Actionable insights on equities, fixed-income, macros and personal finance Start 14-Days Free Trial
Actionable investing insights Get Free Trial

The Inflexion Point for the IT Service Industry (Long)


This is a long post. It’s meant as weekend reading, and I would love to hear your thoughts: deepakshenoy at or comment on this page. I’ve written too much, but I’m sure I’ve not written enough. 

Indian traditional IT sector jobs are going through a massive inflexion point. I’ve said this many times, even on a TV interview with ET Now, that IT service companies are in a soup: they are darlings of inefficiencies rather than efficiencies.

Whoa. Explain.

The IT business has traditionally consisted of billing for headcount. That means you put X people on the project and you get revenues that are a multiple of X. That multiple can be $25 per hour, $100 per hour or even as low as $7 per hour, depending on the kind of work needed. Typically, the idea is to get work done by “freshers” or people with low levels of experience, with oversight by someone who’s experienced (a “project lead”).

Most work is process oriented – you have a checklist to set up and maintain servers, or manage infrastructure requests. Or you take IT infrastructure of a large company which has thousands of computers, servers, backup jobs etc. and do it remotely for way cheaper. (read this TCS Case Study) Or, and maybe what might survive, build software solutions for customers. And then, implement solutions like SAP or Siebel or Oracle or such solutions for customers.

For a long time, US and European companies were happy to send this work to India, because people locally were too expensive to do this kind of work. India was cheaper, even if it put more people on the job than absolutely required. There was no need for efficiency for the likes of TCS and Infy. Even at $20 an hour and Rs. 36 to a dollar (in 1996) the income potential was Rs. 115,200 a month for a 160 hour employee on a project. This is when I started off in IT services, getting paid the grand sum of Rs. 6,500 per month. Even at Rs. 10,000 per month, the company was ludicrously profitable – for every person on a project it could have five on the bench and still make 50% gross margins. (Also, remember, there was no income tax for software exports then).

This skewed the dynamics. The higher margins allowed IT Service exporters to offer higher relative wages (to other sectors), way lower costs to customers (whose equivalent costs were $50 per hour plus insurance plus holidays plus what not) but still maintain great profitability. Guess what? Everyone and his uncle joined the IT service business. In my batch in KREC/NIT Surathkal, in 1996, more than 200 out of a batch of 360 joined the likes of IBM Services, TCS, Infy and the like. It didn’t matter what branch you studied in. (And I made some pocket money by teaching my fellow students COBOL for a test they had to pass to land the job)


Over time, the rupee weakened to 45 in the wake of the dotcom bust, and these companies continued their hiring; they might have lost some business, but as the global economy recovered, what it was for the other countries was a jobless recovery in many ways; the jobs then lost were outsourced to India because it was just cheaper. This was the 2003 to 2008 period when the IT Business absolutely thrived.



Within the IT companies, the path to glory was common:

  • Go through training program which was essentially what-else-to-do-when-on-bench
  • Be assigned to a project. You’re “sold” to a client saying you know X and Y, when you probably don’t
  • Piggyback on project lead to learn stuff that’s important for the project
  • Code, document, test, learn to write proper emails
  • Get promoted to project lead, when you’ll be helping others do the above steps
  • In three years, make project manager. Now you don’t “code”. Yuck.
  • Your tools of choice are word, excel, powerpoint and outlook.
  • Scale up the management tree of the organization, with your power being how many people report to you

This was a given. You work in a company for three years and still code? You must have no skills whatsoever. The idea was that people would mature this way, and many did, and scaled up all the way to even CEOs. As the business grew, the “workers” – the freshers hired at low salaries – were extremely cheap, and hired in the thousands. The guys moving up the ladder made for the middle tier, and so on. Natural attrition into other IT companies kept the middle layer lean (you could double your salary moving to a different job), and the bottom layer saw enough attrition and massive hiring to keep growing. A “pyramid”.

The Paradigm Just Shifted

While this was happening, the carpet was getting yanked out from under the IT companies’ legs, in slow motion. Infra management was changing rapidly, and moving to the cloud. Companies with 3,000 servers could move them to an Amazon or Rackspace or private cloud, which was remotely managed by a handful of very talented engineers and massive layers of automation. Siebel and SAP offered customers cloud based solutions where the whole thing was already setup – all you needed to do was configure your company’s details in, which was complex, but needed 5 people rather than the 50 person teams that would have been needed.

A project manager, a client manager and what not who were usually assigned to customers were no longer efficient; they would now need to manage more projects made of smaller teams.

Alongside, the massive hiring and natural scaling up of the hires in the 2003-2008 period meant in 2014 that you had a large number of people in the middle management as project managers and senior consultants and so on.

The natural “pyramid” had gone out of shape. There were just too many in middle management.[Tweet This]

Salaries, with high inflation, had also gone up substantially. The mid-managers were drawing upwards of Rs. 15 lakh per year, and many were quite happy to see the regular increases year after year.

And then came the assault of the product companies and funded startups. Companies like Google, Facebook and Amazon were hiring for their stuff and they weren’t doing projects for others – they were using the teams to build their own products. The likes of Flipkart and Ola and Zomato also were hiring the best people from colleges and so on. The middle managers couldn’t work here; they had no serious programming skills left, and product management was a different beast altogether. No employee handbooks, no ISO 9000 and no real comfort zone of saying you wanted to be allocated to a different project. So the IT folks, by and large, didn’t join the startup revolution, and distanced themselves from the newer technologies that involved doing great things with smaller teams. And that has been a fatal mistake.

Other IT companies emerged that made themselves niche. A Mu Sigma, for instance, could charge a huge premium, at hundreds of dollars per hour, because they have massive analytics investments that helps their teams deliver solutions to customers faster. Druva, a software company started in Pune, offers automated transparent backup products for computers, mo
bile devices and servers, on the cloud, at the fraction of the cost it would take to hire a large IT service company.

Worse, the rupee at 63 would also be no help. The Rupee has actually appreciated against the Euro, the Yen and the British Pound. The Ruble just devalued from 30 to 60 in a year, and they have great programmers. Against other Asian and Latin American currencies, the Rupee has appreciated substantially.

The EndGame Begins: The Stalled Salary Hikes and IT Firing

As the inefficiencies became apparent, companies tried different ways to do things. Infy, for instance, stopped all salary raises with many quarters of tepid growth. This gave them “natural attrition” as the mid managers found jobs in other IT companies. Some smartly shifted to the more efficient startups.

Companies in western markets have started to shift away to other countries who now have a cost advantage. Faster and smaller projects are in vogue, with the cloud being a huge factor in adoption of customizable products like Salesforce, and has even prompted the SAPs and and Siebels to move to the cloud.

The IT companies had very few product investments. They became nobodies in the space that grew the fastest – search, online advertising, social media and anything to do with cloud/open-source technologies. This would have cannibalized their existing customer base – imagine going to a customer and saying that instead of having a large team of 20, let’s just use this software that’ll do the same work with 5. Eventually the customers saw the cost advantages and reduced sizes anyway.

The fat middle layer, and customer moves to more efficient software solutions meant things had to change and fast. While Infy lost much of its middle layer naturally, the other IT giants were still saddled with them. Then Wipro announced that they wanted a lean middle, moving from a pyramid to “an hour-glass kind of structure“, and would cut workforce by about 33%. Recently TCS is in the limelight for actually firing underperformers (the horror!) which will cut their workforce by over 25,000 people, if numbers are to be believed.

This is just the natural endgame and TCS is likely to be leading the pack.

Where Do People Go?

Roubini talks about this: Where Will All The Workers Go? This was a question they asked in the US when manufacturing jobs left for China, and then service jobs left for India (BPO, IT Services and so on). This is insane. Technology will displace people from jobs, and has done so forever. We IT people took away the jobs of some manual process worker somewhere. Cars took away the jobs of horse drawn buggies. Mobiles destroyed the whole “Public Call Office (PCO)” ecosystem. Online e-commerce will hurt retail. Uber and Ola will hurt the autorickshaw.

The outrage of the IT worker is largely misplaced. The technology we build is designed to cannibalize everything. Tomorrow people will get programs that write themselves without the need for complex programming too.

But the question should be human. Where do these people go? 15 to 20 years of experience and they will get that pink slip.


There is a point to saying people need to be let go better – giving them a month (or six!) of severance rather than having them resign, helping them find other jobs, and in general, accept that you’re not just hurting one person but their families and kids and so on. And it’s incredulous to fire people without providing them money, when you’re hugely profitable, so it’s all the more reason to give people a few more months of money as severance.

What Really Needs To Change

But there’s more to this than a few months of severance. Middle managers – and I know I am in that exact age bracket – need to reinvent themselves. This could be one or more of:

  • learn a new skill – either back to code and processes in newer technologies, or in a completely different domain. This could take months or years, but it’s necessary.
  • invest and create alternate sources of income
  • keep debt manageable so a job loss will leave you with at least a year’s expenses in the bank
  • stay humble: the people who reported to you could be your next boss.

To stick with this might be a better approach than going out on the streets and demanding that TCS hire them back while not hiring freshers instead. Because that is simply not going to happen; if you force it, the IT industry will shut down even faster. Unions and government intervention are the wrong answers.

The broader impact is on real estate and discretionary spends (like cars) and local economies (like Bangalore). But that will need a different post altogether.

I just wanted to highlight the inflexion point in the industry and how it came to be. It was not something you can just blame on higher management. They didn’t see this coming, because to have acted proactively meant loss of revenue and that would have gotten you fired then, not now.

At one level this is a bubble that is now going bust; the great Indian IT service bubble. It was fuelled by the relatively low tax rates by the government – such companies didn’t have to pay taxes till nearly 2010. This maintained that all the juiciest jobs were in IT services. Not in products – you couldn’t get the tax benefits if you built software for India. That created extremely skewed dynamics for everyone that graduated from an engineering college between 1996 and 2008. We could blame the government, but that wouldn’t do us any good either.

It’s time to reinvent ourselves.


Like our content? Join Capitalmind Premium.

  • Equity, fixed income, macro and personal finance research
  • Model equity and fixed-income portfolios
  • Exclusive apps, tutorials, and member community
Subscribe Now Or start with a free-trial