Data Growth
June 4, 2009
IDC published a new white paper. You can get information about the document here. Scanning the abstract and report a factoid warranted snagging; to wit:
IDC found that companies are increasing their file-based storage by 40% to 120% a year
This means that if an organization has one terabyte of data today, storage will be needed for 1.4 to 3.0 more data next year. I heard an executive from a transportation firm say its data growth was 4X per year. What’s clear is that digital data breeds like gerbils. Findability is simple, according to the azure chip consultants. Wanna bet?
Stephen Arnold, June 5, 2009
Exalead’s Vision for Enterprise Search
June 4, 2009
I had a long conversation with Exalead’s director of marketing, Eric Rogge. We covered a number of topics, but one of his comments seemed particularly prescient. Let me summarize my understanding of his view of the evolution of search and offer several comments.
First, Exalead is a company that provides a high performance content processing system. I profiled the company in the Enterprise Search Report, Beyond Search, and Successful Enterprise Search Management. Furthermore, I use the company’s search system for my intelligence service Overflight, which you can explore on the ArnoldIT.com Web site. Although I am no expert, I do know quite a bit about Exalead and how it enables my competitive intelligence work.
Second, let me summarize my understanding of Mr. Rogge’s view of what search and content processing may be in the next six to 12 months. The phrase that resonated with me was, “Search Based Applications.” The idea, as I understand it, is to put search and content processing into a work process. The “finding” function meshes with specific tasks, enables them, and reduces the “friction” that makes information such an expensive, frustrating experience.
Mr. Rogge mentioned several examples of Exalead’s search base applications approach. The company has a call center implementation and an online advertising implementation. He also described a talent management solution that combines search with traditional booking agency operations. The system manipulates image portfolios and allows the agency to eliminate steps and the paper that once was required.
The company’s rich media system handles digital asset management, an area of increasing importance. Keeping track of rich media objects in digital form requires an high-speed, easy-to-use system. Staff using a digital asset management system have quite different needs and skill levels. Due to the fast pace of most media companies, training is not possible. A photographer and a copyright specialist have to be able to use the system out of the box.
But the most interesting implementation of the SBA architecture was the company’s integration of the Exalead methods into a global logistics company. The information required to tell a client where a shipment is and when it will arrive. The Exalead system handles 5GB of structured data to track up to 1M shipments daily. Those using the system have a search box, topics and clients a click away, and automated reports that contain the most recent information. Updating of the information occurs multiple times each hour.
Finally, my view of his vision is quite positive. I know from my research that most people are not interested in search. What matters is getting the information required to perform a task. The notion of a search box that provides a way for the user to key a word or two and get an answer is desirable. But in most organizations, users of systems want the information to be “there”. That’s the reason that lists of topics or client names are important. After all, if a person looks up a particular item or entity several times a day, the system should just display that hot link. The notion of Web pages or displays that contain the results of a standing query is powerful. Users understand clicking on a link and seeing a “report” that mashes up information from various sources.
Exalead is winning enterprise deals in the US and Europe. My hunch is that the notion of the SBA will be one that makes intuitive sense to commercial enterprises, government agencies, and not-for-profit organizations. More important, the Exalead system works.
Stephen Arnold, June 5, 2009
Google Data Centers
June 4, 2009
Royal Pingdom revealed what I have been reluctant to put in my Google monographs. Since the cats are out of the bag, you will find the information in “Map of All Google Data Center Locations” here a useful clip and store document. The total reported is about 31, but the number is close enough for horseshoes. The Royal Pingdom folks include a link to Google’s data center FAQ, which is interesting as well. You can find that document here. Keep in mind that Microsoft has trimmed its spending for data centers at a time when the GOOG continues to build, upgrade, and reengineer.
Stephen Arnold, June 1, 2009
Microsoft Health: A New Thrust
June 4, 2009
Shift your attention from Bing.com to a sector that is a must-win for Microsoft. Ina Fried reported here that Microsoft acquired Rosetta Biosoftware from the struggling pharmaceutical company, Merck. Rosetta Biosoftware is a unit of of Rosetta Inpharmatics. Based in Seattle, the 300 person firm had been hit with cutbacks due to the financial climate. The software unit, which had about 60 employees, was expected to keep it lights on. According to Ms. Fried’s “Microsoft Buys Merck Unit in Life Sciences Push” here,
Microsoft, which has a separate Amalga product family for hospitals, announced in April that it would offer Microsoft Amalga Life Sciences as an effort to help in the drug research software arena. The tools are designed to help manage and analyze the large amounts of data gathered in the process of designing new drugs.
What’s Rosetta Biosoftware’s business? According to a profile here, the company
develops informatics solutions and provides services that enable research organizations to efficiently and effectively conduct life-saving discoveries and develop drugs.
Microsoft’s Amalga, according to Microsoft here, the company
develops its own powerful health solutions, such as Amalga and HealthVault. Together, Microsoft and its industry partners are working to advance a vision of unifying health information and making it more readily available, ensuring the best quality of life and affordable care for everyone.
Looks to me as if the dust up between Microsoft and Google in the health sector is likely to become more intense.
Stephen Arnold, June 3, 2009
The Value of Advertising Search
June 3, 2009
Over the years, I have been asked, “Will advertising boost sales of our search engine?” I am no ad guy, and I replied, “I don’t know.” Now we have a lab experiment that may answer the question. Ina Fried’s “Microsoft Kicks Off Huge Bing Ad Push” here mentioned $80 to $100 million in ad spending. Let’s convert that to revenue in search. For 2009, Autonomy is going to generate around $900 million from its various search related businesses. So at $100 million MSFT is spending the equivalent of nine percennt of Autonomy’s revenue (rough estimate only, Autonomy watchers) to make Bing a household name. Endeca (privately held so I’m using my model here) I estimate to be in the $125 to $150 million range for 2009, so Microsoft is spending the equivalent of Endeca’s revenue for 2009 with some spare change left over. If I look at my estimates for smaller vendors, the $80 million is the equivalent revenue for the next largest 12 search firms I track. Now2 look at the Google. Assume the Google hits $20 billion in revenue in 2009. Also 90 percent of the Google’s revenue comes from ads. The $100 million now looks small. Ms. Fried wrote:
The TV spots are being done by JWT, while Microsoft’s Razorfish unit created the online ads.
In a few months we will know if advertising an information utility like Bing.com “works”. Looking at my rough estimates, I don’t think MSFT is spending enough. The story made one telling point:
“Google is so much a part of everyday culture,” aid Danielle Tiedt, general manager for marketing in Microsoft’s online unit. “It is the verb. If you talk about search you talk about Google.”
Google’s own data mavens will be interested in the ad value for search as well.
Stephen Arnold, June 3, 2009
Search Is Sick
June 3, 2009
I am on my way to the airport for another luxury-suffused flight. I scanned Boston.com’s story “Microsoft Ads Say Search Is Sick, Bing Is the Cure” here. For years, I have been pointing out that search is not sick; search is dead. My research indicates that people don’t like to type words into a search box in the hopes that their guess will unlock the information riches hidden in the system. I still think search is dead, and the efforts to improve key word search result in modest improvements in usability.
Jessica Mintz’s article this morning is an AP story, so I can’t quote from it. I can offer a couple of comments:
- Microsoft is spending big money to point out that search is much better with Bing.com. My thought is, “Will users in a hurry to get info perceive a substantive difference?” There has to be more on offer than overviews of video content, right?
- Bing.com is a decision engine. I struggle with that notion. I am not sure that most people know what a “decision engine” is, and the idea scrapes against my view of information systems that generate high value intelligence outputs. Cognos and SAS are two examples of companies with systems that come closer to my idea of a decision engine.
- Incremental change may not narrow the 50 or 60 percent market share gap between Google and Microsoft. When Microsoft announced Bing.com, Google announced Wave. One looked backwards to search. The GOOG looked forward to real time information.
I will look for the Microsoft ads, but I am not sure my info habits will change too much.
Stephen Arnold, June 3, 2009
Mahalo: More Spin on Search and Money
June 3, 2009
Peter Kafka’s “Jason Calacanis Tries Turning Mahalo into a Wikipedia that Pays” here provides some insight into how an entrepreneur thinks about search and content. The Mahalo search engine was a notable social approach to building an information resource. The idea, like Wikipedia before it, was to rely on humans to provide links and content. For whatever reason, that model does not seem to have the traction needed to keep traffic soaring. Mr. Kafka summarizes the two changes Mahalo has made in its approach. One tweak is for young eyes only; that is, more info on each screen. The second is to implement a Mahalo “bucks” plan. I don’t grasp the notion because I am used to paying people for their services and then doing my thing with the content. As Mr. Kafka explains the Mahalo idea, I sighed. Mr. Kafka wrote:
But now he’s hoping to get Mahalo users to do the work, Wikipedia-style, with a twist–he’ll pay them. The pitch: Calacanis will offer users the chance to “own” a results page, and split any advertising revenue the page generates, primarily via Google (GOOG) AdSense. He’ll be paying users with “Mahalo bucks,” which cash out at 75 cents on the dollar, so users are really keeping 37.5 percent of each dollar their page generates. Calacanis says some of his pages are generating up to $10,000 a year, but most will make far less. Will that be enough to encourage people to build and maintain Web pages on a piecework basis?
I will be releasing a free compilation of my series “Mysteries of Online”, information that originally was developed for talks at various venues. I have a couple of sections about monetization of online information in that 34 page PDF, which becomes available on July 1, 2009. The bottom-line is that unless an information service generates what I call a “clean stream” of revenue, the costs of marketing and administering online services can suck the life out of a useful online service. Paying for content works if the information is “must have” stuff. Examples include certain chemical information, actionable intelligence for financial services firms, and “keep us out of jail” info for a legal matter. Once that high value info is captured, then the marketing and administrative costs kick in. The editorial costs never go away. Lower value info fall prey to the cost of keeping info fresh (hence long update times for certain info) and keeping pace with new info (hence the urgent need to monetize real time info).
I am not sure where Mahalo falls on the spectrum of “must have to nice to have to everyone has”. Perhaps the approach with create lots of eyeballs which can be monetized courtesy of ad outfits. In my opinion, the new improved Mahalo has quite a few moving parts. I like the “clean stream” approach. With the Bingster and the GOOG improving their ad supported results, Mahalo may face a long, hot summer without money for lemonade.
Stephen Arnold, June 3, 2009
Beefing Up Google Blogger Search
June 3, 2009
Short honk: if you have a Web log on the Google Blogger.com service, Search Engine Watch reported in “Google’s Blogger Makes Search Box Available to All Users” here a bit of code magic. Google can make some things dead easy. Wave and the Google Search Appliance are not yet in this “dead easy” category though. The Search Box, said Search Engine Watch:
Search Box automatically detects new posts that it can incorporate into the search results. The gadget also picks up on a blog’s style and colors. Search Box uses AJAX Search APIs to create a tight look that integrates well with the blog. Custom Search helps serve up the results.
Search Engine Watch provides a useful summary of the steps one follows to make us of this service. Coding made easier than Visual Studio. Slick.
Stephen Arnold, June 6, 2009
Criticizing GOOG and MSFT with Angel Feathers
June 3, 2009
ComputerWorld lives on advertising and happy tech companies. However, throwing praise at Microsoft’s Bing Kumo and Google’s Wave does little to set the publication’s “voice” apart. An article that finds fault with Bing and Wave is just what is needed. But there’s a problem. Get too critical and ad dollars and hot tips may go elsewhere. Even more chilling is a letter from Bing Kumo’s or Google’s legal eagles. The result is what I call “angel feather” analysis. A downside is identified but presented in a very upbeat, chipper way. After all, who wants hassles if you are trying to make a go of a publishing business today.
Check out “Bashing Bing, Whacking Wave” here. The weakness of Bing Kumo is that it makes decisions for a user. Too many decisions leave the user in the dark about what’s included and excluded. The flaw in Wave is that it arrives with Google’s legacy and a new fondness for bloat.
Both of these are important points, but ComputerWorld stops short of spelling out what the business implications are of these increasingly similar companies’ approach to online information.
I am going to follow in ComputerWorld’s footprints. I think both services are just swell. The addled goose does not want to know how results are shaped. Furthermore, the need to arm wrestle complicated systems as Google sucks metadata from the human interactions with data is super cool.
Love both services to death.
Stephen Arnold, June 2, 2009
Twitter and Facebook Go Army
June 3, 2009
“US Military Jumps on the Twitter Bandwagon” appeared in the New Zealand 3News.co.nz online publication. You can read the story here. I don’t know how accurate the details are, but I wanted to document the write up. According to 3New.co.nz:
The US military in Afghanistan is launching a Facebook page, a YouTube site and feeds on Twitter as part of a new communications effort to reach readers who get their information on the internet rather than in newspapers, officials said Monday. The effort, which officials described as a way to counter Taliban propaganda, represents a sea change in how the military can communicate its message to foreign and American audiences.
I stumbled upon an interesting Web log several years ago called Company Command. I have lost track of that. I am not sure about the work flow required to get content on these services. Company Command offered some gems and then became a less interesting read. I wonder if the posting process will smooth.
Stephen Arnold, June 5, 2009
-
- Subscribe to Beyond Search
Feature archive
News archive
-
