By Hays Goodman
Associate Editor
Search engine technology functions for a
newspaper Web site in one of two ways: externally and internally. Externally,
newspapers rely on consumer search engines such as Yahoo!, Google and AltaVista
for a portion of their Internet traffic.
Though most people will certainly know to type in
a popular newspaper’s name to access their site, like dallasmorningnews.com,
frequently users will return to retrieve a newspaper’s individual articles. If
a paper makes thousands of articles available to a consumer search engine, then
the potential reach of the Web site expands in a similar fashion, since the odds
for matches and hits on a given search phrase also grow.
As reported on the site searchenginewatch.com,
two major Internet traffic ratings sites, NetRatings and Jupiter Media Metrix,
have a consensus over the most popular search engines and their slice of
traffic. In terms of Web audience reach, the top three are MSN (between 30
percent and 37 percent), Yahoo! (between 30 percent and 34 percent) and Google
(between 28 percent and 30 percent).
According to market research firm StatMarket,
about 52 percent of Internet users found Web sites via direct navigation or
bookmarks. Random surfing accounted for 41 percent and search engines
contributed between 7 percent and 8 percent of traffic to a site. However, that
statistic doesn’t reflect the importance of finding a site initially with a
search engine. Frequently, once a site address is found, it is bookmarked by the
user or remembered so the engine isn’t used a second time for that particular
site. Search engines in that case serve as discovery tools.
Two years ago, Yahoo! and AltaVista, at that time
the two most popular search sites, went to a paid model that offered customers
guaranteed placement of search results in their indexes.
Such guaranteed placement can mean a lot of money
for search engines. Yahoo!, for example, charges $299.99 for its “site express”
service that guarantees evaluation of a site by human eyes, but does not
guarantee placement in its directory. That fee is re-applied every year for
continued placement.
AltaVista has a similar, less-costly plan: $39
per URL for a single site, and then a sliding scale for multiple URLs for a
six-month inclusion in its index. Google uses a different approach whereby
listings are determined by a relative rank of their popularity: If a site has a
large amount of other sites pointing to it as links, and also has text that
closely matches the query, then it ranks high in Google’s directory. Google
does not accept paid placement, but does accept advertising that will be
generated depending on specific search terms that are used by consumers.
On the other side of the firewall, search
technology also serves self-contained sites. News sites in particular can
benefit from search engines that send software “agents” that crawl through
their daily content and their archives. Most search engines will license their
technology to individual sites, and in fact rely on that as their primary
business model now that the advertising market for portals has declined.
Search technology can also be outsourced, as is
the case with Atomz Corp. Atomz search currently powers more than 50,000 sites
including CBS News, Palm, and Macromedia. With outsourced search, no software
needs to be installed on the customer’s Web server. The site is updated at
regular intervals and re-indexed remotely.
One Atomz user is Cincinnati.com, a six-year-old
Web site that houses a variety of media properties including The Cincinnati
Enquirer and The Cincinnati Post (combined daily, 197,399; Sunday, 310,673).
“We were using Netscape on our own servers …
it crashed constantly and even when it was running it was feature-poor and
high-maintenance,” said Jeffrey Tindall, a senior developer who helped craft
the original site.
The crashes resulted in lost data, requiring
lengthy re-indexing.
“We also tried Excite with much the same
results,” Tindall said. “We wanted to avoid administering it in-house if
possible, so our administrative staff could focus their efforts on development
rather than upkeep.”
Atomz, Tindall said, helped alleviate almost all
of the site’s administration challenges.
“In fact, when Gannett corporate decided to
pursue a solution, I strongly recommended Atomz to them.”
Gannett subsequently signed a two-year agreement
with Atomz to provide search tool software for all of Gannett’s newspaper Web
sites. Today Atomz is being used by a number of Gannett’s newspapers,
including The (Nashville) Tennessean and The Argus Leader in Sioux Falls, S.D.
According to Tindall, Atomz monitors inbound
traffic from the referring URLs very closely, and can thus monitor how many
referrals are being generated by various search engines.
“Search engine referrers contribute a very high
portion of our visitors, which is true of all newspaper sites. But we take great
pains to maximize this as much as possible,” Tindall said.
The Deseret News (daily, 67,000; Sunday, 70,000)
in Salt Lake City chose to keep its search engine in-house, integrated with its
archive system. Its site serves the Salt Lake City and Provo metro areas. The
daily has been using Convera’s RetrievalWare for its search engine since 1998.
“We looked through trade magazines, talked to
other newspapers, actually looked at the solicitations coming in, and compiled a
list of possible vendors,” said Dave Schneider, director of new media. “In
’98 and ’99 [application service providers] really weren’t the
considerations they became later. We had, and have, a good systems/IT department
and editorial-content projects are high priority with them. So going out to ASPs
for what we consider an editorial-content service isn’t likely to happen.”
RetrievalWare creates an inventory of all
enterprise assets, then enables users to search more than 200 document types on
file servers, in groupware systems, relational databases, document management
systems and Web servers. This access is managed by RetrievalWare’s
synchronizers, which recognize any changes system-wide and automatically update
the RetrievalWare index.
Search engines, both internal and external, are
clearly an important component to a successful Web site. Five years ago they
were often seen as an exciting option. Today, according to many experts, they’re
seen as a standard operating function of a well-constructed site.
