Technology: The Software-Comp Revolution

Services businesses like healthcare and insurance that will need at least 5 lakh professionals in the next year are pulling out all stops in the scramble for people.

The production practices in this industry are among the most backward. Only a demand-supply equation, strongly skewed towards the former, and high profit margins that absorb the cost inefficiencies inherent in the industry's typical work practices make it the power it is...

"The market for components can only grow as technology matures and allows for a real assembly of applications"
Partha Iyengar
Country Manager (India), Gartner Group

You could be forgiven for thinking these observations refer to the cement business. Their real subject is the software industry which has all the ingredients to make the late messrs. Henry Ford and Taichi Ohno, the pioneers of the flow production system, turn in their graves: a high degree of integration (or a low level of outsourcing); and the desire to, and the practice of, writing every piece of code from scratch. Put simply, there is no software components industry.

Actually, read that sentence in the past tense; there is one, now. It may not be big in terms of numbers, but it does exist. International Data Corporation estimates that the value of the software components market will rise to $2.7 billion by 2004, up from $681 million in 2000. And closer home, companies of every hue, from software heavies like Infosys and TCS, to startups like Persistent Solutions and iSolv are exploring the outer-boundaries of a business that the Gartner Group promises will exhibit ''explosive growth'' in the next few years.

Or, why the software component industry never really took off. On the face of it, make or buy decisions are as relevant in software as they are in manufacturing. Four factors, though, prevented most companies from creating components, and other companies from buying the few that were available. One, companies are loath to invest in creating components and then look for a market for them. Two, in the absence of widely accepted industry-standards the business of off-the-shelf components is extremely risky. Three, developers are reluctant to use components in the absence of any objective quality criteria. And four, most companies that create pieces of software that fit into other, larger pieces of software, believe that by selling them as components erodes their own competitive advantage.

A cross-section of the Indian software components business

Several Indian software companies are creating niches for themselves in the software comps market. A sampling:
Ť TCS, India's largest software company, has made security, workflow and other software components with the stated intent of selling them independently.
Ť The Pune-based Persistent Systems is concentrating on customised components for its clients and calls itself the software companies' software provider.
Ť Hughes Software Systems designs GPRS stacks, frame relay stacks and other components, meant for the telecom industry.
Ť Ruksun Software has sold components for middle ware to Yahoo, msn and ICQ for their messengers.
Ť Aptech Software hawks a number of components developed on the Java platform.
Ť The Pune-based iSolv develops components for middle ware in the wireless telecom space.

Circa 2001, a fundamental change in the way companies get infotech to work for them is accelerating the process of componentisation of software. As Vinod Sood, the Chief Technology Officer of Hughes Software Systems puts it: ''Earlier, almost 70 per cent of a system was driven by hardware; now, I think close to 60 per cent is driven by software.'' That means larger (in terms of volume), and more complex code. And that means no company will be able to write all the software it needs to.

On the supply side, building software comps is a reasonably lucrative business. Typical profit margins in the industry range between 35 per cent and 70 per cent, (the higher margins are because of lesser competition) which compares favourably with the software services industry's 20-60 per cent and the software products industry's 50 per cent and above. That apart, product development cycles are shorter in the software component business. And the risk inherent in the business is lesser. Sums up Partha Iyengar, the Gartner Group's Country Manager for India: ''The market for components can only grow as technology matures and allows for a real assembly of applications.''

The dynamics of the software components industry bear a striking resemblance to those of the auto components one. Focus and specialisation are a must. And component companies can increase their profitability by moving up the chain to sub-systems. Explains Sood of Hughes: ''If you can earn $200,000 by selling a suite of components, you could earn $1.5 million by creating a sub-system around it.'' Seconds Anand Kumar, the CEO of the Pune-based Ruksun Software, which has developed components for Yahoo, msn and ICQ messengers: ''If you develop a platform or application around the component, it may fetch higher revenues.''

One breed of companies that is likely to emerge as a result of the componentisation drive is systems integrators that boast both domain expertise and a knowledge of existing software component libraries. These service providers can help companies achieve their infotech objectives at optimal cost. Says Iyengar: ''The technology hasn't quite reached the plug-and-play level yet, but it will. When that happens companies providing systems integration services will come forth.''

The three most prevalent business models in the software components industry today revolve around functions (like sales, or inventory management, or call centre management), domains (like insurance, banking, or retail), or platforms (like C++, XML, and WML). All three are viable segments. A company that is digitising its processes step by step would be an ideal customer for function-oriented components. Just as one operating in a specific industry, like retail, would be for domain-oriented components. In both cases, though, it is far more likely that a service provider helping the comp-any infotech-and e-enable its processes buys the components and integrates them into a larger, all encompassing solution.

However, to succeed in the code-components business, a company will have to offer either an entire stack of either functional components, or a complete range of domain-specific ones. And companies following all three business models, especially those vending components built around specific platforms, will have to ensure that enough developers buy into their products.

Well, money talks and almost $30 million of HSS' revenues for the financial year 2001, will come from its components and products business. Hype, the software components business isn't, but in the absence of system-standards, companies operating in this area will find it difficult to go beyond a few common applications. Avers Anand Deshpande, CEO, Persistent Systems: ''It isn't possible to use components in areas beyond basic generic systems.'' Besides, companies operating at the nether regions of the components value chain often discover that their customers wish to buy not just a component but also the rights to it. Says Kumar of Ruksun: ''If you are addressing the low- end of the software components market, you need to be thinking on your feet all the time.''

The way ahead for the software components industry, though, lies in creating a range of components that can be used across industries with little or no customisation. Already, the Dublin-based Resolution Technology, a consulting company in which IBM's Irish subsidiary IBM (Ireland) has a stake, is working on creating a suite of components for a Common Business Model-built around those characteristics of doing business common to companies across industries. Evidently, the componentisation of the software business will take time, but when it does happen, it could make 'manufacturing processes' in the industry as efficient as those in the automobile business.