Software and Hard choices

In this world there are many countries which are endowed with natural resources. Some of them were content to extract these resources in its most basic form and sell and some countries built up industrial bases that add value to these natural resources. The latter prospered and the former often stagnated especially because this abundance in one area acted as a disincentive to growth in other areas. The stagnation might have happened when the resources ran out or more prices dropped or with falling demand or arrival of more competitive suppliers.

We have the potential for a somewhat similar problem in our software industry which is growing to be a significant sector of strength and opportunity for India. This could be on account of some factors that are holding us from rising above mediocrity. If we don’t address these, eventually we may end up paying price for this as a nation.

Demand Growth: There is an increasing acceptance for use of Information Technology in most sectors of economy in India; e-Governance, hospital administration, educational institutions, manufacturing industry banking and financial services. This offers huge potential for the IT industry. In India, our focus and strength is in application development more so in building bespoke applications and less in hardware and system software. This, in addition to the outsourcing opportunities for undertaking developments for international clients, creates a burgeoning demand for software industry in India.

Customer Awareness: However the in house appreciation among this large consuming sector within India is quite primitive and therefore unable to demand sophistication and quality from their vendors. On the other hand, a significant part of the our outsourcing contracts are for relatively low-end programming as per the solutions defined by in house CIO and his team or based on the solutions architectured by high-end international consulting companies. Thus there is very limited incentive among the programmers to worry about the performance of their output but encourages to focus on functionality and features.

Increase in Diversity: Fast paced  introduction of new tools, more sophisticated databases, and diverse programming languages encourages the programmers to be familiar with this diversity than develop deeper expertise in any of the systems to extract the performance efficiency. The are happy to be "Jack of all trades but Master of none"

Leaping hardware technology: The hardware is progressing in sophistication so fast that it is reducing the cost for processing same volume. When shoddy system design and program quality put strain on performance with increase in volume, the developers recommend more iron. Since these new machines process more volumes, the senior management of customers the gets lulled by the apparent reduction in cost and ask few questions because in most other areas they are used to increase in cost with increase in volume.

Impact of IT cost: Major consuming industries for our software developers are manufacturing, financial services and government. For manufacturing IT costs as a factor of their total cost is relatively low. As they get their revenue from selling products (cars to drugs to chemicals to consumer durables and non-durables) their attention is more on the technology for making and selling their products. IT is seen as an enabling component or a fad and gets lesser attention on performance.

Similarly in Financial services with the revenue being a function of the value of transactions than the number of transactions they pay less attention in cost per transaction. In e-governance application also the situation is same with less attention to performance but more on functionalities.

Measurement of Performance Efficiency: In software industry, there hardly few good measures of performance efficiency that are widely used and fewer bench mark values against which performance can be measured; especially when it comes to cost per transaction. With nobody taking the ownership of the total solution, when problems occur, providers of each component like hardware, system software and networks point finger on each other. Even very few system integrators own the performance of the total solution, and but shift performance responsibly to the components. (Read “Learn to count- both Blessings and Failures” for some more thoughts on this)

So What?

The more discerning users for whom cost per transaction is critical like those of EBay, Google and Facebook have their own in house team whose focus is to squeeze out efficiency and reduce the cost and hardly few of our software service providers do any meaningful work for this high-end computing.or develop unique solution

There are many smart Indians in the development teams out there working on such solutions. But our domestic software industry is often drunk with the $ from being technicians and cybercoolies and not architects and engineers. The user requirement study is "Tell me what exactly you want met to automate, I will program it"  and not "Tell me what your dream is, we will work together on how technology can make it happen"

But if there is a large demand for low and middle end computing should we not supply it to earn our dollars and be happy? Of course yes. What is the risk in this?

The transaction volume in our consuming industry is growing leaps and bounds. The shoddy designs will soon show its weakness in handling this ballooning volume. The users will ask for more performance. They will ask for sophisticated analytics, pattern recognition, trend analysis and statistical modeling with the goldmine of data that has been amassed. Then our conventional solution providers will have nothing to offer as against those companies who have given more attention to high-end computing and more sophisticated model building often having outsourced the run-of-the mill programming to us.

That is why in this time of plenty there is a need to invest in building high performance solutions, develop a culture of fine-tuning systems for performance, learn to offer solution to a client's problem and not just code what he ask for, develop capabilities for building models and so on. Even in our public policy we should start factoring this and shift the incentives from profits of software export to investment and profits from high-end products or solutions development for the global markets.

Talent without discipline is like an octopus on roller skates. There's plenty of movement, but you never know if it's going to be forward, backwards, or sideways.H. Jackson Brown, Jr.


If you liked this post, share it with your friends.