Adam Smith and the Software Industry
Published: October 1, 2001
Published in TDAN.com October 2001
“ . . . every individual . . . intends only his own gain, and he is . . . led by an invisible hand to promote an end which was no part of his intention.” Adam Smith, in Wealth of Nations, Book IV, Chapter II. Paul Samuelson said it with more finesse: “The oil of self-interest will keep the gears working in almost miraculous fashion. The market will answer all things.”
This paper has two parts. The first relates my company’s experience from when we commenced software development in India three years ago. The second surveys the macro picture at the national level.
In the beginning . . .
A decade ago, a company that I founded and took public developed the client-server accounting suite USL Financials. Last November, the Journal of Accountancy featured USL Financials as "one of the ten leading products" for its market out of a total of 200 to 300 available products. Development costs of that suite through alpha testing were about $2.5 million. Together with our software development team in Noida, India, we recently brought its successor product to the same point for $800,000.
In the fall of 1998, specifications for the successor product were completed, but my company did not have its own development staff. We had a reliable but limited source of funds, and our options were to either raise a lot more money, assuming we could find a suitable investor at this early stage, or find a way to get the job done on our limited budget. We preferred the latter, knowing that our valuation would be much better after our software suite was operational.
We were aware of the offshore development phenomenon, but had no exposure to it. The person in our organization who seemed to be in the best position to know – he was a director of Computer Sciences Corporation – explicitly warned against it; too risky. Still, researching the option wouldn’t do any harm, would it? So the research commenced, consisting of numerous Internet searches using every search engine available. We found 112 offshore software development firms, from the Baltic to the East China Sea. Incidentally, a similar search had been conducted as an academic exercise 18 months earlier, and fewer than a dozen sources were found.
Given this large number of potential sources, it seemed a pity to leave it there – especially with free near-instant communications available via the Internet. So we wrote a Request for Proposals and sent it via email to 112 companies. We posted the bulk of our requirements on our web site where they could be downloaded by each vendor. Our project required MS Visual Basic and SQL Server skills. The team needed to include one senior developer, with the rest of the team having two to four years experience. We asked for time and material bids, and left considerable leeway for the vendors to propose their staffing and schedule plans.
Within three weeks, we received 54 proposals.
A Supplier vs. A Partner
One of our requirements was that the developer had to take a substantial part of his compensation in equity. We wanted a partner with a long-term stake in the success of the project. The development process would also constitute an investment by us in the project staff, as they would know our product suite thoroughly by the time they finished. We wanted to retain as much as feasible of that investment toward product maintenance and the development of future releases. Having the developing firm as a shareholder was deemed to be a way to accomplish that.
That requirement was spelled out clearly enough in the original RFP, or at least we thought it was. Many did not understand the requirement, or perhaps thought it was a negotiating stance. It wasn’t, and several proposals were ultimately eliminated due to this issue. Still, 40 firms said they were willing to take half or more of their compensation in equity.
Narrowing Down the Field
The firms that made the first cut were of these nationalities:
My personal preference at the outset was to find a partner in the former Soviet Union. It seemed so obvious. Here was a technologically advanced country going out of business, with millions of unemployed but well-educated workers, and 10 companies to choose from. Three issues changed that preference:
The short list finally consisted of six Indian firms, each of which could be visited within one day of a central point. Billing rates of the short-listed firms could not be compared with precision because of variations in proposed staffing and payment plans, but tended to range from under $10 to $15 per hour. Others that didn’t make the short list ran as high as $60. The extreme range created some questions at the time as to where reality lay, as all firms were essentially drawing from the same labor pool. Subsequent experience and relationships have provided assurance that the top quoted rates were beyond excessive. The very best firms continue to provide staff with the skills we were seeking in the $10 to $20 per hour range. Incidentally, these rates include all equipment, materials, etc.
I flew to India in February ’99 and visited four sites. A fifth was scratched because of its inaccessibility. A primary concern was the ability of the organization to take on the project and finish it. One firm with especially skilled people was rejected because more than half of their entire staff would be assigned to this project. They could just as easily be pulled off if a more attractive project come along; say one that was paying 100% cash.
The award was made to a firm with more than 100 developers present at the site I visited. I also inspected their massive in-house power generator, a necessity to keep computers operating during frequent public power failures. We finalized terms and signed a contract on the last day of my visit.
I continue to spend a week on site at roughly six-month intervals. We require a monthly report from each staff member who is billed to us, and a summary by the team’s manager. One objective of the periodic visits is to provide an audit of billing through discussions with each individual developer. In the interim, we use ICQ, daily if necessary, for chat sessions and data transfer.
To conclude the ‘personal experience’ component of this paper, we were lucky and we did well. Our contractor has become a real partner with a significant stake in the company and determined to help us succeed. We expect to continue the relationship through both maintenance and development of future product versions.
The Bigger Picture: An Overview
There has been a severe and growing shortage of software developers in the United States for several years, and the Information Technology Association of America predicted that U.S. employers will be unable to fill nearly 850,000 information technology jobs this year. The H-1B visa program was created by the Immigration Act of 1990 to allow U.S. companies, universities, hospitals, state and local governments, and other employers to hire skilled foreign nationals for up to six years. For fiscal year 2001, the cap on H-1B visas for foreign nationals was increased to 195,000. The shortfall of 655,000 has been temporarily eased by the economic slowdown following the recent terrorist attacks, but they have not erased it and the long-term trend, which has been developing for years, appears likely to continue.
Why India: Government and Industry Initiatives
There are some obvious reasons why India is the primary country picking up the slack in software development:
The Indian Government has taken deliberate steps to improve on their native advantages. Declaring that software is “one of the extreme focus areas for growth of exports”, it established The Software Technology Parks of India (STPI) to promote industry development. There are 21 STPI centers throughout the country, each providing the infrastructure necessary to operate a software venture. For example, STPI Noida, near New Delhi, promotes itself thusly (copied from their web site; www.stpn.soft.net):
“STPI-Noida offers Built-Up Ready Made Premises with all the Infrastructure which has most of the locational advantages like the Proximity of educational Institutions, Hotels, Restaurants, Hospitals, Domestic and International Airport, Banks, Excellent Residential Complexes, Sports Clubs which are all located in and around of the Capital - New Delhi- which becomes the advantage for IT Industries to set-up a STP-Unit.”
Businesses that do not wish to locate within the STI Center itself can nevertheless take advantage of tax and other benefits mentioned below, by becoming an “STP-Unit” in their own premises anywhere within the “Noida Jurisdictional Area”, which covers virtually the whole of Northern India.
Many of the benefits of the STP program can be appreciated only in the context of Indian business, but the following are understandable in any society:
For a summary of additional and continuing government initiatives, visit the web site of India’s National Association of Software and Service Companies, www.nasscom.org.
The Issue of Quality
The software industry of India has been compared to the automobile industry of Japan, and Watts S. Humphrey of the Software Engineering Institute (SEI) at Carnegie Mellon University has been called the Deming of India. Just as the quality of Japan’s automobile production was guided by an American guru to surpass that of Detroit, so India’s software quality may be passing that of the US, guided by the teaching of another American.
Mr. Humphrey established a standard of measurement of software development practices, named the Capability Maturity Model Ladder (CMML), commencing at ‘ad hoc’ and progressing to ‘mature’. Some 60% of all the companies worldwide that have reached maturity on the SEI ‘ladder’ are Indian.
Last February saw the dedication of the Watts Humphrey Software Quality Institute (www.watts-sqi.org) in Chennai, India. Credit for founding the Institute is usually given to Girish V. Seshagiri, an Indian native who is CEO of US-based Advanced Information Services Inc. Mr. Humphrey attended and participated in the dedication ceremony.
The Institute is reportedly training a faculty of 400 to conduct courses to executives, project managers and engineers in the methods promoted by Mr. Humphrey and SEI.
Following are examples of the dozens of Indian subsidiaries of foreign firms whose role is to develop software on behalf of their parent companies. Note the Korean firms among them.
3 Com Asia Ltd.
Adobe Systems India Pvt. Ltd.
CA Computer Associates India Pvt. Ltd.
Citicorp Overseas Software Ltd.
Citrix Software Ltd.
eSamsung UMIT Infotech Pvt Ltd
Hewlett Packard India Software Operation Pvt. Ltd.
Honeywell India Software Operation Pvt. Ltd.
IBM Global Services India Private Limited
Intel Asia Electronics, Inc.
Lucent Technologies India (P) Ltd.
Microsoft Corporation (India) Pvt. Ltd.
Onward Novell Software (I) Ltd.
Oracle Software India Ltd.
Philips Software Centre Pvt. Ltd.
Rational Software Corporation (India)
Samsung Electronics Co. Ltd.
Santa Cruz Operations, Incorporations
Texas Instruments India Limited
Xerox Modicorp Ltd.
The following chart shows the growing value of the Indian software market.
No company with major software development expenses can be competitive today without operations in India or a comparable offshore location, and there are very few alternatives. The low cost of computers accompanied by the migration of applications to the Internet has created demand for software that is far beyond what the West can produce. Software development is, after all, a labor-intensive process, and Adam Smith’s law of comparative advantage dictates that it will migrate to where that labor is available.