In a competitive parry against Google, MSN is beta testing a service that aggregates news headlines and stories from multiple sources and arranges them based on their likely relevance to the user. The service hints at how future MSN sites and services might enable a more personalized experience for Internet users.

Competing with Google

The MSN NewsBot, which is being beta tested on MSN sites in France, Italy, Spain, and the United Kingdom, gathers stories from more than 4,000 sources, then arranges them in categories such as "World," "Business," and "Entertainment." The order of headlines within each category is determined by factors such as how recently the story was posted, the number of users who have previously accessed it, and the number of news sources covering that same story. If users have signed in with a Passport ID, MSN NewsBot can also factor other personal information, such as a user's postal code or a particular news source that has been frequently selected, into the ordering algorithm, according to the NewsBot FAQ.

The MSN NewsBot closely resembles the Google News search tool introduced in 2002. Competition between Google and Microsoft has intensified since spring 2003, when Microsoft declared its intention to build a better algorithmic search engine than Google. In turn, Google has begun venturing into Microsoft's core territory—software—by offering a Google Toolbar, which hooks in to Internet Explorer and allows users to search Google without visiting the site. The company is also testing a Google Deskbar, which allows users to search Google from within any Windows application without opening a Web browser.

Focus: Personalization

The NewsBot signals a new focus for MSN: creating personalized sites and services based on information gathered from users.

This concept has surfaced in several ways since Apr. 2003, when MSN said that personalization would be a key feature of its search engine. Joshua Petersen, who helped develop the recommendation engine for Amazon.com—a program that remembers which products a customer previously read about or purchased, and recommends new products based on that behavior—is now part of the MSN Search team, and helped develop the MSN NewsBot.

In November, MSN hired Paul Ryan, the former chief technology officer of Overture, the Web's leading provider of paid-search technology (search results in which advertisers pay for specific terms). Ryan has been put in charge of the cryptically named Monetization business unit, which MSN Vice President Yusuf Mehdi describes as taking the paid-search business model and applying it to other areas, such as music downloads. The paid-search business model is based on the fact that search engine users are self-selecting—somebody searching for "Ford" is more likely to be interested in advertisements for a Ford dealership than the average Web site visitor. This allows paid-search providers to offer better conversion rates (that is, clicks that actually lead to sales) compared with other types of Web-based advertising, letting them charge higher rates to advertisers.

MSN could use information collected about Web site visitors to create self-selecting groups for other types of services. For instance, a user who frequently clicked on news stories about a certain musical group might be offered an early opportunity to download new music from that artist; then, based on the user's Passport information, MSN could offer concert tickets when the artist passes through that user's region.

Google has also been active in this area: in Oct. 2003, it acquired Kaltix, a start-up company that is working on technology for personalized and context-sensitive search.

An English-language version of the MSN NewsBot, along with links to versions in other languages, is at uk.newsbot.msn.com/worldwide.aspx.