Google Losing Its Cool
March 21, 2009
Not my phrase. This wordsmithing appeared in the Gawker story “Google’s Data Fetish Drives Away Its Top Designer.” The phrase that caught my attention was not the resignation of a Googler. The red light was:
Bowman’s reasons for quitting are fascinating — and they show why Google’s losing its cool.
You can read the Gawker story and unravel the motivations of a Googler who sought greener pastures. The quest for quantitative data reminded me to recommend that you read John Ralston Saul’s Voltaire’s Bastards. You can get a copy on Amazon sometimes. Go for it here. Google is the high point of a data driven management culture. Zeta function, anyone?
Stephen Arnold, March 21, 2009
Evri: Semantic Smack Down
March 21, 2009
I don’t know much about Evri. Semantic technologies intrigued me a few years ago, but the shift is toward real time content processing. Semantics are important but in my mind plumbing that operates as a contributory component. I did write about the company’s deal with the Washington Post here. The Washington Post needs every (no pun intended) advantage it can get. Ad revenues are down. The Treasury is printing money like one of those fake countries in South American pot boilers. Even upscale restaurants’ business is down in the ultimate Power Lunch town.
I was surprised, maybe shocked, that Evri was shedding staff. Venture Beat here published “Semantic Search Engine Evri Cuts Staff by 25 Percent.” My impression was that Evri was going like a Harrod’s Creek mine worker on his way to the local watering hole. The most interesting comment in the article was:
Even so, Roseman [the president] says the company is pleased with its traction and progress, drawing more than 20 million monthly users.
Google AdSense on 20 million uniques should generate big money for Evri if properly monetized courtesy of Mother Google. Plus, Evri has received about $8 million from a Seattle investor. With strong uptake and big traffic, I wonder if staff cutbacks are a sign of the times or a signal that semantic search may be suffering in a down market for publishers such as those Evri has nailed as customers.
Stephen Arnold, March 21, 2009
Vivisimo’s Stan
March 21, 2009
Clusty has a new friend, Stan. Stan, said Vivisimo, is:
an everyman. (Hey, that rhymes!). He could be you, or your colleague or anyone frustrated with their company’s ability to search internal data (or inability to search). Stan is the man!
You can learn more about Stan here. The FAQ here provides more information about this personification of a user of an enterprise search system. I learned that Stan has a Web log. You can read his musings here. I learned from the March 17, 2009 posting:
I heard about this enterprise search thing. Frankly, it sounds too good to be true. Searching across multiple repositories from a single search box. Presenting results into topical clusters. Tagging and rating documents to impact future search relevancy. Sharing results with other users. All of this while respecting my individual security rights. Seems like pie in the sky to me. Is anybody actually doing this, or is it just some marketing hype? So I’ve started this Web site and blog to figure things out for myself. But I need lots of help. So bookmark the site, friend me on Facebook and follow me on Twitter. Give me suggestions for future videos and blogs postings. Maybe I can actually find something at work now.
Like the a recent consultant’s report about search, Vivisimo has jumped on Google’s “simple” marketing bandwagon. Google’s game plan calls for its sales professionals to state that Google’s enterprise search solution is simple. When one licenses the Google Search Appliance, the customer eliminates the months of fiddling that go along with some search systems. The administrative console makes set up quick and painless. Google makes available to customers who want to do more the OneBox API and expert Google partners like Adhere Solutions. But the message is clear. Google reduces the hassle and hence the cost pain of setting up and maintaining some enterprise search solutions.
Stan is at http://meetstan.com
Is the CMSWatch report right? Is Stan right? I have no idea. I get asked questions about search because the caller has discovered that what looks easy was not for that caller’s particular situation. In general, search is hard. Even Steve Ballmer said that search was tougher than it looked.
Simple is good. Great teachers make the complex easy to understand. In search, there are a number of issues to simplify. Getting one component wrong can create some challenging situations. Some create budget havoc. Others lead to legal matters.
I suppose that if most people insist that search is simple, a minority opinion is irrelevant. At some point, a person may sit down and figure out what has to be done to make electronic information available for specific use cases in a particular organization. I don’t think that is a simple task. Someone may have to figure out how to transform content to make it crunchy. Someone may have to figure out how to create a 24×7 system with near real time updates. I don’t think that is simple either.
For now, I am going to suggest that making issues clear and easy to understand is good. I hope that mythical Stan does that. Simple as a marketing tactic may not be appropriate for some information access situations such as eDiscovery, medical information, law enforcement, or similar applications. Just my opinion. Simple.
Stephen Arnold, March 21, 2009
Interwoven: Autonomy IDOL Integration
March 21, 2009
The deal closed one day and the next day Autonomy announced that it had integrated its IDOL (intelligent data operating layer) with Interwoven Teamsite. You can read the story by Ben Pitman here. The integration, said Mr. Pitman, will:
enable a host of additional IDOL functionality within Autonomy Interwoven’s Web Solutions. Customers will benefit from a combination of legacy keyword and advanced conceptual technologies and an order of magnitude increase in performance and scalability over the previous release of TeamSite.
My take on this is a bit different, which is the domain of the addled goose. Forthwith:
- Microsoft bought Fast Search & Transfer in April 2008 and the integration has not yet been completed. Autonomy’s velocity makes clear how serious the company is with regard to leveraging its acquisitions. Microsoft’s progress? I will leave that to you to draw a conclusion.,
- Vivisimo, Interwoven’s search partner, now finds itself under additional pressure. There’s the general marketplace squeeze that affects every company and the Autonomy IDOL squeeze which is quite particular.
You can season my interpretation with your own experience and that of your search consultants to taste.
Stephen Arnold, March 21, 2009
Navgle: Nifty and Useful
March 21, 2009
I don’t like the word or phrase mash up, usually written mashup. I don’t like the phrase programmable Web. Heck, I don’t like much of the wordsmithing generated by the trophy kids who swarm over Web companies. I do like a service called Navgle. You must try it here. The name is a combo of Naver.com (South Korea) and Google.com (intergalactic). I looked at this service on the laptop of a Google partner not long ago. Naver.com dominates search in South Korea and has revenues of about US$1.0 billion. More information about the company is here.
I needed to locate information about Cleveland business failures, and I was not too happy with lots of Crain hits and the baloney about how wonderful the northern Ohio business climate was. I decided to give Navgle.com a try.
What happened? See for yourself. Go to Navgle.com. Plug the phrase “Cleveland business” in the search box, and you can see what I found useful. Here’s the result screen for my home run query a couple of days ago. Navgle.com updates frequently, so your screen will probably look different from this one.
The first result–note, the first hit!–reported that Crain’s Cleveland Business was chopping employees. When you run your query, you will see that Navgle.com returns results by type of content, delivering the top hits from each in a useful relevance ranked order.
On this one page, I could see that Cleveland is in a world of hurt. There were Tweets. I had Google search results. I had useful Web log entries. I had snippets from Google Groups. In short, I had a useful way to get an overview of the Cleveland business scene.
Is Navgle.com perfect? No, but it is a heck of a lot more useful than most of the search systems I use on a daily basis. I can think of a number of ways to use the Navgle.com system to make this Web log more useful to my five or six readers. In fact, I am going to write the Navgle.com “contact” and see what the company has to say. Maybe I can wrangle a way to put some search related info links on this site to provide more context for my opinion pieces.
I keep getting emails from readers who ask, “Who are you? Why do you act like a stupid goose?” Navgle.com might slap some bacon around my write ups and the grousers can live with my view of the world.
Bottomline: a nifty service. Not much latency. A very useful service. What’s with the tie up between Naver.com and Google.com? In my opinion, more than meets the eye.
Stephen Arnold, March 21, 2009
Google and Microsoft: Core Competencies
March 20, 2009
One of my readers sent me a couple of links to remarks by Steve Ballmer, Microsoft’s top gun, this morning. The news story contained one revealing observation here attributed to Mr. Ballmer:
I’ll tell you the search business has been harder for me and for other senior people to learn than most businesses…
Interesting because Microsoft has invested in search and watched as its market share in Web search has declined. The company has not fared well in the enterprise arena either. The $1.2 billion investment in Fast Search & Transfer has yielded since April 2008 when the deal concluded:
- One Norwegian police action
- A Web part
- A roadmap.
But Google’s YouTube.com executive quoted in Mediaweek’s “YouTube Exec: Google Not Good at Content” here reveals a similar incompetency. Kevin Yen, director of strategic partnerships at Google allegedly said:
We’re not good at content… We create platforms that allow other people to succeed.
After investing in multiple content-related businesses, creating a legal stewpot with copyright hawk Viacom, and launching Knol (yep, I know. What’s a Knol?)–Google admits that it is not too good in content.
Step back.
What do these quotes reveal about two of the most influential companies in software, electronic information, and systems?
My thoughts:
- Competence in one area does not transfer competence to another area
- Believing one is able to solve every problem does not mean that one can actually solve those problems
- Technology does not equip Google and Microsoft with magic wands.
Just my opinion about two influential companies with reservoirs of confidence in their abilities to climb every mountain.
Stephen Arnold, March 20, 2009
We Have the Data, Now What
March 20, 2009
ZDNet’s Tom Espiner’s “Gov’t May Track All UK Facebook Traffic” here. I found the post interesting because of this comment:
Under the EU Data Retention Directive, from the 15 March, 2009, all UK Internet service providers (ISPs) are required to store customer traffic data for a year. The Intercept Modernisation Programme (IMP) is a government proposal, introduced last year, for legislation to use mass monitoring of traffic data as an anti-terrorism tool. The IMP has two strands: that the government use deep packet inspection to monitor the web communications of all UK citizens; and that all of the traffic data relating to those communications are stored in a centralised government database.
Let me be clear. I have zero problem with law enforcement taking such steps to fulfill their mandated duty. If that is a problem for you, quit reading my Web log. My concern is that lots of data poses a search and retrieval problem. When those data change, the problem gets bigger and quickly. Based on my research, there are only a handful of companies with the technology to tame these exascale data sets. One is Google. The other is Exalead. If you know of others, please, let me know. Just bring facts, gentle reader.
Stephen Arnold, March 20, 2009
Financial Times: Try, Try, Try
March 20, 2009
Flashback. FT.com year 2005. I was a paying subscriber. I got a user name and a password. I logged on. Ran a query and the system timed out. Flash forward to 2007. FT.com licenses Fast Search & Transfer. I tested the system. Slow. I was asked to test a semantic system under consideration by the Financial Times. Useful but slow, slow, slow. Now the Financial Times has tapped another point and click vendor for a “deep” search experience. Time out. The Financial Times, arguably one of the two bigger franchises in business information, has been a laggard in online search for quite a while. The FT’s parent owns a chunk of the Economist, another blue chip in business information. I was a subscriber to * both * the print and online editions until late 2007. Why did I drop these must read news sources? Too much hassle. I hope the FT’s new system moves from the “deep” to the daylight. I hope the FT monetizes successfully its content. I hope that I will be able to play in the World Cup, but I am a realist and recognize that hope not mean accomplishment. If you are cheerleading for a dead tree outfit that once owned a wax museum, read the Guardian’s “Financial Times Launches Business-Focused Deep Search Service” here by Kevin Anderson. The article included a useful description of what the FT hopes to do with indexing:
The service allows users to search easily by news topic, organisation, person, place or theme. If a user searches for stories about business in China, the search can quickly be refined to cities in China, showing stories about Beijing, Shanghai or Hubei. Greenleaf described this as a “know before you click” model so that users can see related topics and the number of stories available for each sub-topic. In addition to automatic tagging, Newssift editors have also added other relationships to the service relevant to their business audience so that if someone looks for news about Ford Motor Company, they can also see related content from Ford suppliers.
This type of metatagging is useful, but it is computationally and human intensive. But the main difference between this most recent try in FT’s quest to develop an online service that makes up for the precipitous loss of revenue from its traditional dead tree business is the economy. Too late. I wish the FT team success, but I don’t think this most recent service will deliver the cash needed to get the ship squared away for even rougher seas ahead. Red ink ahead in my opinion.
Stephen Arnold, March 20, 2009
Emotional Search
March 20, 2009
There’s social search. People search. now there’s emotional search. This new species is explained by Patricia Skinner in her article “Search is Getting All Emotional Thanks to Twitter” here. I have been getting email explaining that my Twitter stories are chasing off some readers. Not much I can do about that. Free Web log and all that apply. Her point was:
Because Twitter is essentially a huge network of millions of tiny networks (you and your friends), talking about what matters to you and what you think about. So a kind of ‘innate, emotional-level understanding’ is built in.
Ms. Skinner includes some examples, a bit of information about behavioral marketing, and some examples. I agree with most of her points. Is Twitter important to search? In my opinion, yep. I don’t care much for the emotional angle, but tapping real time content flows is a big deal.
Stephen Arnold, March 20, 2009
Microsoft Tom Tom
March 20, 2009
Tom Tom is not rolling over. Microsoft wants to choke off what it seems to see as improper use of its FAT32 innovations. I don’t think Microsoft is advancing on Tom Tom. I think the Tom Tom action is a probe prior to deciding what to do on a far bigger front. You can read “TomTom Fires Back at Microsoft with Lawsuit over Streets & Trips” here to see how messy litigation becomes. The matter is now one involving not guns and butter but patents and baloney. Take a moment to read some patents. Check out the claims. Do you know what the invention does? I have to read these stylized documents multiple times to get a sense of what the heck has been “invented”. My view is that Microsoft wants to make a point about companies who choose not to use Microsoft software, systems, and methods without Microsoft’s blessing. When it comes to search, I see the Tom Tom action as a flash of lightning against the distant horizon. The storm may be arriving quickly. Will the Tom Tom gambit work? Who knows? Where legal processes operate, mere mortals can only speculate.
Stephen Arnold, March 20, 2009
-
- Subscribe to Beyond Search
Feature archive
News archive
-
