A Newspaper-Bred Yahoo! Is It Possible?

By: Steve Outing Lots of companies in the Internet business want to be a "portal," though few have achieved that exalted status to date. America Online is certainly the most successful portal (a.k.a. gateway) to the Internet for its millions of customers. Yahoo! is probably doing the best job of being a portal site, chased by Excite, Infoseek and Lycos. CNET's Snap! project is a portal strategy. And Microsoft is developing a portal strategy that it calls Start. Netscape has similar ambitions to be the primary place where Internet users go to find their way around the Web, with its NetCenter.

Notice that none of the major players mentioned above are traditional news media companies. As reported in this column in recent weeks, Southam Newspapers in Canada has adopted a portal strategy for its Canada.com Web news network, accomplished by partnering with a search engine company, Inktomi (the company behind the Wired HotBot search engine). Newspaper new media consortium New Century Network, before its demise earlier this year, was working on developing a portal strategy for use by its affiliate publishers -- but that too was likely to have been in partnership with an existing Internet navigation company.

No "old media" company is yet developing an ambitious portal strategy on its own. None has taken a look at Yahoo! and said, "We will do it better." Yahoo! is without doubt king of the hill now, and newspapers can do nothing but drool over its traffic. An estimated 20 million people generate close to 100 million page views a day using Yahoo! to find stuff on the Web.

But maybe it doesn't have to be this way. Maybe, just maybe, a newspaper company could "out-Yahoo! Yahoo!" by itself, without partnering with one of those companies mentioned in the first paragraph of this column.

That's the vision of Peter Rinearson, president of Alki Software in Seattle, Washington, and a former newspaper journalist who once won a Pulitzer Prize before escaping to the world of high technology. It sounds preposterous, but Rinearson says he has technology that can create a better Yahoo! without the gargantuan expense that common wisdom would suggest is necessary to create a Web portal that would compete with Yahoo! Still "a newspaperman at heart," Rinearson hopes to convince a major newspaper company to use the technology to create an Internet portal that will give it a way to keep people on its own sites and stop them from flocking to Yahoo! whenever they need to find something on the Web.

My friend, Bill

Alki is a small company, best known for its add-on software for Microsoft Word. (Rinearson has the ear of Microsoft's Bill Gates, having helped Gates write the billionaire's book, "The Road Ahead," as well as Gates' column for the New York Times News Service.) Since 1995, Rinearson has been steering his software firm on an Internet path, and has developed a technology called InType, which can be used by publishers to create various Web services -- restaurant guides, city directories, classified services, even news. It is a database publishing technology that can be used to create the kind of services that the newspaper industry currently gets from companies like Zip2/Citysearch. And it could be used to help a publisher get rid of old legacy editing systems and replace them with a "write once-publish many" database publishing system to collaboratively operate a newspaper's print and digital operations. (Knight Ridder has announced that it is working with Alki to use the InType technology to do just that. KR uses Zip2 as its vendor for its online city guide ventures.)

Rinearson's company had been working on this technology when NCN folded, and NCN's uncompleted portal strategy became public. He says that that news made him realize that the newspaper industry -- desperately in need of a portal strategy, even if many in the industry didn't realize it yet -- could use the developing InType technology to create an alternative Yahoo!

Yahoo! is, of course, more than a search engine that indexes the contents of the Web. Its Silicon Valley-based staff also exercises editorial judgment, finding and cataloging Web sites for its directory structure. You can go down a Yahoo! directory tree and find information on the Web based on your topic of interest -- e.g., show me sites about skiing for those with physical disabilities. On pure search engines, you can simply search using those words and get a long list of possibly relevant sites returned. The search AND directory approach is largely what makes Yahoo! so popular and useful.

Publishers who adopt only a search engine approach and call it a "portal strategy" are missing a key element, Rinearson says. To outdo Yahoo!, you need to mimic it to some extent. And that means creating a directory structure. Yahoo! mostly uses humans to categorize Web sites and pages it finds and put them into the correct directory branches -- an expensive endeavor when you consider the amount of information on the Web. But the InType approach uses technology to largely automate the process.

(This is the next wave in Internet navigation. Yahoo! competitor Lycos just last week announced that it has developed an automated search directory. Where Yahoo! has its employees surfing for new sites, Lycos will use automated processes to gather information about new sites, according to the company's president.)

What the InType concept is capable of, says Rinearson, is sending spiders out to new Web sites and coming back with a base level of detail -- what the topic is about from reading META tags, when the site was last updated, the name of the Webmaster, etc. But clearly no automated process can do as thorough a job as a human who reviews a site and determines, "That belongs in categories A, D, F and P." So the next step is to "use the community of Webmasters" to fill in the missing details, explains Rinearson. Instead of having that staff of hundreds surfing the Web and cataloging sites, have the sites' owners fill in the information for you, and automatically categorize the data that they return and infuse it into the database.

The ideal implementation of this idea is to have the system determine basic information about a site, then assemble it into a form that is sent to the site's Webmaster. The form says something like, "Here's what we know about your site, which will be included in our directory service. Please correct anything we have wrong, and tell us more about your site by filling in the questions on this form." When the Webmaster returns the information, the system automatically updates the database record for the site and parses the completed data set into appropriate categories in the directory service. This is how you create a Yahoo! competitor without the need for millions of dollars to pay humans to do the job.

Rinearson thinks that a large media company, employing his technology, could make a run against Yahoo! by utilizing the Webmasters themselves to provide information rather than a paid staff. It's thus not so daunting to compete with Yahoo! as it might have appeared in the past.

Still, no amount of automation can fully match what human intelligence can perform. So Rinearson envisions a news organization's staff getting involved over time. A newspaper company could have its science writer shepherd the science components of the directory service, utilizing the automated procedures to do the bulk of the work, but adding his own comments to site reviews, for example. In this regard, an established news organization has a distinct advantage over the cyber companies because of its existing brain trust (a.k.a. editorial staff). The more a publisher's staff gets involved in improving the directory service, the more it improves over the product produced by the automated system.

The next big thing

Rinearson is a proponent of the concept of "community publishing," an idea that the newspaper industry has been slow to adopt but which he thinks is the most significant opportunity in the Internet space in the coming years. It's the community publishing concept -- getting the citizenry to provide the information about their activities and organizations and providing them with simple tools to get their information into your database instead of relying on staff to collect and process the information -- that provides a means to compete with the Yahoo!s of the Internet world, he says.

Community publishing concepts are the basis for building the online community guides that can be created with the InType technology. Restaurants, for example, can create fancy directory listings including everything about their businesses including menu items simply by filling out a Web form. Alki has a demonstration of that on a Web site called A Table in Seattle.

Should some newspaper company play this scenario out to its full potential, what it could develop using InType would be not only a replacement for Yahoo!, but also the ultimate guide to its local community. Local information directories likewise would be created by the data input of the community itself -- from movie houses to schools to government agencies, etc. (Alki's technology thus is in competition with that of Koz.) In that local respect, a local-based news company employing technology to collect local data would have a tremendous advantage over a company like Yahoo! that is trying -- from the outside -- to create local community guides.

Rinearson says he wants very much to see the newspaper industry embrace this concept. "For me, it would be gratifying to be involved in the reinvention of the newspaper industry," he says. He thinks "there's still time for the industry to do this," but in two more years it may be too late; by then the Internet portal space will be locked up by others if publishers don't act. Will the industry act? "Newspapers have great potentials to be portals. But they're not moving fast enough," he says.

Alki is not betting the farm on newspapers, however, as any number of industries could embrace its technology and become the Internet portal for their communities. He says he's not talking to telephone companies in the U.S. about this, but he is in discussions with them overseas, for example.

His hope is to find a large newspaper chain to embrace the concept, rather than try to deploy it across the whole industry. He's not interested in doing 150 deals with newspapers. A full implementation of the technology probably will have to come with a company that's not already tied up with Zip2/CitySearch, which has pretty much of a lock on the newspaper city guide business. Rinearson says that he doesn't want to go up against Zip2/CitySearch, yet he does provide an alternative choice in a space that is basically dominated by that one vendor.

Perhaps most importantly, Rinearson believes that the newspaper industry should do this on its own. Partnering with a search engine company ultimately does more to benefit that company than it does for the newspaper, even if in the short term it brings in a few million dollars in new revenue for the publisher, he says. Newspapers should be positioning themselves to be the portal to the Internet for their customers, not help another company achieve that goal.

(Alki has a new Web service, called BabyNamer, that was built using some of the technology in InType. Take a look at it to get an idea of InType's capabilities.)

Contact: Peter Rinearson, peterr@alki.com


To receive an e-mail reminder each time we publish a new Stop The Presses! column, sign up here.

Archive of columns
This column is written by Steve Outing exclusively for Editor & Publisher Interactive three days a week. News, tips, and other communications may be sent to Mr. Outing at steve@planetarynews.com

The views expressed in the above column do not necessarily represent the views of the Editor & Publisher company


No comments on this item Please log in to comment by clicking here

Scroll the Latest Job Opportunities From The Media Job Board