The competition in India’s mainframe market is hotting up

 

Source: The Telegraph 

 

Developed economies also tend to be information societies, with high investment in information and communication technologies. ICT has transformed economic relations, enhanced productivity and created new services and markets. It added directly to the gross domestic product through employment and value addition. It also added network effects and added indirectly to the GDP by lower transaction costs and rapid innovation, thus increasing efficiency and productivity of user sectors. Among the Asia-Pacific countries, India spent 6.83 per cent of the GDP on ICT, China 6.34 per cent, Malaysia 7.42 per cent, and Singapore 10.54 per cent. In 2005, in absolute terms, India spent $55,304 million while China spent $142,313 million, Malaysia $10,137 and Singapore $12,299.

In India, understanding the market for computers — which is about to transform itself from mostly personal computers and medium-range or entry-level mainframes and mere hardware purchases to solutions orientation and hence purchase of services — is important for strategic reasons. Understanding the nature and extent of competition and the practices adopted to maintain dominance also helps regulation of competition. Indicus Analytics and ICRIER have completed a research study that sheds light on this little known market and competition within it.

The trend towards computing is accelerating. Telecommunications, banking and financial services are the major drivers, together with many other sectors like airlines, insurance, manufacturing and industry, e-governance initiatives of the government and so on. The unique identity project will give a further fillip to information technology, directly as well as indirectly, given the many uses it will spawn. Managing growing volumes of data is going to result in growing demand for high-end servers. The Indian high-end server market is showing accelerated growth, expected to average 10 per cent annually until 2012.

Server products can perform many functions, including web applications, email or messaging, front-end web serving, streaming media, high performance computing, data warehousing, online transaction processing and running infrastructure applications. Among servers, mainframes serve the needs of a whole enterprise, providing the capability for extensive workload management and uninterrupted performance and security. Workload requirements have been small and India today has only 25 mainframes in use.

The high-end server market includes mainframes that have grown in speed and efficiency and are carrying more and more of the data load for business and industry. IBM and Fujitsu are the two principal manufacturers. In 2009, IBM derived 1.10 per cent of its India revenues from mainframes compared to China’s 17.44 per cent. Indian users and prospective users consider Hewlett-Packard’s Superdome and Sun Microsystems’ Sparc system as capable substitutes for the mainframe. Industrial dominance is the least in the high volume entry-level market, more in the mid-range, and the most in the high-end server market.

Many factors enable suppliers of high-end servers to achieve market dominance. The applications influence the choice of the operating systems. Once the choice is made, it becomes very expensive to switch from one to another because of the loss of time involved and the capital cost of switching to a new system. This is heightened because systems are designed to discourage inter-operability between manufacturers’ systems. This leads to a ‘legacy’ effect. Someone who has bought from one supplier finds it difficult to move to another because of the high costs involved and the loss of business owing to the time taken to switch systems.

Many sectors try for ‘legacy’ applications to achieve loyal customers and tie customers to their products. Thus a Gillette three-blade razor can use only a Gillette blade. A user of a HP computer printer is warned to use only HP ink cartridges if he wants to safeguard the machine. In project work, contractors supply their own equipment; its regular servicing locks the customer to them. Suppliers are therefore willing to price the razor, printer, or project quote low and make their profit on the repeated purchases of their high-priced ‘tied’ products. By locking in the customer, the manufacturer assures long-term revenues and good profit. Mainframe manufacturers also aim to lock in customers. Users must find ways not to pay extra or lock in and not be able to switch to another supplier. High costs of switching make the server market competitive ex ante when the choice is being made, but not ex post once the choice has been made.

A comparison showed that IBM prices in India were lower than in China: the average price of the mainframe with ‘z operating system’ is one-and-a-half to ten times more expensive in China. This might exemplify IBM’s determination to lock in customers in an expanding market like India.

Though the explosion in chip memories has made cheaper entry-level computers capable of processing and storing vast amounts of data, IBM has increased its dominance in the last seven years from 17 per cent to 34 per cent of the mainframe market, with increasingly powerful machines. With the information explosion and a variety of large applications, IBM, in the last one year, has acquired 35 brand new customers across the world for its mainframes, of which five were in India. The numbers in India are bound to increase in the coming years as large databases become more common in many sectors.

IBM’s strategy has centred on protecting its monopoly in the mainframe market. One way was to ‘tie’ the sale of its mainframe software only with its mainframe hardware. This proprietary nature of IBM’s operating system made it difficult for customers to move to other platforms. This strategy can lead to high prices and high-cost service contracts.

‘Cloud’ computing (linking servers together) will not stop lock-in and tied sales. Linked computers in one cloud may not be compatible with those in another cloud, making switching expensive and time consuming. In the past, companies have promoted their own mutually incompatible standards and formats and this might also happen with clouds. Further, the risk of loss of privacy is an important consideration (for example, with financial data like credit cards) when sensitive data is stored and handled. For applications which are mission-critical, there is also the risk to the safety and security of data. Hence cloud computing is unlikely to displace the dominance of the big three mainframe makers (IBM, HP and Sun Microsystems, and possibly Cisco) in the next five years. Regulators must watch for this dominance extending to software and professional services.

The Indicus-ICRIER research suggests that large enterprises prefer to buy hardware bundled with the operating system. If application software were standardized and available off the shelf from other software providers, it would deny the integrated IT vendor a comparative advantage by tying them together.

To avoid being exploited by the dominant suppliers, Indian enterprises moving to large mainframes must form their own skilled in-house IT teams which can, in certain cases, reduce the monopoly power of the vendor. However, in-house skills cannot handle maintenance of mainframes and the supplier will have a virtual monopoly, unless it is possible to use open systems for which service support is easily available (though security concerns may make their use impossible). User clubs that share information and experience can help avoid supplier dominance.

Large public sector units and government departments already follow the Indian government’s practice of procurement through a fresh tender for bids every five years or so. This negates any apparent benefits to incumbent IT providers. This practice must continue.

In Europe and the United States of America, IBM is alleged to have successfully extracted huge profits from proprietary mainframes for many years. To prevent their repetition in India, studying past and future developments in developed markets can forewarn us.

Even in this most complex market, market analysis, keeping tab on market practices of the same firms in other countries, and application of competition law can protect users from exploitation.

The author is former director-general, National Council for Applied Economic Research 

---

---