Login

Bala Iyer

Thursday, October 02, 2014 5:24 AM
     

Current Articles | RSS Feed RSS Feed

Mashup network picture -- Feb 24

Posted by Bala Iyer on Sun, Feb 26, 2006 @ 01:15 PM


  


Comments (2 )
  
  
s



Two presentations about Google (from the inside?)

Posted by Bala Iyer on Sun, Feb 26, 2006 @ 12:48 PM

Last week, Prof. Hal Varian visited BU to talk about his recent research using Google as the source for his data. While I did not take any notes :-), I found a website with some notes from another talk by Prof. Varian that was very close to his talk at BU. One thing that I realized very early in his talk was that Google had a very sophisticated (theory driven) auction process for their advertising engine. Prof. Varian showed how this process was very robust and scalable. An interesting aside..  to a question from me about APIs, Prof. Varian did not think that Google would charge for API calls. However, he did leave the door open for charging once the volume of transactions from any source picked up. I am reminded of Eric Schmidt's concept of URL -- ubiquity first, revenues later!

Another presentation that I viewed (thanks to Prof. Venkat) over the weekend was a recorded presentation by Paul Strassman about Google's system architecture. A key point that Strassman makes is that history may not remember Google as a search engine company, but as a company that created the architecture for large scale distributed data and application management leading to the realtime enterprise. To that end, Google has invested in hardware clusters, operating systems, and metadata management that give them high scores on reliability and response time.BTW, Strassman told the audience that programmers should learn API-related skills on core platforms to be relevant.


Comments (2 )
  
  
s



Google's analysts' day on March 2nd

Posted by Bala Iyer on Sun, Feb 26, 2006 @ 12:29 PM

Google is planning to meet with analysts this week. Although in the past they have provided very little guidance, they have asked for inputs from analysts on what they should cover during the day. The videos and presentations from the event should be available and interesting.


Comments (2 )
  
  
s



Can mashups have a business model?

Posted by Bala Iyer on Mon, Feb 20, 2006 @ 09:14 PM

Recently I read an article about how it is difficult to design a business model around mashups. In the world of mashups there are API providers, mashup developers, mashup enablers and consumers (see link). Let us look at this closely. A business model has four components: value proposition, marketplace offering, defendable resource system, and a revenue model.

A value proposition includes a target segment of consumers, focal benefit and a rationale for superior delivery. This is a requirement for any business, whether created by a mashup or otherwise. From a marketplace offering perspective, a mashup provides a service or just information. A resource system must include partners offering services such as API providers, and enablers on which a mashup is built. Included in the resource system should be a core benefit that the mashup creator must provide. This core benefit could be some processing logic or business logic that the developers has developed on their own. For a revenue model, the developer could simple run ads on the site and survive through ad revenue or sponsored ads. The other option is to create a revenue sharing model with API providers.  In the case of ebay motors, ebay could share a piece of the transaction revenue with the mashup developer based on the traffic that ebay motors direct their way. Finally, developers could provide some basic service for free and charge for premium services. The big risk is that the API providers may decide to charge more or even abandon a service. This is common even for software product development where an application provider or a middleware provider could do the same thing. In such instances, the API developer loses credibility if they change business model abruptly. Even if they do so, module/product developers could switch to other service providers. At a basic level, a mashup developer could be a system integrator that develops products using pre-existing products. In a more advanced setting, the mashup developer could be using APIs to develop new products by adding new functionality, logic and data to the basic building blocks.


Comments (2 )
  
  
s



Mashup network picture -- Feb 15

Posted by Bala Iyer on Wed, Feb 15, 2006 @ 02:10 PM


 


Comments (2 )
  
  
s



Oracle buys Sleepycat Software Inc

Posted by Bala Iyer on Wed, Feb 15, 2006 @ 02:00 PM

In my research I have found that acquisitions within the same layer of a stack do not payoff from a market evaluation perspective. Since Oracle is acquiring another database management company, what is the payoff here? Oracle's attraction here maybe to learn how to operate in the OpenSource and hybrid (open/proprietary) arena. If in the future every product has some opensource elements to it, how do companies manage IP issues? The other factor to consider is innovation. Do companies that support opensource development innovate better? Finally, regardless of the method used to develop code and innovate (open/proprietary), platform providers get a piece of the action. In the case of proprietary code it is licensing revenue and in the case of opensource code -- maintenance or support contracts.


Comments (2 )
  
  
s



Mashup network picture -- Feb 2

Posted by Bala Iyer on Thu, Feb 09, 2006 @ 06:47 AM


In this picture, node size is based on the number of APIs provided by the component. Link thickness, as before, represents the number of mashups made available using those components. Nodes representing components owned by the same firm are filled using the same color. My thanks to John Musser of programmableweb.com for providing me with access to the data.


Comments (2 )
  
  
s



Managing distributed work

Posted by Bala Iyer on Thu, Feb 09, 2006 @ 06:13 AM

In recent times, as I look at IT departments struggle with issues like outsourcing, vendor management and domain based services, my conclusion is that their core struggle is with distributed work management. While the first stage of outsourcing involved taking a big piece of functionality (IT, H/R or R&D) and giving out to one or two vendors, today's reality is more like giving out smaller pieces to more vendors.

One interesting example is described in Tom Friedman's book. In this instance, a firm on Wall Street provides some research work to Office Tiger. Office Tiger breaks the work down into chunks and distributes the execution to other companies located in different parts of the world. Once each chunk gets completed they are assembled back again by the Wall Street firm and used to service their clients. Many difficult operations were performed to complete this work order. First of all, a service platform allows all the providers to interact in a friction free manner. Who owns this platform? Should the Wall Street firm own all of it or just the piece that contains the integration logic? Second, a complete work order was broken into logical chunks and monitored remotely to meet time and quality requirements. No firm other than the Wall Street firm has complete visibility into the process. Finally, all the subcontractors got a piece of the revenue that was given to Office Tiger. Clearly there is a service delivery platform that lies beneath this story. At BU we are launching a project to study these platforms.


Comments (3 )
  
  
s



What is your core competency?

Posted by Bala Iyer on Mon, Feb 06, 2006 @ 10:25 PM

Just came out of a very interesting round table discussion with several CIOs and Mr. Azim Premji from Wipro. In describing Wipro's core competency, Mr. Premji claimed that it was talent management. Outsourcing has moved from data center management to software development to business process outsourcing to services management. Management of services such as traditional R&D for product development and knowledge work requires talent management. This is what Wipro has developed over the years.

From the demand side, given that companies can outsource design and production, what should they protect as core? Could the core competence be the ability to dynamically allocate work? Once allocated, could the integration of work be a supporting function? This raises interesting possibilities. Companies may get work done by the same providers and differ mainly in their ability to integrate and reallocate work. Furthermore, basic research and development may be done just to build their ability to understand new developments or to simply evaluate the work done by the service providers.


Comments (6 )
  
  
s



Google and analysts

Posted by Bala Iyer on Wed, Feb 01, 2006 @ 02:59 PM

Another one of those puzzling things about Google. They don't providence guidance to the street about quarterly results. Once they announce results, the market reacts positively or negatively based on their exceeding (in most cases) or coming under analyst expectations. Google then explains why they came under or over. I thought Google did not care about analyst estimates! Why do they bother explaining the actual to the estimated values? Google should simply stop this charade and start providing guidance -- this way atleast the fund managers would sleep well!

 



Comments (2 )
  
  
s



© 2014 www.balaiyer.com