Friday, October 21, 2011

A new STPI to promote innovation?

    The purpose of STPI (Software Technology Parks of India), as stated on their website is to “promote the development and export of software and software services”. Under the STPI scheme, a software firm can get a ten year tax holiday on their profits from export of software and software-related services. The tax holiday scheme is obviously to promote and encourage the software export business, and has been quite successful. According to NASSCOM estimates, the IT and BPO sector accounts for around 6% of India's GDP in 2009 and 2010. However, a troubling trend has been that these Indian software firms have been concentrating largely on software services, rather than invest in software products. A quick look at TCS and Infosys' revenue break up from Q3 2011 (Sep 30, 2011) and 2010 whole year, respectively, reveals this disproportionate skew towards services.

To understand the reason for this services-skew, it may be necessary to first define how a software product and a software service differ. A software product is, essentially, licensed. The buyer does not get the intellectual property rights to the software, but a license to use it. The seller of the software product is still responsible for defending the IPR in the software, as well as indemnifying the buyer of any infringement claims by a third party. A software service, on the other hand, does not involve any kind of licensing. It is a pure sub-contracting of software activities from the buyer to the provider of these services. In these cases, usually, the service provider does not own any intellectual property rights to the software being delivered. The service provider gets paid as the service is being produced.

The reason for the disproportionate skew towards software services is due to this very fundamental difference in software products & services. A company building a software product needs to make some risky investments. Firstly, the product has to be developed to some level of completeness before it can be sold. This means that software resources have to be used before they can be billed. Here is the first risk... if the company cannot sell the product, then their investment goes down the drain. Secondly, selling software products means that the company has full responsibility for the IPR in the product. If someone sues the buyer, claiming that the software product infringes on their patents, then the seller has to step in to depend it's IPR, or bear the consequences. A level of software IPR maturity is needed, besides investing in a legal department.
Software services, on the other hand is a low-risk business. Software resources are allocated only when the buyer has signed on the dotted line (or is very likely to do so soon). The risk is decidedly lower than in software products. Besides, the IPR is usually (but not always) the buyer's responsibility. Many of the software subcontracting contracts state that any IPR that comes into the picture due to the execution of the software services would transfer to the buyer. So, in layman's terms, software services is a low-risk way of making a quick buck! But is there anything wrong with making a quick buck? Surely it's in the best interests of the shareholders?
The problem arises when one analyzes the business model of software services. Here I try to use a very common strategic management framework, known as Porter's Five Forces model. While this model is used to analyze an industry, and not a business model, it can be tweaked to consider the not-too-hypothetical case that all companies in the industry are providing software services only.

  • Threat of new entrants – Very high, as the barriers to entry are low. The business model can be easily replicated by any software firm in a low cost geography.
  • Bargaining power of buyers – High, since they usually are larger and more influential firms than the sellers.
  • Threat of substitute – Medium to Low. The substitute here being that the buyer executes the software activities using their own internal resources. The threat is low since many customers of Indian software firms do not have the geographical low cost advantage.
  • Rivalry – High. It is well known that as customers try to drive down their costs, they would look at evaluating multiple software suppliers and selecting the ones that best matches their cost vs value target.
  • Bargaining power of suppliers – Medium to High. The suppliers for a software services focused company are the human resources. Engineering graduates from our many engineering colleages and institutes of technology, as well as employees from their rivals who are “looking for a change”. Attrition has been a problem with most Indian software companies, with 15% to 25% being the norm.

This model confirms what many people already know... if all software companies focus on software services, then the industry will not remain profitable for long. It would end up with too much competition, and too little differentiation to remain profitable.
Software products, on the other hand, have many advantages when compared to services. While it is indeed riskier than services, software products offer several benefits. The firm needs to invest once, but can sell the same product to many buyers, with little or no customization. This breaks out the problem of revenue growth linked to number of employees (a problem that services-focused companies acknowledge). Revenue now, would not be tied to the number of software developers, but with the utility and innovation of the software product. Moreover, the barriers to entry are a bit raised now. A competitor has to invest a reasonably high amount of R&D effort in order to come up with a competing software product. A higher premium can then be charged for the product, as there would be not too many competing substitutes. Motivation to innovate would increase, as the more innovative the product the more profitable it generally is.
Hence, I strongly feel that a new Software Products Technology Parks of India scheme should be introduced, replacing the existing STPI and even SEZ schemes when it comes to the Software Industry. The SPTPI scheme should provide tax breaks only for export of software products, and not for services. Maybe even provide a tax break for domestic sales of software products, thus ensuring that firms cater to domestic demand. It would be fairly easy to check if a particular contract involves a software product or service... the Income Tax Department need just check if the seller owns the IPR or not. Quite simple & straightforward!

No comments: