Pradeep Singh, chairman of Aditi Technologies, has rebuilt his company around product development services

The reception of iSoftTech Solutions' Chennai office defies convention. So does its business model. Strange corners protrude into the small, yet tastefully done, lounge space. Three brightly-lit corner pillars are covered with inspirational text written by Rabindranath Tagore, Ayn Rand, etc. Beyond the reception, there are five meeting rooms and about 100 open cubicles, each of which is painted in bright purple. The cubicles come in two sizes - a subtle reflection of the flat two-tiered organisational structure not uncommon in start-ups. The first type is six-and-a-half by six-and-a-half, meant for the product engineers. The second is six-and-a-half by ten feet, meant for managers. Even president and COO Vijay Babu sits in one of them.

A month from now, about 20 of the cubicles will be vacant. The engineers who inhabited these for over three years have just been acquired by Yahoo!.

It was from these very cubicles that they developed a software product called Bloomba. Simply put, Bloomba is an email client with a powerful search engine that does away with the need for folders. Heavy email users with hundreds of emails in their inbox will find Bloomba much easier to use than Microsoft Outlook.

But iSoftTech does not own the intellectual property of Bloomba. It is owned by the US-based Stata Labs. The product was its brainchild. But it was completely developed by iSoftTech. The latter was acting as an outsourced product development company for Stata Labs.

Yahoo! recently acquired Stata Labs along with the 20 iSoftTech engineers working for Stata Labs. Yahoo! now plans to use Bloomba to take on Microsoft's Outlook and Google's Gmail.

This is the biggest breakthrough yet for the three-year-old iSoftTech, but not in the way you might imagine. The deal may have brought in a substantial, though unspecified cash inflow. It might even have enhanced the company's global reputation, but that's not why this is such a big deal. This is exciting because it offers evidence that iSoftTech's unique business model is working.

The $1.4-million iSoftTech is not an IT product development company. It does not dream of owning world-changing software products. Neither is it an IT services firm. It doesn't hope to cut bills worth hundreds of millions of dollars through run of the mill development and maintenance contracts; iSoftTech treads the middle ground. It is a product development services company operating in the cusp between products and services.

Recently, quite a few small to medium-sized companies have ventured into the unexplored space of outsourced product development. Bangalore-based Symphony Services was among the early players (See 'A Different Music', BW, 19 July 2004). Then there are Aditi Technologies and Aztec Software (Bangalore), BrickRed (Noida), and Aspire Systems (Chennai). But what is a product development services company?

Typically, when any company starts to develop a software product, it would have a unique differentiator or core of the product. That core generally accounts for 15 per cent of the development work of the product. This would invariably be developed in-house. But the product would also have a lot of other modules. These are essential, but they are not big differentiators. This is where product development services companies step in - they develop these non-core modules.

There are also occasions where companies opt for completely outsourced product development - they own the core idea, and get everything done outside. That's exactly how Stata Labs used iSoftTech to develop Bloomba.

Gowri Shankar, CEO, Aspire Systems

There are several reasons why companies developing products are willing, even keen, to outsource. It helps developers cut the time to market. It also frees up the product development company's management resources to concentrate on the most critical core of the product. This could turn out to be a big opportunity for Indian companies.

A Nasscom-McKinsey analysis forecasts that the product and technology services revenues of Indian companies could reach $8 billion-11 billion by 2008. Over 23,300 software companies in the US, with combined revenues of over $103 billion, spend 18-25 per cent of their revenues on product development. That adds up to a $19 billion-26 billion annual spend on product development (See 'Who Needs Outsourced Product Development?'). Theoretically at least, almost all of it can be outsourced. But there is one problem.

Very few companies in India - we are really entering rarefied strata here - have the skills needed to develop world-class products. Besides, why would IT products companies ever want to make a business in giving away such products to other companies?

There is a reason. Software products fail more often than they succeed. Good, and even exceptional products, have failed for lack of financial or marketing muscle. In many cases, new, disruptive technology can make a product obsolete in no time. Even large product companies face huge risks. Start-ups have limited chances of success in full-fledged product development. And who better to tell you that than an entrepreneur who tried it and gave up?

Pradeep Singh is a feisty entrepreneur. He was among the earliest Indians to set up an IT products company. The former general manager of the Microsoft's Windows 95 mobile services group chucked his job in Redmond to float Aditi Technologies. Aditi developed a CRM product called Talisma in the mid-1990s, and later spun it off as a separate company. But nothing that he had studied at Harvard Business School or learnt at McKinsey prepared Singh for what was to come.

Over the next decade, Talisma sucked in about $70 million investments, mostly funded by venture capitalists. Singh, who did not match those investments, was left with a dwindling minority holding in the company.

In 2002, he gave up his CEO job and in 2003 resigned as its chairman. Talisma is a global company today and is doing well, under new investors with deep pockets. But Singh has no role in the new scheme of things. So, he went back to Aditi. "Aditi was left rudderless when I moved to Talisma. It was time to give it a new direction."

Singh is a product creator at heart. He was not interested in the IT services business because it offered no scope for intellectual property creation. Moreover, the services business was now all about scale. There was no way Aditi could catch up with a billion-dollar Infosys or Wipro. But, he also knew that he would be unable to sustain another new product. "I have lost my appetite for taking big risks like Talisma. Creating a new product company in today's environment is difficult." That was why he decided to explore the middle path.

Now, Aditi has stopped developing products on its own. It only provides product development services to other companies. It has revenues of about $20 million, and expects to grow 50 per cent this year. "But, even if we are creating products for someone else, the excitement of product creation still remains with us," Singh explains.

Vijay Babu, COO, iSoftTech

A good example of this is a product that facilitates secure sharing of digital content (music, games, etc) that Aditi recently developed for a start-up company. The start-up approached Aditi with an idea to tackle online piracy and illegal sharing of digital content. It wanted Aditi to convert the idea into a product that could be sold to retail websites. Over the next 12 months, about 40 engineers from Aditi worked on the idea. The team took end-to-end responsibility for the product, from discovery to delivery - from creating the marketing requirements document, product specifications, prototypes, design/architecture and development to the final shipping. The result was a product that ensures that consumers who buy online music are not able to pass it on. Using this product, the start-up has established itself with leading music distributors and secured the rights to retail almost 450,000 songs. The product is now being used in one of the world's largest e-commerce sites.

It is only recently that Indian companies are getting opportunities to develop complete products on their own. Bloomba is the other such example. Such opportunities used to be rare; but they are now becoming more frequent. Still, a bulk of the product development services pertains to smaller modules. For example, iSoftTech is now working with a large European company to build a new router that will take on the likes of Cisco and Juniper. The core of the router - a data path engine - is being developed by the company. But, iSoftTech has been given a three-fold responsibility: it has to develop the network management software for the router; it will be responsible for about 80 per cent of the testing of the completed router; and it will have to develop all the embedded software for the router, jointly with the European company. This project started off with two people, but has since grown to over 20 engineers.

Similarly, Aspire Systems, a Chennai-based company, has bagged a contract to test a media management product. (It controls screening of ads, billing, slot management, etc., for television channels.) The $3-million Aspire is also developing a data driver for a Houston-based company. This driver will enable PC users to access data on mainframes with relative ease. It is being developed by a 20-member team; ten from the Houston company and ten from Aspire (See 'The Different Shades Of Outsourced Product Development').

Companies generally start off by developing smaller modules before they graduate to developing a full-blown product. But, even this calls for substantial skill-sets, most of which are simply not available with even large IT services companies. First, these product development services firms need to maintain a constant hotline with the customer. They need to be privy to almost every discussion, debate and development at the client-end. Else, they face the risk of lagging behind in the development. "We work as if we are the extended engineering team of our customer," says Babu.

Second, the very approach to writing code is completely different. IT services companies seldom look at the code-base. If the functionality of the program is fine, they are happy. If the front-end works for the user, that's enough. But product companies drill deep into the code of the modules that they have outsourced. Code-efficiency and quality is important; it is not just enough if the module works. "Initially, we had great difficulty adjusting to this mindset," concedes Gowri Subramanian, CEO, Aspire Systems.

But for companies that are able to make the transformation, the bounty is plentiful. First, margins in the business are much better than in IT services. "Billing rates per engineer per month can be as high as $7,500, compared to the $2,000-$4,500 that services firms make," says Babu. Sources say billing rates are at least 30 per cent higher than the IT services industry. "The client is ready to pay higher bill rates for predictable high-end product delivery."

More importantly, these companies have the opportunity to do bleeding-edge product development. "It is as if a small company is actually working on a group of products, but without the associated risks," says Babu. There can be no greater turn-on for many hi-tech start-ups. Finally, even one successful product can bring global recognition and a consistent stream of revenues. Product upgrades, new versions, etc., keep product development services companies busy even after the initial launch.

But despite these obvious benefits, one uneasy thought remains. Are the outsourced product development companies making a mistake by giving away all the intellectual property rights (IPR)? The answer is no.

Even if they do own the IPR, Indian companies are in no position to exploit it. Most of them do not have either the financial or the marketing muscle required. And product development services companies follow this business model only because they don't want the risks associated with new products. Besides, this is the only way most product companies will play the game. If they are forced to share the IPR, their inclination to outsource may dissolve.

But a few like Babu believe that it is possible to create IP even if you are working in the product development services model. "We continue to build our own IP," he says. Now, iSoftTech is developing reusable components that could be licensed out to clients. For example, it has developed a workflow engine which an IT company can use to manage its projects. It has also developed a customer support management software for call centres. This is being marketed jointly with another complimentary product owned by a Delhi-based firm. But this model does have its limitations.

Other interesting solutions are also emerging. An organic foods company in the US is seeking to disintermediate the distribution chain of the multi-billion dollar industry through an IT product. The development of this product has been outsourced to Aspire Systems. Though Aspire will not own the IP of the product it develops, it will receive an equity stake in the organic foods company. This will ensure that Aspire also gets to share the long-term benefits of the IP it has created. Aditi Technologies too has operated in the equity-share model in the past, and may do so again in the future. Such an arrangement could be a win-win in the long run.

For the moment, outsourced product development is seen as one way for small IT service firms to differentiate themselves from the larger firms. This model may not appeal much to the biggies because it is not as scalable as IT services. But for smaller companies, outsourced product development is a terrific learning opportunity. More importantly, these companies are helping build a product development ecosystem within the country. Who knows? One day, one of these companies could develop a killer product of its own.

With inputs from Mitu Jayshankar