• Home
• About
• Writers
• Landscape
• Wizards Index
• Fraud
• Wanna Be Like Larry?

Exogenous Complexity 2: The Search Appliance

I noted a story about Fujitsu and its search appliance. What was interesting is that the product is being rolled out in Germany, a country where search and retrieval are often provided by European vendors. In fact, when I hear about Germany, I think about Exorbyte (structured data), Ontoprise (ontologies), SAP (for what it is worth, TREX and Inxight), and Lucene/Solr. I also know that Fabasoft Mindbreeze has some traction in Germany as does Microsoft with its Fast Search & Technology solution. Fast operated a translation and technical center in Germany for a while. Reaching farther into Europe, there are solutions in Norway, France, Italy, and Spain. Each of these countries’ enterprise search and retrieval vendors have customers in Germany. Even Oracle with its mixed search history with Germany’s major newspaper has customers. IBM is on the job as well, although I don’t know if Watson has made the nine hour flight from JFK to Frankfort yet. Google’s GSA or Google Search Appliance has made the trip, and, from what I understand, the results have been okay. Google itself commands more than 90 percent of the Web search traffic.

The key point. The search appliance is supposed to be simple. No complexity. An appliance. A search toaster which my dear, departed mother could operate.

 The work is from Steinman Studios. A happy quack to http://steinmanstudios.com/german.html for the image which I finally tracked down.

In short, if your company operates in Germany, you have quite a few choices for a search and retrieval solution. The question becomes, “Why Fujitsu?” My response, “I don’t have a clue.”

Here’s the story which triggered my thoughts about exogenous complexity: “New Fujitsu Powered Enterprise Search Appliance Launched in Europe Through Stordis.” The news releases can disappear, so you may have to hunt around for this article and my link is dead.

Built on Fujitsu high performance hardware, the new appliance combines industry leading search software from Perfect Search Corporation with the Fujitsu NuVola Private Cloud Platform, to deliver security and ultimate scalability. Perfect Search’s patented software enables user to search up to a billion documents using a single appliance. The appliance uses unique disk based indexing rather than memory, requiring a fraction of the hardware and reducing overall solution costs, even when compared to open source alternatives solutions…Originally developed by Fujitsu Frontech North America, the PerfectSearch appliance is now being exclusively marketed throughout Europe by high performance technology distributor Stordis. PerfectSearch is the first of a series of new enterprise appliances based on the Fujitsu NuVola Private Cloud Platform that Stordis will be bringing to the European market during 2012.

No problem with the use of a US technology in a Japanese product sold in the German market via an intermediary with which I was not familiar. The Japanese are savvy managers, so this is a great idea.

What’s this play have to do with exogenous complexity?

One of the hottest trends in search and content processing is what I call the Endeca partner model. Endeca, prior to its acquisition by Oracle, was an ageing technology with an MBA turbo boost. The firm’s consultative approach generated reasonable margins, but the company hit a revenue “glass ceiling.” As Autonomy sped to $900 million in revenues and a $10 billion buy out by Hewlett Packard, Endeca garnered about $150 million in annual revenues and a $1.5 billion buy out from Oracle. Neither company offered a search appliance.

In fact, the winner in the search appliance world is Google. According to my very unreliable sources, Google has placed about 60,000 GSAs since its rollout more than seven years ago. Index Engines has an appliance which has found traction in the niche of organizations needing to search back up archives. I have a list of appliance vendors in a file somewhere. At the top of that list is Thunderstone, arguably the outfit which created the first viable enterprise “search toaster.” Thunderstone has been quiet in the last 18 months, but the firm delivered a reasonably priced, reliable solution in the “toaster” format; that is, a server arrives. The administrator plugs it in, works through the set up screens, and the Thunderstone gizmo indexes what the administrator directs.

Endeca’s innovation was to enlist a number of partners with quite a range of capabilities. Some partners were specialists in integration; others did ecommerce; and others still were consultants, hardware resellers, user experience specialists, or makers of add in components like sentiment analysis. To build buzz for Endeca’s solutions, the company marketed at the top of the management pyramid using “experts” from well known business schools.

The issue, however, became one of revenue. The complex partnering set up produced money, just not Autonomy scale money. That’s one of the hallmarks of exogenous complexity. A company assembles a number of business partners, and there is traction. But my argument is that the financial payoff is inhibited by the exogenous complexity of managing the partner relationships. I think that Fujitsu may encounter these headwinds:

1. The amount of money that flows to Fujitsu may have a glass ceiling. Autonomy’s method was direct; the Endeca method consumes time and resources without generating the upside that Autonomy’s frontal assault delivered.
2. The cross cultural issues are likely to make for some interesting discussions. The need for speed in making sales will be hampered by trying to reconcile how key players scattered across the world want to operate. Wow. Quite a management job.
3. The niche in which Fujitsu is competing is crowded, and the model for delivering search is undergoing rapid change. If I hear one more vendor talk to me about search apps, I am going to go to work at one of Amazon’s new retail stores as a greeter.

In an effort to build a business, managers have lost the simplicity of just solving a problem. The complexity and variables of building a search business increase as you pile on the partners.

Maybe Fujitsu can prove that its approach to search will deliver Autonomy scale payoffs. My hunch is that the Endeca experience may the best the company can achieve. Hey, a $1.2 billion payoff on $150 million in revenues is pretty good. The problem, of course, is that Fujitsu (PINK:FJTSY) has $4.5 billion in revenue, which has been declining. To move the needle is going to be difficult.

Another victim of exogenous complexity may be making its way into business school case studies. In short, complexity outside of the box itself. Interesting.

Stephen E Arnold, February 15, 2012

Sponsored by Pandia.com

Comments

Copyright © 2008 Beyond Search · ArnoldIT.com · Disclaimer · Revolution News theme by Brian Gardner · Log in