Cloud Computing: What’s Required
October 20, 2008
Seeking Alpha ran a long analysis by Gregory Ness titled “Cloud Computing: What Are the Barriers to Entry and IT Diseconomies.” I thought the analysis was quite good. Not surprisingly, I had several thoughts occur to me, but I find it stimulating to read thoughtful work by an individual who approaches a subject in a helpful, informative way. You can find the full text here. The most useful portion of the write up for was the discussion of the infrastructure. The gap between Google and the also-rans in the Web search game boil down to plumbing. Mr. Ness understands its importance. I don’t agree with his assertion that we have entered “Infrastructure 2.0.” My view is that Google built on AltaVista.com’s experiences and applied itself to addressing fundamental issues such as file and record locks and unlocks, minimizing message overhead in massively parallel systems, and confronting the problems of traditional Codd database structures in its first year or two of existence. Since that time, Google has continued to make incremental improvements in its decade old system. Companies trying to catch Google are not going to get very far if those firms try to embrace Infrastructure 2.0 as more than a word envelope. Amazon–a company which seems to get more mileage from modest R&D and information technology investments than others–has made good progress, but I doubt that its engineering foundation is as robust as Google’s. But Google, like Amazon, can fall over as the recent Gmail outage proves. Nevertheless, plumbing is important. When I was wandering around Crete, I saw some ruins that we thousands of years old. Those ruins had terracotta water drains visible. Plumbing is old stuff. I don’t think archaeologists talk about “Plumbing 2.0.” Despite my dislike of the “2.0” reference, this is a good bit of work. A happy quack to Mr. Ness.
Stephen Arnold, October 20, 2008
Boom Is Lowered Gently on Yahoo
October 20, 2008
Kara Swisher lowers the boom on Yahoo gently in “What Yahoo’s Looming Costs Cuts Actually Mean (Not as Many Layoffs as You Think), which appeared on October 17, 2008. The hook for the write up is Yahoo’s firing people. I won’t cite a number because whatever that number is it won’t mean as many as I think. With regards to Yahoo, I don’t think much about layoffs. These are inevitable, and regardless of what the company will do in the next three or four months, Yahoo’s sitting on a cost time bomb. Nuking employees won’t do much. If you are a believer in Yahoo, you will enjoy the new announcements cogently summarized by ReadWriteWeb here.
Here’s what my research has turned up.
Yahoo has numerous search systems, search licenses, search initiatives, and search technologies. Today it is desirable to have a less heterogeneous technical sandbox. Not at Yahoo. Overture has a primitive search system, which I could no longer find on the redesigned Yahoo site. No problem because traffic for Yahoo advertising seems to be stable or gently undulating like long slow waves in the moonlight. There are two “flavors” of email and search delivered from the Stata Labs acquisition. No problem. Since the acquisition of Stata Labs, I can find email in the Yahoo system. There’s the Web search. Again no problem it is neither better or worse than Google’s Web search but Google has carried the day for now. There’s Flickr search. There’s other search systems kicking around. One reader reminded me that Yahoo’s real shopping search is Kelkoo. More information here. You could fiddle with the InQuira powered help search system until recently. I like using it to locate “cancel service”. Give Help a whirl here. For a laugh look at this attempt to “improve” Yahoo help.
If I am happy with these different search systems in general, why do I think collectively these very same systems are Yahoo’s cost time bomb. Three reasons:
- It costs money to maintain different systems. Staff, consultants, hardware. The more an organization has, the more it must spend for information technology.
- Heterogeneous systems means staff are not easily interchangeable. This means that Yahoo has to either hire more consultants or live with hacks that may operate like small roadside improvised explosive devices. Yahoo doesn’t know when a fix is going to create a problem elsewhere. These are unbudgeted fixes until one goes pop. CFOs don’t like this type of pop.
- Adding a new feature or function means that Yahoo either has to pick a horse to ride, thus keeping other systems in a position of imposed obsolescence or find a wizard who can produce a fix that works across heterogeneous systems. If this path is followed, see item 2 above.
Yahoo is busy creating new, new things. The hard fact for Yahoo is that much of the underpinnings are old, old things. You don’t fix these problems by firing people. You fix these problems by facing the reality of the infrastructure and making even more difficult decisions about technology, actions, and services. Firing people is expedient, and it will grease the skids for whatever Yahoo’s current pet consultant company recommends. But these steps, like Ms. Swisher’s analysis, lowers the boom gently on a ship that is struggling with flawed engineering. The ship, gently reader, she is not sea worthy.
Stephen Arnold, October 21, 2008
Yahoo Imposes Unilateral Profile Changes
October 19, 2008
I have a Yahoo email premium account. I have written before about killing Yahoo for fee services. Since I analyzed Yahoo’s email search system for a paying customer, I just left the account sitting in cyberspace. As part of the test, I created a custom news profile, slapped some sources on the page, and fiddled with the point and click color and layout functions. I check the site periodically to see what’s new. In the last year, the layout changed so an email link is sometimes hard to find. Eh, so what? Then there were new themes. None of which seemed particularly useful to my 64 year old eyes. Eh, so what? Then there was the sharp deterioration in the shopping search. Eh, so what? I did not pay much attention because Yahoo was morphing into a less and less relevant service for my needs.
Imagine my surprise when I found out that Slashdot posted another Yahoo change. You can read the original Slashdot snippet here, dated October 19, 2008. Yahoo explains what it did and why here. As far as I am concerned a free service can change any time it wants. For me, Yahoo’s fiddling around with open source, its Web log asking for help to improve its help, or this shift in profiles are irrelevant. This addled goose is not going to flap his wings or make a sound.
However, it seems that some users are annoyed with the blank profile delivered to them. I logged on and took a gander. Here’s what the new blank profile looks like:
Amazing. This blank layout is easier for me to read. I can even spot the tiny links to email at the top left hand corner of the display. I don’t care too much for the weird handling of USA Today content, but I skip that drivel regardless of color. The nice red of the stock market declines leaps out at me. Although not shown in the list of financial results is Yahoo’s share price at closing on October 17, 2008, at $12.90, down about $20 from Microsoft’s offer earlier this year. That delta of $20 speaks volumes about Yahoo.
The company is adrift. Grand stand plays like making everything open source won’t work. Even the helpful Yahooligan who reminded me that the real Yahoo shopping service is Kelkoo.com, not the big shopping search link on the splash page’s search box. You could have fooled me. I thought that when the main page’s search box’s shopping label was clicked, Yahoo would deliver the goods so the speak. Nope, that’s not the “real” service. I also pine for Mindset, a Yahoo experimental search service that was somewhat more helpful for me than the “real” search service. Mindset disappeared without warning in the last 12 months or so.
VideoSurf Looking for Wave of New Users
October 18, 2008
VideoSurf, a new online video search engine, is inviting people to try out its beta engine. No username or password is required.
The engine is built on “computer vision” – VideoSurf has designed it to search and “see” inside videos to index content rather than depending upon tags and descriptions that can produce spam. The goal is to return more relevant results on keyword searches.
VideoSurf’s competition is Google Video Search and blinkx. We wrote about blinkx back in May here.
VideoSurf boasts more than 10 billion videos indexed. Visit their site for more information.
As for this beta user? A test search on VS of “Simon Pegg Star Trek” listed the top result as the “Star Trek” teaser trailer, followed by an interview on “Friday Night with Jonathan Ross”, then several more trailers before other Pegg errata. Google’s top results returned only movie trailers, while blinkx listed a couple trailers followed by several interviews and media event clips.
Jessica Bratcher, October 18, 2008
Yahoo: Pragmatic Advice
October 16, 2008
Silicon Alley Insider does a good job of identifying Yahoo’s weaknesses and pointing out some obvious remedies. The consultants racing around Yahoo will have to lay out options for Yahoo, prioritize them, and dress the painful ones in a Project Runway gown. Yahoo has big problems, and you can get up to speed by reading “Yahoo Cracks $12 , Valuation Now Officially Ridiculous” by Henry Blodgett here. I wanted to add one point to Mr. Blodgett’s analysis. Yahoo’s heterogeneous approach to platforms and software adds another, more troublesome problem to the mix. Some fixes can’t be made because the time, cost, and complexity mean the job is just too big. Other fixes work for one service, but the features can’t be made available seamless to other services. For an example, just navigate to Yahoo and run a shopping query. Now navigate to Google’s shopping service and run the same query. I am running these test queries from the UK, so you may have to rekey the search phrase I used, “quad core motherboard”.
Which set of results makes more sense to you? Yahoo has some bright people, but the platform is looking more like a major liability Yahoo or its eventual owner must address.
Stephen Arnold, October 16, 2008
Yahoo Search: More Needed than an Ad Campaign
October 16, 2008
On October 14, 2008, Yahoo revealed new search features. You can read the story here. You can read ReadWriteWeb’s discussion here. The publicity suggests real change. I just see an ad campaign, including radio spots. With its share price below $14, Yahoo needs to deliver muscle, not window dressing. The economic downturn will exact a toll on Google, Microsoft, and Yahoo. Of the three, Yahoo will be least able to turn the dip into a scoop of ice cream.
That Microsoft buy out offer looks better than a new ad campaign for Yahoo search. I think it still looks good to some Yahoo shareholders. For Yahoo fans, this announcement may be a big deal. For this addled goose, it’s one more example of a company that went from leader to AOL clone in more ways than one.
Stephen Arnold, October 16, 2008
Google and Media Searching
October 15, 2008
Google seems to be serious about media searching. How do I know? Larry Page, one of Google’s founders, received US7,437,351 “Method for Searching Media” on October 14, 2008. What makes this invention interesting is that Mr. Page did not have any co inventors. The Google founders may be thinking about green energy and their real estate development at Moffett Field, but Mr. Page takes time to do real Googley work. Here’s the abstract from the patent document which was filed in 2003:
The present invention is directed to a computer-implemented method and apparatus for searching in response to Internet-based search queries using a search engine and an electronic database. According to one example embodiment of the present invention, data sets representing published items are input, for example, scanned-in or sent electronically, and stored in a searchable database. Each data set includes text from at least one published item. Responsive to the search query, a search engine searches for and identifies relevant Web pages and data sets representing published items and, in a more specific embodiment, ranked characterizations are returned for the relevant web pages and published items. An electronic path can be provided with the published item for accessing further information about the published item. In one embodiment, the electronic path is a hyperlink from a characterization of a relevant published item to a more complete electronic representation of the relevant published item. Publishers provide authorization to display copyrighted materials through a permission protocol.
You can obtain the complete document from the USPTO.gov Web site. My take on this invention is that it is plumbing to allow Google to operate on and manipulate binary objects so more sophisticated and newer methods can be used to unlock media files; for example, the voice to text system to permit searching within the sound in a video file. Sergey Brin invented that technology for Google, by the way. One final comment to Cyrus, the Googler who is somewhat unfamiliar with the conventions of Google patent documents. “This patent document has the same lousy diagrams that Google includes in most of its filings. Check it out if you have time between emails, SMS messages, and other Google sales activities. Try Google’s own patent service, but it is not as useful as some for-fee services, however.”
Stephen Arnold, October 15, 2008
Yotify: A Social Search Engine
October 15, 2008
Technology Review, a publication with the imprimatur of the Massachusetts Institute of Technology, analyzes Yotify.com with some able help of the Kelsey Group. Andrew Freiburghouse’s “Making Search Social” reveals quite a bit about Yotify.com and about what the MIT-endorsed publication finds newsworthy. First, click here to read the full text of the October 10, 2008. You will have to dismiss the MIT-endorsed pop up advertisement and then you can read the full-text of the Yotify.com review. Man, I hate those pop ups, but if MIT likes them, what’s the opinion of an addled goose matter? No a whit I assure you.
Mr. Freiburghouse explains that a Yotify query is an alert with smarts. Yotify.com uses the word “scout”. Users provide information about their interests, and Yotify combines ads, collaboration, and search. To be fair, Mr. Freiburghouse notes that Yotify.com is skewed to shopping. And, towards the foot of the article, Mr. Freiburghouse quotes the Kelsey Group regarding Google’s market share. A consumer oriented search system, even with the lift of social search and software agent technology has its work cut out to gain market share.
I gave Yotify.com a quick look to refresh my memory. I ran a query for “quad core processor” and the system at 8 10 pm Eastern displayed:
A user can share results, get alerts hourly or daily, and use the related searches feature to refine the query.
This type of system will appeal to an individual who wants to obtain shopping, travel, and articles listed in classified ad systems. I fiddled with the Web log function, but it was less useful than the shopping content. Over time, the Yotify.com index will add poundage.
Why did Technology Review rev its engine when it learned about Yotify.com? My speculation is that Yotify.com is different from Google. Yotify offers social and collaborative functions. Yotify has lots of buttons, controls, and options to make the user feel as if he or she has control over the results. I think an hourly update on a “quad core CPU” might be just what the doctor ordered for a Technology Review editor needing a break from the intellectual cage match at the publication.
I am waiting for Yotify.com to create an enterprise version. Granted the Craigslist.org content may have to be swapped out for something more substantive. Procurement teams and information technology professionals looking to deploy a search system that works may find the Yotify.com technology applied to a regulated industry like pharmaceuticals just the cure for ails information access. Until there is an enterprise version, I like Yotify.com as a system that offers some interesting features for consumers. If you are intrigued by consumer search systems, give Yotify.com some attention.
Stephen Arnold, October 14, 2008
Smart Money versus Start Ups
October 12, 2008
Matt Marshall’s “Expect to See Start Ups and VCs Hit Standoff over Valuations” is a very important article. The piece appeared in Venture Beat on October 10, 2008. The hook for the story is that a lousy market puts VCs at odds with the companies these firms funded. Mr. Marshall provides useful color about the different approaches some VC firms use when looking for the next Google. The examples are ones you will want to tuck in your notebook. Insider info like Mr. Marshall’s is hard to find.
For me, the most interest comment in the article was this passage:
It may be next year before we can give a serious assessment of the true fallout for start-ups. Expect to see more companies go out of business too, as VCs in some cases decide not to invest at all.
The impact on start ups will be immediate and continue for at least a year. I agree.
So, what’s the impact on search and content processing? I will be giving this subject considerable thought in the weeks and months ahead, but I have some preliminary thoughts. I want to capture these before they dissolve from this addled goose’s mind. Also, keep in mind that I may change my views as I obtain more data and do more critical thinking. Here goes:
- Backlash. I think there will be a backlash against consultants who promise quick, easy, and cheap fixes to problems with search, content processing, and content management systems. The notion that a dab of Neosporin and a bit of tape will make the pain of flawed information systems go away will be dismissed out of hand. Martin White and I have written 250 pages that explain the methodical approach needed to back out of a search system disaster. A big problem cannot be resolved overnight, so management expertise in budgeting and controlling work becomes more important than “recipes” or “silver bullets”. Tough times demand management resolve, not placebos and truisms.
- Push back. Companies offering platform solutions that are not will have a difficult time closing new deals. In fact, I think the economic climate will encourage organizations to seek point solutions that can, if warranted, be scaled to handle larger jobs.
- Protectionism. Vendors will escalate their efforts to create lock ins for their existing customers and whenever possible set up deals that lock out competitors. I learned about one large company that is solidly Microsoft and the procurement team is looking only to Microsoft for a solution. The goal is “one throat to choke” for the customer. For Microsoft, it is control of the account. The problem is that Microsoft does not have a solution that will work, so the loser in this deal with the naive licensee who will spend millions and end up with the same information problem. The goals of each party deliver a problem wrapped in what looks like an ideal solution. The fur will fly in 18 months. Today, customer and vendor are drinking to one another’s health.
- Attrition. I encounter too many entrepreneurs who believe their approach to search is the “next big thing.” In most cases, these companies will find that revenues will be tough to generate. I talked with one company three weeks ago and encountered paranoia about my call. The irony of this call is that it was prompted to put the company in a major consulting firm’s “watch” database. The call was, therefore, a “good news” call, but the business owner heard only the veiled threats his own mind whispered in his ear. The issue was resolved, but this “fear” will close off opportunities for some companies leading to less likelihood for revenue magnetism. Fear and paranoia are not as appealing in tough economic times; pragmatism and common sense are pretty charming in my opinion.
- Skepticism. Prospects won’t believe much of what some vendors say unless the vendor is already in the fox hole with the customer.
- Baloney. Lots of Buffy and Trent marketing and PR information will be generated. I wish I was 23, filled with energy, and able to invent new buzzwords to describe functions and operations that are 50 years old.
If you want to add or modify the items on this preliminary list, please, use the comments section for this Web log. Don’t write me directly. I am on the road, returning to the US in about nine or 10 days. My email systems perform miserably when out of the country, but the Web log system is pretty reliable.
Stephen Arnold, October 12, 2008
Ask: In Trouble
October 11, 2008
CNet ran a Web 2.0 deathwatch article on October 10, 2008. The company that caught my attention was Ask.com, Barry Diller’s killer search system. Well, Mr. Diller terminated with extreme prejudice Jeeves, the cartoon logo that I liked. Bang, Jeeves. Now, according to Rafe Needleman, Ask.com may be troubled. You can read his story “11 Troubled Web Companies: The Next Kozmos?” here. The write up about Ask.com is short, so I won’t quote any of Mr. Needleman’s material. My question is, “Is this a surprise?” Maybe I should start my own search deathwatch? I have some patients in mind. Just a thought. Just a thought.
Stephen Arnold, October 11, 2008
-
- Subscribe to Beyond Search
Feature archive
News archive
-
