Microsoft and Yahoo: What's Scale Got to Do with It?

By Darryl K. Taft  |  Posted 2009-07-31 Print this article Print

Microsoft's search technology deal with Yahoo is about improving the software giant's share of the search market by adding the strength of Yahoo's search capabilities, but it is also about improving the Microsoft Bing technology and the overall experience for users and advertisers. In short, the deal is about scale.

In the Microsoft-Yahoo deal, scale is a weighty matter.

Microsoft's search technology deal with Yahoo is about improving the software giant's share of the search market by adding the strength of Yahoo's search capabilities, but it is also about improving the Microsoft Bing technology and the overall experience for users and advertisers. In short, the deal is about scale.

"Scale" is the keyword Microsoft CEO Steve Ballmer and other executives have continually used when talking about the opportunity Microsoft's deal to license Yahoo's search technologies brings to the company. Most recently, Ballmer and company discussed search and scale at the Microsoft Financial Analyst Meeting at the company's Redmond, Wash., headquarters on July 30.

"First, let me start on the strategy side," Ballmer said. "Microsoft, we get a chance by putting together our scale and Yahoo's scale in the United States and in some other countries where Yahoo has scale. ... But scale is actually a tool for product improvement. I'm not talking about money improvement. I will get there in a minute. It's a tool for product improvement. The more queries you see, the more you can tune your product. The more scale you have, the more advertisers advertise on your system and the more relevant they make their ads for your users. It is not about money. It is actually about relevance because the advertising is part of the actual user experience."

Describing the process a bit further, Ballmer said:

"If there's a thirteenth-century Italian art dealer in the world, they're going to buy keywords on Google, and they may not buy them today on Yahoo and Microsoft. Yet, if you're purchasing for thirteenth-century Italian art, you don't probably want a book on Italy, which might be at the most relevant ad one place or more relevant somewhere else. So, this partnership is about product improvement, customer understanding and scale for advertising relevance."

Some observers have asked whether Microsoft, by pooling its search assets with Yahoo's, is actually improving the condition of its search capability or simply adding share. Well, it is doing both. But, because of issues of scale that Ballmer and Qi Lu, president of Microsoft's Online Services Division, laid out, Microsoft is definitely improving its lot in search with its Yahoo deal.

Data about Web pages is stored in an index database for use in later queries. The purpose of storing an index is to optimize speed and performance in finding relevant documents for a search query. Without an index, the search engine would scan every document in a vast body of knowledge, which would require considerable time and computing power. And the usefulness of a search engine depends on the relevance of the result set it gives back. With increased use-in this case, a bigger pool of users making more and more queries-the system improves as more and more relevant results are indexed and the search process is refined.

Somewhat in line with this, Microsoft recently hailed a game created by Microsoft Research called Page Hunt that is designed to help improve the search accuracy of Bing as more people play the game

According to the press release detailing the Microsoft/Yahoo search deal, "Ballmer said the agreement will provide Microsoft's search engine, Bing, the scale necessary to more effectively compete, attracting more users and advertisers, which in turn will lead to more relevant ads and search results."

Lu said many search engine features are sensitive to the overall usage of the engine-again, that ever important pool of users. "For example, if you go to a search engine and type in 'query,' there is a drop-down query suggestion," he said. "That's a very simple algorithm behind that. How good the suggestion is depends on how many people are using the search engine. By aggregating scales, we will be able to get immediate benefit on consumer experience on a number of fronts. And it is important to call that out."

Lu also said Microsoft has a strategic framework with different layers of activities, including a base layer that he called foundations. "These are large-scale R&D operations that you have to do just to be in the game," he said. "That includes crawling the whole Web, building a huge index-tens of billions of Web pages-and keeping the index fresh. You need to keep it super fresh and do all the R&D, the scientific discoveries to understand user intent-which is generate a distribution of possible meanings, possible needs or interests, and then match that against what we work with out of the Web today."

Another layer is differentiation, Lu said. "So we have to produce a search experience that is materially different and is competing at the same time," he said. "So our focus is on a few things, focusing on user intent and understanding competition, what the users are trying to accomplish, and help users to complete the tasks better and make more informed decisions faster."

Moreover, despite taking the audience of investment analysts deep into the innards of the search arena, Lu spelled out the bottom line. Said Lu:

"The important challenge we're facing is because this is a scale-driven business. Even if you have one user, you have to crawl the whole Web doing R&D. So when you operate at subscale, you will incur a certain amount of financial losses. It needs to get to certain scales. When you are at scale, this can be a hugely profitable business. But at subscale, there is a financial cost to be incurred."

The Yahoo deal moves the scale.

Indeed, Lu added, "Scaling is key because the Web grows every second." 


Darryl K. Taft covers the development tools and developer-related issues beat from his office in Baltimore. He has more than 10 years of experience in the business and is always looking for the next scoop. Taft is a member of the Association for Computing Machinery (ACM) and was named 'one of the most active middleware reporters in the world' by The Middleware Co. He also has his own card in the 'Who's Who in Enterprise Java' deck.

Submit a Comment

Loading Comments...
Manage your Newsletters: Login   Register My Newsletters

Rocket Fuel