Search All Information: The Categorical Affirmative Is Alive and Well
July 18, 2008
I think I took a logic class in 1964 from Dr. William Brown, a Ph.D. with an encyclopedic knowledge of Will Rogers, the American humorist. Dr. Brown could whip out one liners to make a point. I watched as my classmate when he said, “Will Rogers said, ‘A man only learns in two ways, one by reading, and the other by association with smarter people.'” My classmate avoided categorical affirmatives and negatives for the remainder of the semester. I’m not sure if my classmate understood the quotation or if the attention Dr. Brown directed at my classmate drove the lesson home.
This morning my news reader presented me with this headline: “We Offer a Single Place to Search All Information”. I clicked the link and read an interview conducted by Ruth Samson for DQ Channels, a publication with which I was not familiar. Her interview subject was Sanjay Manchanda, Director, Business Division, Microsoft India. You must read the interview here.
The subject of the interview is Microsoft, Fast Search & Transfer, and assorted closely-aligned issues. Here’s my list of what caught my attention:
- Demand for SharePoint Search comes from “companies that have huge amounts of data and need to search quickly and constantly.” My notes to myself: What’s huge? SharePoint native search dies at 50 million documents, sooner if the documents are multi-megabyte jobs with SmartTags and other goodies stuffed inside the wrapper.
- Unique selling proposition: “we have built our search capabilities so deeply that it is seamless. We offer a single place to search information of different types. Our recent acquisition of FAST, a provider of enterprise search solutions, also gives us an edge over other players in the segment.” My notes to myself: No way. Search is a crazy quilt. Fast Search is not one thing. Powerset is older technology given a new coat of paint. Baloney. Seamless! Not true. The stitches are evident, far apart, and made with a nail, not a needle.
- Microsoft has 3,600 partners. My notes to myself: Hmm. Are these SharePoint partners or total Microsoft partners? 50 to 60 partners in India. Why the spread?
- Cost of SharePoint is variable. Starts low, then increases as the number of users goes up. My notes to myself: Ah, exactly like the IBM and Oracle models. Business model may not work in the present economic environment. Search server in India costs R20,000 or $470. Now that’s a deal with Fast Search targeting $150,000 in early 2007.
- Search is “the next killer app”. My notes to myself: Nope, search is disappointing, a utility function, and not what users want. If this Microsoft expert Sanjay Manchanda is right, Microsoft is chasing the wrong ambulance.
The interview struck me as a marketing effort. Not only was the “all” spurious, the idea that Microsoft search is “seamless” is not congruent with my experience. What’s seamless is loading a third party search system such as Isys Search Software and getting something that doesn’t take a platoon of Microsoft engineers scrambling around for days.
The interview has some one-liners that would make Will Rogers smile. Dr. Brown would still be fussing over the “all”. I am hung up on “seamless”. Quite an interview.
Stephen Arnold, July 18, 2008
Search and Text Analytics in New Fall Wardrobe
July 18, 2008
Clarabridge, according to CRM Today, has announced a new program for its business partners. You can read about the deal here. The most interesting portion of the announcement is the positioning that Clarabridge is adopting. I think of these fresh, clever ways to package search and text analytics as the fall collections of fashion houses. Clarabridge is described as “a leading provided of customer experience management solutions”. The acronym is CEM. In order to deliver better CEM, Clarabridge employs search and other technologies. If you want to know more about Clarabridge, click here. The headline on the Web site describes the company as one which delivers “customer experience intelligence”. I assume that the CEM and CEI are closely related. Clarabridge’s founder did a stint at MicroStrategy, a company that has an interesting track record in business intelligence.
Stephen Arnold, July 18, 2008
Google: Stepping on the Enterprise Search Gas Pedal
July 18, 2008
Richard Martin’s “Pressed by Rivals, Google Accelerates Enterprise Search Efforts” is a revelation. The story appeared on July 17, 2008, and you can read the full text here. (You may see an annoying fly over ad and then be redirected. The wacky url worked for me.)
In my opinion, The most interesting point in the scoop was:
Unifying a single search interface across those varied systems, which often use different protocols and programming languages, has become something of a holy grail.
The religious metaphor is lost on me, but I think Mr. Martin’s point is that search is supposed to make it possible to access data, regardless of its location or format, from one interface.
I agree.
My view is that this is not a search problem and, therefore, the GOOG, Microsoft, and the companies mentioned in the article will not solve this problem with search technology.
The challenge is transformation of data in such a way that new operations are possible. Who wants to search? I want to know where a particular item comes from and how certain I can be that the item is accurate or good enough to use in a decision.
In my opinion, talking about enterprise search is a waste of time. Google is definitely going to deliver, but the solution will not be a search technology. The problem is better resolved in terms of data management. Google has some interesting technology in this practice area, and I discuss it in my new study Beyond Search, published by the Gilbane Group.
Stephen Arnold
Explicit User Input: The Future of Web Search
July 17, 2008
TechCrunch has an intriguing Web log post called “Is This the Future of Search?” After scanning the text, watching the video, you will want to read the comments to this post. I find it difficult to summarize multi-media Web log posts. My take is that Google is showing some users a Digg-like interface; that is, a user can provide input with a click about the value of a result, add a link, and with little effort on Google’s part, provide a full complement of user input functions. With the acquisition of JotSpot, Google can out Digg Digg with a user creating in the midst of a query or other Google action a “new” Web page. I provided raw material for an Outsell analysis last year, and few took notice of Google’s ability to morph into a publisher. Digg-like features are a fraction of Google’s content acquisition and creation capabilities. Traditional publishers, are you tracking Google as a digital William Randolph Hearst?
Among the more than 100 comments is one that I think centers of what is going on. Ben writes:
Ouch. I’m betting Kevin Rose is wishing he cashed up when he had one of those many opportunities.
I’d be a little worried if I were Mr Calacanis also.
Based on my research, my view of these Digg-like features are a bit like a double-edged sword equipped with those nifty barbs and hooks like this:
Source: http://www.myarmoury.com/talk/viewtopic.php?t=9497
Here’s how I interpret Google’s Digg-iness:
- Google wants to brandish a threatening weapon without actually striking anyone. This is digital saber rattling. Google shows strength, maybe intimidates or causes a competitor to think about selling or partnering with Google
- The GOOG is pushing the tip of the sword into the flesh of a opponent, forcing the opponent to get out of the way. If the opponent does not move the way the Google Jedi wants, a quick thrust and twist ends the opponent’s life. Death comes but it is painful.
- Googzilla is taking practice swings in order to use the cutting edge of its technology to hack its way through the social search opponents. Limbs and other digital body parts will be summarily removed as Googzilla stomps forward swinging the deadly blade.
I think the Digg-iness of Google is designed to position Google to buy one or more entities, force one or more entities to become partners in the manner of Nokia-Symbian’s sudden love affair with the GOOG, or outright war.
Stephen Arnold. July 17, 2008
Symbian: Simpatico toward Google
July 17, 2008
Symbian [http://www.symbian.com/] makes open source operating systems for mobile phones and has been building reference designs for telecommunications companies for years. Nokia, its major stakeholder, recently announced that it’s going to buy the rest of Symbian’s shares and turn it over in its entirety to the Symbian Foundation, a new group backed by several mobile phone companies that have pooled their knowledge and resources to create one big, happy, open-source, royalty-free, customizable Symbian platform.
Enter the Googlemeister. It’s about to launch a new product: Android [http://code.google.com/android/]. Android is (drum roll) “the first complete, open, and free mobile platform” and was developed by a group of more than 30 technology and mobile companies.
Symbian’s CEO said in a July 16 article here [http://news.yahoo.com/s/pcworld/20080716/tc_pcworld/148477] that the company has a good relationship with Google and cooperation between the two companies is possible. But he also defends Symbian and basically questions why Google is pitching Android as if Symbian didn’t exist.
Perhaps Symbian sees the hammer about to fall.
Google has grown far beyond just being an Internet portal and is a serious player in the telecommunications industry. You know the “walk softly and carry a big stick” credo? There isn’t a big enough stick for Symbian to carry to ward off Google. It might not be a bad idea to embrace the GOOG’s services and functions.
Jessica Bratcher, July 17, 2008
Blinkx Actions and Rumors
July 17, 2008
Blinkx is in the news again. As you may know, Blinkx is a video search service. Beta News reported on July 16, 2008, that the company is taking offensive and defensive action. You can read Jacqueline Emigh’s story “Rumored Google Target Blinkx Teams with Microsoft, Others” here.
Ms. Emigh cites another Web log. The paragraph of interest to me was:
“Speculation is rising around a potential acquisition of Blinkx, a video search engine, by either Google or News Corp. (and possibly Yahoo),” noted a blog posting in May by Heather Dougherty, director of research at the Hitwise market research firm.
Blinkx reports that it you can search over 26 million hours of video from Google Video, YouTube, MetaCafe, and others.
Among Blinkx’ new deals are tie ups with:
You can learn more about Blinkx here. The Register offers a different angle here.
I am not sure what to make of this story. In 2005, Blinkx was an independent company according to this posting in Search Engine Watch. I have heard that the company is linked to Autonomy, but I need to look into this alleged tie up more closely.
The phrase “sharpened its defense and offense” continues to puzzle me. Does Blinkx want to be purchased by Google or Yahoo? Does Blinkx want to remain independent? More information is needed.
Working Hypotheses
My working hypotheses are:
- This is a PR play, designed to increase visibility of Blinkx and focus attention on what are partnering deals, probably not big money makers yet.
- Blinkx is a hot property, and the company’s management is aggressively positioning the search engine for acquisition
- Video search is a money pit, and the company is pursuing tie ups in hopes of finding a formula to boost revenues and get the stock moving upwards.
The honking goose will fly over its available information and update this Blinkx coverage. My earlier essay about Blinkx is here.
Stephen Arnold, July 17, 2008
Enterprise Search: It’s Easy but Work Is Never Done
July 17, 2008
The Burton Group caught my attention with its report describing Microsoft a couple of years ago as a superplatform. I liked the term, but the report struck me as overly enthusiastic in favor of Microsoft’s server products.
I was surprised when I saw part one of Margie Semilof’s interview with two Burton Group consultants, Guy Creese and Larry Cannell. These folks were described as experts in content management, a discipline with a somewhat checkered history in the pantheon of enterprise software applications. You can read the first part interview here. The interview carries a July 15, 2008, date, and I am capturing my personal thoughts on July 16, 2008. That’s my mode of operation, a euro short and a day late. Also, I am not enthusiastic about CMS experts making the jump to enterprise search expertise. The leap can be made, but it’s like jumping from the frying pan into the fire.
The interview contains a rich vein of intellectual gold or what appears to me to be sort of gold. I jotted down two points made by the Burton experts, and I wanted to offer some color around selected points. When you read the interview, your conclusions and take aways will probably differ from mine. I am an opinionated goose, so if that bothers you, quit reading now.
Let me address two points.
First, this question and answer surprised me:
Question: How much development work is require with search technology?
Answer by Guy Creese, Burton Group expert in content management: It’s pretty easy… Usually a company is up and running and can see most of its documents without trouble.
Yikes. Enterprise search dissatisfies anywhere from half to two thirds of a system’s users. Enterprise search systems are among the most troublesome enterprise applications to set up, optimize, and maintain. Even the Google Search Appliance, one of the most toaster like search solutions, takes some effort to get into fighting shape. Customization requires expertise with the OneBox API. “Seeing documents” and finding information are two quite different functions in my experience.
Second, this question and answer ran counter to the research I conducted for the first three editions of Enterprise Search Report (2004-2006) and my most recent study Beyond Search (2008).
Search technology has some care and feeding involved. How do companies organize the various tasks?
Answer by Guy Creese, Burton Group expert in content management: This is not onerous. Companies don’t have huge armies [to do this work], but someone has to know the formats, whether to index, how quickly they refresh. If no one worries about this, then search becomes less effective. So beyond the eye candy, you have to know how to maintain and adjust your search.
“Not onerous” runs counter to the data I have gathered in surveys and focus groups. “Formats” invoke transformation. Transformation can be difficult and expensive. Hooking search into work processes requires analysis and then customization of search functions. Search that processes content in content management systems often require specialized set up, particularly when the search system indexes duplicate or versioned documents. Rich text processing, a highly desirable function, can wander off the beaten path unless customization and tuning are performed.
Observations
There are a handful of people who have a solid understanding of enterprise search. Miles Kehoe, one of the Verity wizards, is the subject of a Search Wizards Speak interview that will be published on ArnoldIT.com on July 21, 2008. His company, New Idea Engineering, has considerable expertise in search, and you can read his views on what must be done to ensure a satisfactory deployment. Another expert is my son, Erik Arnold, whose company Adhere Solutions, specializes in customizing and integrating the Google Search Appliance into enterprise environments. To my knowledge, neither Mr. Kehoe nor Mr. Arnold characterizes search as a “pretty easy” task. In fact, I can’t recall anyone in my circle of professional acquaintances describing enterprise search as “pretty easy.”
Second, I am concerned that content management systems are expanding into applications and functions that are not germane to these systems’ capabilities. For example, CMS needs search. Interwoven has struck a deal with Vivisimo to provide search that “just works” to Interwoven customers. Vivisimo has worked hard to create a seamless experience, but, based on my sources, the initial work was not “pretty easy”. In fact, Interwoven had a mixed track record in delivering search before hooking up with Vivisimo. But CMS vendors are also asserting that their system is social. Well, CMS allows different people to index a document. I think that’s a social and collaborative function. But social software to me suggests Digg, Twitter, and Mahalo type functionality. Implementing these technologies in a Broadvision (if it is still paddling upstream) or Vignette might take some doing.
Third, SharePoint (a favorite of Burton if I recall the superplatform document) is a polymorphic software system. Once it was a CMS. Now it is a collaboration platform just like Exchange. I think these are marketing words slapped on servers which are positioned to make sales, not solve problems. SharePoint includes a search function, which is improving. But deploying a robust search system within SharePoint is hard in my experience. I prefer using third party software from such companies as ISYS Search Software or the use of third-party tools. ISYS, along with Coveo, offer systems that are indeed much easier to deploy, configure, and maintain than SharePoint. But planning and experience with SharePoint are necessary.
I look forward to the second part of this interesting interview with CMS experts about enterprise search. Agree? Disagree? Quack back.
Stephen Arnold, July 17, 2008
Google: Web Search Market Share Increasing, Again
July 16, 2008
Silicon.com ran Stephen Shankland’s essay “Google’s Search Share Continues to Creep Up.” You can read the full text here. In the flurry of news about Google-Viacom, Google-Microsoft-Yahoo, and Google everywhere, I missed this point:
Its share increased from 68.29 per cent in May to 69.17 per cent in June, the analyst firm said. Over the same period, Yahoo! dropped from 19.95 per cent to 19.62 per cent and Microsoft dropped from 5.89 per cent to 5.46 per cent.
In my little world, a company under such intense media scrutiny and in the midst of legal hassles on a number of fronts, this is interesting. In fact, I can’t recall reading about a company increasing market share while coming under attack from many different quarters, in different countries, and in different technical areas.
I looked up the term monopoly to refresh my memory. The Wiktionary provided this memory nudge to me:
A situation in which solely one company exclusively provides a particular product or service, dominating that market and generally exerting powerful control over it; An exclusive control over the trade or manufacture of a commodity; A company dominating a market in one of the above manners.
If one is a stickler, the loop hole is exclusively. Google is not the only provider of no-charge Web search. I can choose to use Live.com, Yahoo.com, Baidu.com (or have one of my interns tackle Chinese for me), and hundreds of other services which I list here.
In Google Version 2.0, which you can learn more about here, I don’t spend much time discussing Google’s domination of Web search. I accept that hegemony as the status quo. After all, Google’s been grinding forward for 10 years without significant opposition except from Baidu.com and Yandex.com and maybe a couple of other companies operating outside of the ken of US pundits. In North America, despite lots of hand waving and public relations, no vendor has been able to focus the technical effort on search to leap frog Google. As a result, Google has had time to improve its operation and increase its lead as Mr. Shankland points out.
The real challenge that Google presents is that its Web search and ad business is so dominant that getting a better or clearer view of the company is very difficult. In my research, I have learned that Google has an application platform. Search and advertising are just two very successful applications running on the Google infrastructure. This means that Google could enter other markets at low incremental cost.
The company is moving into other markets now, but taking baby steps and following an unorthodox approach. Much has been made of the departure of Google executives who have cashed out or grown tired of the idiosyncrasies of life at the GOOG. Some of the recent information I have gathered suggests that Google is adapting to its work force changes. Those adaptations are likely to make it even more difficult to predict what Google will do next. Consider that:
- Google is hiring smart, young engineers and giving them freedom to make decisions as long as there are data to back up those decisions. This approach increases risk to a certain degree and it also makes surprises more likely.
- Google’s innovation process of pushing out products and services has slowed, which reflects more management discipline. Nevertheless, if a beta generates a positive result, Google flows resources to that service in response to clicks. The addition of “voting” to search results is a current example of this Darwinian behavior.
- Google is concatenating its services in roll up patent applications. Google is becoming more seamless which reduces complexity to some degree and allows faster response to market conditions.
Google’s not the exclusive provider yet. But unless the competition gets into gear, only lawyers and Google’s own management missteps will prevent Google’s extending its reach.
Stephen Arnold, July 16, 2008
Shooing Away Legal Eagles
July 15, 2008
This lawyer stuff is not to my liking. I read the Google Web log post “The Law and Your Privacy: An Update” and I shivered. I recall my debate coach in high school criticizing my linking of a broad generalization with a personal argument. He used a fancy Greek term I didn’t understand. What I did understand was his telling me that I would lose if my opponent drove a truck through the argument and over me. You can read the Google Web log post here.
Then I scanned Michael Arrington’s “Google/Viacom Agree to Preserve User Anonymity in Data Shakedown”, and I thought my debate coach would have smiled that chilling grin he used when he knew he had a winner. Mr. Arrington summarizes what the various legal eagles worked out, and then he added this remark with which I agree:
I for one have no further objections to this data being handed over from a privacy standpoint, although I still urge Viacom to stop the endless litigation and consider more innovative business models around their content.
By the way, the TechCrunch post includes the court document issued by Judget Stanton.
I have had the opportunity to serve as a consultant in some legal matters, and I think my health issues in early 2007 were blow back from the stress of these jobs. Once a legal process starts moving, in my opinion, the matter takes on a life of its own. The logic is exactly that honed in university rhetoric courses and then refined in law school.
Once inside the argument, the reasoning becomes sharply faceted. For those without much experience getting ripped by one of the winningest high school debate coaches in Illinois’ history, you have to keep fingers out of the moving machinery. A misstep can mean real trouble.
Mr. Arrington’s point is one that speaks clearly and wisely. Throwing money at a crusade is fine and dandy for a while, but most legal skirmishes can be resolved by talking. In this matter, I know I cannot tell who is “right” and who is “wrong”. Once the legal drag line starts ripping earth, you see the big gouges and piles of dirt. It’s pretty tough to envision what was there before the process began.
In my view, big media ignored the grassroots revolution that has been building. I wrote an essay about Napster years ago in which I portrayed media’s view of that service as the spawn of Satan. The angle I took, as I recall, was that the people doing the bulk of the file sharing, copying, and trading were the sons and daughters of the big media people and guys like me. I recall watching students at a major university engaging in mass audio ripping and sharing when I visited my son’s dorm. Now there are young people who just don’t understand why everyone doesn’t behave as they do.
Today, with traditional media business models failing, controlling ephemeral zeros and ones is pretty tough. No, not tough. Impossible. Google bought YouTube.com and with it Google acquired traffic and an opportunity to monetize. The traffic is still there, but the service has been difficult to monetize. Meanwhile savvy young people have seized upon online media as their next big thing.
Young people see online video as the way to snag chunks of crunchy information. I have done several research projects probing into young people’s attitudes about online information. I don’t understand what I learned, and I can sum it up easily:
Online is the standard TV-telephone-juke box-hang out super enabler. I am a dinosaur. I think it is admirable to preserve traditions. A crusade, as long as one is not killed, can be enervating. However, with each passing year, the swelling number of young people who are going about their business oblivious to the needs of courts, Google, Viacom, and adults in general is growing. I would assert that if Google and Viacom did not exist, the inexorable march of executives’ children will continue into the fluid, copyright indifferent digital future.
Based on my research, in my opinion, it is more productive to work out a pragmatic solution and turn one’s attention to serving the needs of this growing and soon to be dominant majority consumer mass. I have a wonderful gold Waterman fountain pen, a relic from my Booz, Allen & Hamilton days in the late 1970s and early 1980s. I don’t use it. I can write this essay on paper, and most people can read my writing. But I have had to adjust to the new world. My expensive Waterman is sleeping in its velvet lined storage box, but I don’t think I will use it again–ever.
Adjust and adapt. Otherwise the rising tide of young people’s behavior will marginalize BOTH new companies like Google and the old guard like Viacom.
Agree? Disagree? So educate me, already.
Stephen Arnold, July 15, 2008
Google: SLAs Here and QoS’s Not Far Behind
July 14, 2008
Stephen Shankland’s essay “Google Aims to Earn Business Trust in the Cloud” provides a good summary of Google’s service level agreement for Gmail. An SLA specifies a level below which a an operation will perform. Mr. Shankland’s focus is on the white hot topic of cloud computing. Organizations are reluctant to trust any information to a service that is flaky. The most important point for me in his essay was:
Google is trying to communicate better with users and customers…
SLAs alone are not enough to reduce risk for an organization contemplating moving services and information “out there” in the cloud. This summer is what I have called “the summer of transparency”. Everywhere I look I see Google executives chattering about the technical innards, the child care services, privacy, and other Google issues. One of my clients said, “Enough Google already.”
But there is an interesting aspect of Google’s SLA that warrants mentioning. In late 1998 or early 1999, Google engineers began working on a Quality of Service invention. You can read more about this invention by downloading US7142536, Communications Network Quality of Service System and Method for Real Time Information, Filed December 14, 2000. Granted November 28, 2006. Google’s SLA is just one component in a broader set of guarantees which extend into the methods often associated with assuring system level functions, not application level functions.
With “guarantees” like these, Google can assert that risks associated with moving to the cloud are to some degree ameliorated.
Stephen Arnold, July 15, 2008
-
- Subscribe to Beyond Search
Feature archive
News archive
-
