Betting On The Millennium Bug

Only a 45-minute drive from Chennai is the sleepy industrial township of Ambattur. But in the world of software, it counts as one of the busiest centres of activity. This is where Indias largest software company, Tata Consultancy Services (TCS), has set up a huge software factory. Housed in this 75,000 square-foot facility are 1,000 software programmers scanning and repairing millions of lines of software code everyday. Arguably, it is the biggest Y2K project in the world.

Here, time is of the essence, because within two years, the market for which TCS and many others are gunning for, will cease to exist. So its now or never. Come January 1, 2000, and it will all be over.

 

The Year 2000 problem, or Y2K as it is popularly known, is about a quirk in the software that might cause computer systems to go haywire as we cross over to the next millennium. Put simply, on that fateful day, computers will reset their internal clocks, and in the process, wreak havoc on precious databases and spreadsheets. Banks, financial institutions and many other corporations who run mission-critical applications are, therefore, bringing in software experts to solve this problem.

Much ado about a two-bit repair job, you wonder? According to the Gartner Group, a well-known IT consultancy, the Y2K business could add up to no less than $500 billion (Rs 2,000,000 crore). (Though today, most observers believe that the figure is a bit of an exaggeration. And there are concerns that a large majority of firms are simply refusing to do anything about the problem.)

So in the end, the business may not live upto its hyped up potential. But that is small deterrent for local software firms, as they rush in to carve up the opportunity.

The Strategist pans on the four software firms, who have in their own way, fashioned unique responses to the same business opportunity.

Software from the shopfloor

Software programming is a creative work.

At TCS, the Y2K work is organised on the assembly line manufacturing model. Nerds like to work flexi-time.

At Ambattur, there are shifts in operation.

Making software is different. You can work at home, seek help from a team in another city, or even in another country.

Nearly 1,000 people have shacked up at Ambattur working exclusively on the Y2K project.

So it is a software job with a difference. And there is a good reason why it allows a company like TCS to take a radical approach. To make a software program work smoothly during the millennium transition period, or make it Y2K compliant, as the industry jargon goes, is essentially a repair-and-test job. Identify patches in the software code with date occurrences, make modifications, test the program and ship it back to the client.

At least thats how the Y2K business opportunity appears to TCS. A job with low value addition, and low margins. To the Rs 600 crore software giant, focusing on high volumes therefore seemed liked an obvious course of action.

S Ramuadorai, CEO of TCS describes his organisations approach thus: We are dealing with a short time window during which a large amount of Y2K work needs to get done. Our factory at Ambattur operates like an assembly line and we have the capacity to do over a billion lines of code before the next millennium starts.

TCS has already done 300-400 million lines of code. It has bid extensively for Y2K projects all over the world on different types of machines and softwares. Our ability to optimally use the resources depends on the number of projects that we get at any point of time, says S Mahalingam, executive vice president, TCS.

It is not very often that software specialists concern themselves with the subtleties of production theory. But these days Mahalingam talks in terms of manufacturing imperatives. The main characteristic of this work is the fast turn around time. It needs to be extremely well planned and monitored.

In essence, TCS is attempting to prove that nothing changes, even if the manufactured product is just a bundle of bits and bytes. So each Y2K assignment is broken down into discrete processes, with each process automated to the extent possible, and then specialist teams attack specific sections of the code.

To extract maximum efficiencies out of this assembly line operation, TCS opted to consolidate the entire Y2K work under one roof rather than spreading the work among different centres. Unless the client agrees to upload his code to Ambattur, the company does not take up the Y2K assignment.

The Rs 50 crore complex has massive IBM mainframe machines which can hook up nearly 1,000 computer terminals. And a high speed communication network with some special characteristics allows TCS to uplink with any client in Europe or in the US within just three weeks.

But what will the company do with these resources after the project ends in less than three years? Not so much the hardware but the hordes of software professionals whom TCS has hired and trained for the Y2K project specifically.

Again, the solution has come from the shopfloor: outsourcing. TCS has brought in smaller software companies to work on Y2K assignments at Ambattur. Nearly 60 per cent of the Y2K teams comprise of non-TCS engineers.

The big-volume mindset has begun to pay off. When The Strategist met the CEO of TCS in October last year, he had projected the total Y2K business at around 500 million lines. Last week, their executive vice president revealed a substantially upgraded target of almost one billion lines of code. Even the most conservative estimate would indicate that TCS would net $ 350 million business on Y2K assignments alone.

Pinning on realism

Big is beautiful may only be a partial truth. Or at least there may be more relevant verities for other players. To a smaller company, an exploding market opportunity could present a peculiar dilemma. Investing in fixed assets and other resources for a market that will disappear in a couple of years may not offer cut and dried responses.

Should it pitch itself for a cheap one-stop shop for hundreds of millions of lines of code and hire hordes of software programmers a la TCS? Because other things being equal, employing more people can enable a larger volume of software code to be tackled and Y2K projects are labour-intensive. And most of the opportunity available to Indian companies is in the high-volume code.

Hiring too many people for a quick ramp-up can mean a lot of deadwood after year 2000. If an organisation is not able to grow proportionately, it will not be able to absorb the people hired for Y2K, says Uday Gharpure, managing director, Origin Information Technology (India) Ltd. On the other hand, if the company hires on the lower side, the size of its business will be very less. And it may lose an excellent opportunity to kick-start its growth.

Or should it look for a longer-term relationship with its potential customers? (It can then use its various other software services to amortise its investments and then throw in the Y2K service as an add-on.) So what will be an optimal level of investment for the Rs 30 crore company like Origin then?

Origin (India) is a subsidiary of Philips NV and does IT consulting and other services for the enterprise resource planning software market. It has been growing steadily at 70 per cent every year and by the year 2000, it hopes to hit about Rs 80 crore. But doesnt Y2K present that once-in-a-lifetime opportunity to break into the big league very quickly?

No, seems to the cautious response coming from Origin. It has chosen to peg its investment norms very tightly on its projected growth rates, with the Y2K business not getting bigger than 20-25 per cent of the total turnover. We dont want to exceed our exposure beyond this level, says Gharpure.

A breach of this level can lead to problems, he says. Utilising the hardware is not a big issue, but people are. Origin does not want to recruit people with only Y2K capability. Our selection procedures are uniform, for Y2K or for any other line of business that we have. We see no point in assembling a rag-tag team of COBOL programmers, says Gharpure.

Why? A lesser qualified person may require further training and even become a cost to the company. Since the company is looking at a longer time-frame (through long-term contract), its pricing for fixing the Y2K problem does not reflect the cost of additional hardware and people hired for the Y2K problem. So if the hired person is not competent enough to contribute to other businesses, the company has a laggard on its hands. And whoever it sacks or retains, the transaction costs are still high.

At the same time, the person hired for doing just Y2K work may not bring the level of commitment to his job. The idea is to give that person a career path and not just let him off after the project is over, says Gharpure.

Both these factors make sense for Origin. It needs both these skills for its various service lines in no small measure. To maintain a steady growth, we are not actively bidding for business. If someone comes upto us, then fine, claims Gharpure. This also allows the company to play a stronger hand, while negotiating prices. When a Y2K customer walks in through Origins door, there is little doubt about who will set the terms.

Clearly for firms like Origin, Y2K is merely the means to an end. The bigger the end, the greater will be the thrust on an apparently non-core business like Y2K.

Does it fit my core business?

Certainly for NIIT, it doesnt. Y2K is mainly a problem on chunky mainframes. These monstrous machines are gradually ceding ground to the new networking platform of client-server architecture. NIIT would rather get a piece of action in emerging client-server market.

Arvind Thakur, senior vice-president, NIIT, says that it takes a long time to establish credibility in overseas market. We have positioned ourselves as a leading edge technology player in the US. We dont want to dilute that positioning by trying to become all things to all people.

Thats why the company is not seeking any Y2K business in the US market. With a sizeable chunk of NIITs business coming from that market, Thakur believes that Y2K would send confusing signals about the companys core capability.

But at the same time, Y2K is too lucrative an opportunity. So why pass up the lucre? A large chunk of Y2K business is on the mainframes. And NIIT is into mainframe maintenance business. So the company has trained its sights in another market: Europe. The company does not have a strong foothold in Europe, and therefore, could build its business portfolio afresh. And with Europe moving to a single Euro

currency, the promise of a bigger business opportunity in the

future is clear.

While on one hand, NIIT did not want to commit large resources to the Y2K project alone, breaking into a fresh market couldnt happen overnight either. And yet, the time on the Y2K opportunity was running out fast. So NIIT had to seek assignments through a few local systems integrators like the $1 billion Logica, with whom it has done business in the past. Clients hesitate to transfer their software codes to a company unknown to them, says Thakur, so we needed a local interface which these systems integrators provides.

The local company bids for business, does the initial business and technology impact studies for the client. Only once the problem areas are identified, does NIIT steps in to repair the code and then test it. Its charges are determined by the man-hours it commits to the project.

This low profile approach has led to only a small amount of business to come to the Rs 320 crore company. The goal: get about Rs 40-50 crore in Y2K assignments and dont let the share of Y2K business exceed 10 per cent of the total turnover. Will this body-shopping approach and the small quanta of business targeted bring NIIT into reckoning in the European markets?

If anything NIITs experience is a reminder that no market is homogeneous to be cracked open with one mindset. And a players capabilities and aspirations determine the share and the nature of the market it can take home. So if you want to go higher in the foodchain, you need better vision and faster reflexes.

Segment and augment

So look carefully and there shall exist markets with markets, not ad infinitum but large enough nevertheless. This is the Y2K world, according to Syntel. For this $125-million company with 700 people employed locally, Y2K is really the aperitif. The main course is the huge IT outsourcing market, growing really rapidly world-wide.

We have a product called Intellisourcing, for which we want to create a market. Therefore, we are selling Y2K services only to strategically important clients, says Baru S Rao, chief operating officer, Syntel India. However, just because Syntel wants to sell Intellisourcing, there is no reason for the market to buy it. If anything, Y2K is considered to be a service that does not add value to existing system, and an unnecessary bother for customers.

Technically, the customers should be content with exorcising the ghost in their machines and get on with it. But Syntel has been winning contracts in transportation, insurance and investment banking in the US, which have all been tied up for Y2K work.

The reason apparently lies in the perception of the opportunity and presentation of the solution. If a Y2K positive customer allows Syntel to put its foot in the door, then Syntel can, while checking the code, understand the innards of the business system. Then it can suggest a better business system which the IT system describes to be adopted. If it sounds eerily like the stuff IT consulting outfits do, rest assured. It is.

Understanding the language of machines is one thing and mastering the fundamental of various businesses is another. So Syntel has started hiring people with understanding of businesses and not merely code-cutters. So in addition to hiring the ubiquitous engineer, Syntel has been hiring chartered accountants, PhDs in mathematics and operation research and people with degrees in transportation engineering. Some of the candidates have been given further intensive training in insurance business, apart from software engineering refreshers.

Such methodical approach and preparation necessarily affects pricing. Bra presents Syntaxs pricing mechanism enigmatically. Nor do we price ourselves by the number of lines of code neither do we go for a fixed-price contracts, says Bra Rae. But the process has to be expensive, right? Bra says, Let us say just one thing: We are not cheap.