Rethinking the Microsoft Corp Search Cost Burden

July 31, 2009

I am waiting to give a talk. I have been flipping through hard copy newspapers and clicking around to see what is happening as I cool my heels. Three items caught may attention. Both the New York Times and the Wall Street Journal reported that the Yahoo deal is good for Yahoo. Maybe? What I think is that Microsoft assumed the cost burden of Yahoo’s search operation. Since my analysis of Yahoo’s search costs in 2006, I have gently reminded folks that Yahoo had a growing cost problem and its various management teams could not do much about these costs. So Yahoo focused on other matters and few people  brought the focus back on infrastructure, staff, and duplicative search systems.

Now Microsoft has assumed this burden.

I scanned John Gruber’s “Microsoft’s Long, Slow Decline”, and I noted this comment:

Microsoft remains a very profitable company. However, they have never before reported year-over-year declines like this, nor fallen so short of projected earnings. Something is awry.

Dead on. What is missing is thinking about the challenge Microsoft has in search. My thoughts are:

First, Microsoft has to make headway with its $1.2 billion investment in enterprise search. I think the uptake of SharePoint will produce some slam dunk sales. But competitors in the enterprise search sector know that SharePoint is big and fuzzy, and many Microsoft centric companies have here and now problems. I think there is a real possibility for Microsoft to cut the price of Fast ESP or just give it away if the client buys enough CALs for other Microsoft products. What I wonder is, “How will Microsoft deal with the punishing engineering costs a complex content processing and search system imposes during its first six to 18 months in a licensee’s organization. Microsoft partners may know SharePoint but I don’t think many know Fast ESP. Then there is the R&D cost to keep pace with competitors in search, content processing, and the broader field of findability. Toss in business intelligence and you have a heck of a cost anchor.

Second, may get access to Yahoo’s Ford F 150 filled with software. But integrating Yahoo technology with Microsoft technology is going to be expensive. There are other costs as well; for example, Microsoft bought Powerset and some legacy technology from Xerox Parc. Layer on the need for backward compatibility and you have another series of cost black holes.

Finally, there are the many different search technologies that Microsoft has deployed and must presumably rationalize. Fast ESP has a better SQL query method than SQL Server. Will customers get both SQL and Fast ESP or will there be more product variants. Regardless of the path forward, there are increased costs.

Now back to Mr. Gruber’s point: a long, slow decline requires innovation and marketing of the highest order. I think the cost burden imposed by search will be difficult for Microsoft to control. Furthermore, I hypothesize:

  • Google will become more aggressive in the enterprise sector. Inflicting several dozen wounds may be enough to addle Microsoft and erode its profitability in one or two core product lines
  • Google’s share of the Web search market may erode but not overnight’. The Googlers won’t stand still and the deal with Yahoo strikes me as chasing the Google of 2006, not the Google of 2009.
  • Of the more than 200 competitors in enterprise search and content processing, I am confident that a half dozen or more will find ways to suck cash from Microsoft’s key accounts because increasingly people want solutions, not erector sets.

In short, Microsoft’s cost burdens come at a difficult time for the company. Microsoft and Yahoo managers have their work cut out for them.

Stephen Arnold, July 31, 2009

New York Times: Two Indexing Methods

July 31, 2009

Teragram, a unit of SAS SAP, provides software that automatically indexes content for the New York Times’s Web site. I saw a tweet on my Overflight service that pointed out that the newspaper uses humans to create the New York Times Index, a more traditional index. You can find the tweet here. If true, why won’t the Teragram system do both jobs? When financial corsets get yanked tighter, something has to give. My thought is that if the tweet is accurate, is redundancy cost effective? An indication that neither works particularly well? There is a political logic, not a financial logic, at work?

Stephen Arnold, July 31, 2009

SurfRay Management Change

July 31, 2009

Short honk: Earth Stream ran a story called “New Management Team for Search Engine Factory, SurfRay”. My last bump into SurfRay was its appearance on a Xerox Web page. I think the new chief executive officer for SurfRay is Soren Pallensen, a former VC. He has had some Vulcan mind meld with Microsoft. The company seems to be chugging along. I am not sure what a “search engine factory” is, but it is an interesting way to present the firm’s business. I wrote a monograph about Japan’s beavering away at making data. I called that study “Information Factory”. Useful word for me in 1992 or so.

Stephen Arnold, July 31, 2009

Microsoft Business Intelligence on the Horizon

July 31, 2009

Coming “someday” from Microsoft: Business intelligence. Computerworld reports that BI and analytics will be included in Windows Azure, internet-scale cloud services platform, but it’ll be a wait. Microsoft won’t have it ready until 2013 or later. See the Computer World story “Microsoft to Bring BI to the Cloud.”

Microsoft already has three major corporations set to use PerformancePoint, a business performance management software, and the company’s “self-service” BI component is set to release in Word and Excel next year. Generally common functions of business intelligence technologies are data mining and analytics, performance management, benchmarking, etc. If Microsoft is working these functions into MS Office for the enterprise organizations, I can see that being an important tool. They’re working on it, but it’s not here yet.

Jessica Bratcher, July 31, 2009

Google and Games

July 31, 2009

Editor’s Note: Stephen E. Arnold delivered this talk at the games conference held in Louisville, Kentucky, on July 31, 2009.


I want to provide a quick review of Google’s approach to games and gaming. I want to show some screenshots that make clear that simple and more complex games are available with more games becoming available everyday. I then want to describe how Google views the notional topic of games for users of computing devices. I want to conclude by putting my remarks in a timeline that carries the subject of Google and games to the year 2015. The date is not arbitrary because Google works in chunks of three to five years. Google’s approach to games won’t change too much in the next 16 years, but the scope, application, and monetization of games and game technology will. My conclusion may surprise you. By 2015, Google may be one of the leading game platforms with a broad range of products and services that use gaming technology in interesting, revenue-boosting ways.

Google People

Most users of Google’s systems don’t know individual Google engineers by name. The company has nearly 19,000 professionals on staff and about two thirds of them are engineers, computer scientists, mathematicians, or physicists.

Quite a few people today play games. The devices range from the high-end, state-of-the-art platforms like the Microsoft Xbox, the Nintendo Wii, and the Sony PS3 to the grandma friendly games on Yahoo or mobile phones. One store in rural Kentucky where I live sells a $5 keychain with a simple game for bored adults and affluent seven year olds to play when stuck behind a horse in Harrod’s Creek. New platforms bring new people to the “game party”. Add in the influx of mobile device users, and the stage is set for a “game revolution”.

I want to highlight two Google engineers and mention some of their work to give you an idea about how deep game technology has been embedded at Google.

Steve Lawrence, an Australian, is a gamer. In addition to authoring technical articles that have been referenced more than 5,000 times, he is the author of Game sports betting markets, Sandip Debnath, David M. Pennock, C. Lee Giles, Steve Lawrence, ACM Conference on Electronic Commerce, 2003, pp. 258-259. Dr. Lawrence is one of Google’s most prolific inventors, and his technical skill has influenced inventions ranging from user interface (US7272601) to personalized network searching (US2008/0215553).

Consider Ross Koningstein. He was a graduate student at Stanford’s Aerospace Robotics Laboratory when he contributed to the development of Chuck Yeager’s Advanced Flight Trainer II and Car & Driver Text Track. Google’s advertising system is based on bids. The methods used are dependent on calculations used in traditional games like horse racing. Mr. Koningstein has been working to bring game-like features and interface elements to Google’s advertising management system. The idea is that a person at an ad agency can use a game-like method to model what certain types of ads and a specified amount of ad money will generate for the advertiser. Dr. Koningstein wants to make modeling ad spends and ad management more of an interactive game experience. You can read more about his approach in Google patent documents US20050228797 , US20050096979, US20060224444, US20060224447, US20050114198, and several others.

You get the idea. Dr. Koningstein is not dabbling; he is inventing systems and methods that have roots deep in the interactive game experience.

Keep in mind that other Googlers have equally deep roots in gaming.

Google Technology

I want to do a quick fly through of Google technology and provide you with some screenshots of applications that are available today.

First, Google is a platform, and it offers a range of software development kits, application programming interfaces, and “sandbox” toys. The idea is that a developer with online basic programming skills can use the Google platform. At the other end of the spectrum, a professional developer or a company focused on game development can create applications that run on the Google platform.

Keep in mind that the platform is a one way street. This means that you can put code into Google but it can be difficult to repurpose that code for another platform. Therefore, the best way to think about using the platform for a game or some other application is to create a game for a platform such as the iPhone and then recycle the graphics and other useful bits for the Google Android platform. You will learn in a few moments that this recycling approach may be the path forward for the next few years.

Second, Google tried to cut a deal with Yahoo for online games in the 2005 to 2006 period. My sources suggested that the tie up did not make sense. Google on the surface has not played a major role in commercial game development. In fact, the model today is influenced by Google’s need to be perceived as an open source company that is not a monopoly. The point is that if you get into the Google development space with a game, you will be operating in a competitive but open environment. At some point in the future, Google could change its approach, but there is little downside for experimenting with the Google platform. New tools such as Google Wave will be forthcoming. Coupled with Chrome (Google’s virtual machine and container system) and Android (a chunk of the Google operating system), Google now offers a usable platform for game development.

Third, the forthcoming Google Wave technology (a component of Google’s dataspace initiative) appears to be a significant new component of the Google service suite. The idea behind Wave is a plastic bag. Put carrots or small parts in the bag, and you can manage them. Wave allows a developer to create a space – a digital freezer bag. Activities can take place within the bag. Wave makes it possible to have the objects in the digital bag interact. The idea is to make it possible to create new types of social interactions with information objects. The most important feature is that states can be saved. It is possible to slice and dice the objects and the interactions by time. If you think about this functionality, new opportunities for games and game like experiences can be built on these multidimensional functions. Let me give one example: lectures, lab experiments, and student interaction. I think certain types of instructional constructs where traditional game like features and time can be combined with social interaction in useful ways.

Keep in mind that Google’s technologies pivot on programming languages that many developers know. These include JavaScript, php, python, and Java, among others. The point is that you can hit the ground running with Google’s sample code and your favorite programming language. At this time, there’s no fee, just a Google registration.


In the time I have remaining, let me look at two different doors that are now opening. Each door is a metaphor for a way to exploit Google as a game platform as well as a platform for building game like applications. In short, I want to suggest that the notion of a game must be viewed in two ways.


Google has completed much of its next-generation computing platform. Consumer applications such as games are now a potential growth area with Android, Chrome, and Wave as enablers.

First, I think it is wise to look at the Google platform as one in which 10 percent of one’s development effort should be invested. The reason is that Android and Wave are immature or not yet built out. Therefore, the idea is to take a simple game idea or an existing game feature and recycle it for Google. Within the next two to five years, additional development resources should be directed at the Google platform. Five or six years out, development for only the Google platform is likely to be possible.

The reason for this is that the game platform and game device market does not change as rapidly as some believe. The high end, dedicated game devices will persist in the market. In the short term, the Apple iPhone is a more viable mobile game platform. However, over time, the shift to mobile computing and cloud computing will change the equation.

In short, learn and recycle. Don’t bet the farm on Google.

Second, I think it is important to recognize that Google moves in small, incremental steps. The company does this in order to avoid alerting competitors to its broader strategy and to minimize antitrust actions. Nevertheless, you should plan on allocating your time based on how the Google market shapes up. This means that delay in learning how to code for Google is a bad idea. Among the technologies to learn are SketchUp (Google’s drawing program), Android (the visible part of the Google operating system), Wave (collaborative spaces), and Google Apps and OneBox APIs. These functions are, at a minimum, the way in which to obtain the Googley expertise you need.

In closing, let me make three observations about Google, games, and the game like applications that will be the norm in computing in the future:

Read more

AOL May Presage the Future Trajectory of Yahoo

July 30, 2009

Larry Dignan reported that AOL revenue slid 24 percent in its second quarter. With a Googler at the helm, will AOL be able to reverse its slide and then get back on the growth highway? Interesting question, but one that I find more intriguing is, “Does the performance of AOL provide a glimpse of Yahoo’s future?” Yahoo has to come to grips with its new role as a clone of AOL. Yahoo and AOL are both portals. Both companies depend on advertising. New services are rushing past AOL and Yahoo. By the time the companies react, the opportunities have shifted. What about a merger of AOL and Yahoo? Whatever angles Yahoo explores, the company has to make some headway in today’s choppy financial waters. Yahoo has a six pack of search systems. Maybe even a case of search systems. Managing these services and pumping up revenues will go a long way to show me that Yahoo is not an AOL.

Stephen Arnold, July 30, 2009

Dot Net Clean Up Utility

July 30, 2009

Microsoft SharePoint, Fast ESP, and Dot Net news has been in the hammock resting. The goslings discovered a utility that may provide some sun tan lotion when the Dot Net sun gets too intense. You can download the Dot Net Framework Clean Up Tool 6.0.3790 from files. Use freeware at your own risk. The goslings live dangerously; you might not want our brand of excitement. Have you even seen the creatures that live in our pond refreshed with mine run off?

The tool will remove selected versions of our pal Dot Net. Our thought is that you should run some tests on a Dot Net installation to get a feel for how this buggy handles the curves. When you run the utility, you can kiss shared files and registry keys used by other versions of Dot Net if you have multiple versions installed. Back ups are usually a good use of time and disc space.

Stephen Arnold, July 29, 2009

Newspaper Woes in Harrods Creek

July 30, 2009

I subscribe to four hard copy newspapers. I don’t think too much about the delivery of the newspaper because I am not too far from a large, confused supermarket. This outfit gets four or five copies of the newspapers so my copies turn up. Yesterday, I did not receive a USA Today. This is a Gannett publication known in Harrods Creek as McPaper. Today I got the USA Today but it was soaked. That’s better than the New York Times’s dead tree offering. The New York Times paper did not arrive. A tip of the hat to the floundering Courier Journal (where I used to work) and the zippy new Wall Street Journal. Both arrived. I resubscribed once the WSJ stopped sending me spam to become a subscriber. Net net of this is that when you have an old geezer who pays for hard copies, maybe the newspaper folks should deliver the paper and get it there in a readable form. Must be really tough in today’s financial climate. No wonder there is some reluctance on the part of busy people to go with old news that arrives wet or not at all.

Stephen Arnold, July 29, 2009

Books Do Not Have Lift

July 30, 2009

Mike Hendrickson posted his mid-2009 State of the Computer post on O’Reilly Radar at He’s tracking sales of computer industry books, and sales are going down, down, down in a possible reflection of the economy. On the other hand, if you’re looking for technical information and how-to’s for systems like Windows Vista, Mac Programming, or SharePoint, wouldn’t you start on the Internet? We would. There’s so much good real-time data out there to crunch (if you’re willing to put in a little extra work ). That combined with the exponential growth of the Internet is going to put more nails into traditional media’s coffin. Book readers are mostly oldsters and the younger folks seem to relate to fresher media.

Jessica Bratcher, July 30, 2009

Future Microsoft SharePoint ESP Features Revealed

July 30, 2009

Stephen Bell wrote “Microsoft Fast Search to Mate with Social Networking.” The title brought a grin to this addled goose’s bill. I like the idea of Fast ESP “mating” with a social network. The metaphor brings a number of images to mind and invites a wide range of double entrendre. Maybe the ComputerWorld editors were tickling the goose’s funny bone?

The story asserts that a fellow named Steve Letford, a Microsoft wizard in New Zealand, revealed at a SharePoint conference that Fast ESP will be “mated with SharePoint in a number of ways.” What? I thought the headline said Fast ESP, and now it is SharePoint.

Which is it? Fast ESP gets social? SharePoint gets social? SharePoint is a collaborative content management platform with search available upon installation. I guess I don’t understand what is happening.

Mr. Bell reports:

In ordinary office productivity work, the Fast search tool will give more meaningful results than an ordinary search engine on unstructured data, because of its ability to recognize the semantics of such elements as a date or an address, Letford says. Its pipeline architecture enables different inquiries to be made concurrently on a stream of documents as they pass down the pipe, meaning multifaceted queries can be executed more efficiently. Documents retrieved by the search will be able to be further interrogated interactively and the documents and their interrelationships presented in graphical format.

Now that would be something. I need “meaningful results”. I am not sure if an address is a “semantics” component but entity extraction or a tagged field can be interpreted under the fuzzy wuzzy banner of semantics. Precision is not part of the toolkit of a marketer.

The article references a now long-i-the-tooth IDC study about the cost of looking for information. In the present economic climate, my hunch is that the organizations with such high costs might be road kill, but that type of thinking does not slow the rush of the Microsoft argument Mr. Bell reports.

The killer comment for me was:

Customers of the enterprise search tool will have to buy a specific license and purchase a dedicated search server. Enterprise search is not something you install on your PC or laptop, he [Letford] says. A third common use for high-end search capabilities, Microsoft predicts, will be in monetizing website offerings by guiding users to content for which they will be willing to pay, and bringing up relevant advertising messages. Letford confirmed, in answer to a question from the audience, that Microsoft’s public search-engine, Bing, will evolve to use FAST-style technology.

What does this statement mean for Powerset, the Xerox Parc plumbing, and the glue code created to create the systems? More evolutionary change? Sounds like a major overhaul to me. If this statement is accurate (keep in mind that it comes far from the search think tanks in Redmond and Oslo, Microsoft has big plans for Fast ESP. SharePoint, and

Let’s do a flashback to the pre-Fast ESP sell out and before the Norwegian investigators looked into the alleged fraud at Fast Search & Transfer.

Read more

Next Page »