. See the meeting agenda etc at:

http://www.kelseygroup.net/ddilm2003/index.htm

Great question, Loren, and nice timing!

David Yancey
http://www.vivante.com
"Searching the web your way"




Received on Tue Oct 21 2003 - 08:54:26 CDT