I guess because IT firms typically bill by time and material, there is a perverse disincentive to adopting AI in their day to day work! Since the time savings need to be passed on to the customers !
And so I guess they use AI only in their AI practices ? As an outsider this sounds weird to me!
Strong piece. The most important shift you highlight is the inversion of adoption AI winning at the individual user level first and then forcing enterprise scale-up. That is structurally different from cloud/ERP cycles and does compress timelines.
That said, two constraints may slow the “P(AI)N” curve in India more than markets currently assume:
• Wage arbitrage still mattersAI must beat already low offshore costs, not US salaries.
Near term, this looks more like headcount deceleration + margin expansion rather than outright disruption. Medium term is where the real pressure builds especially in commoditized coding, testing, and L1/L2 work.
The signal to watch is simple:
Revenue growth vs. net headcount over the next 4 to 6 quarters.
If that gap keeps widening, the thesis moves from narrative to inevitability.
Interesting piece as always. But I found it bit strange you used SaaS and IT Services inter-changeably. All the big Indian IT ones are IT services not really SaaS. How AI will playout will be very different for these two categories. Well even SaaS is not one category, so much differentiation within them.
Check this out below on why SaaS-ocalypse is over stated as market seems to be running on emotions on short term,
Short term, the IT firms will be fine because these firms will have locked-in contracts for 1-3 years but I think medium term the pressure will build when the clients don’t renew the contracts (Alex Campbell from Campbell Ramble had a good point on this, that software isn’t dead but the service economy built on it is and he was talking specifically about listed Indian IT players)
Unless these firms pivot to idk some hardware business (data centres i guess, which might be labor intensive to build but it is not labor intensive to operate) or migrate into something more high value like AI consulting (pretty sure that no way requires this much head counts), it is difficult to see all the Indian IT players coming out on the other side unscathed.
It’s of course not just a problem for IT companies but the whole knowledge outsourcing economy at large. But the whole AI thing still has some level of Knightian uncertainty to it, so we’ll have to see how it plays out in real time as it’s difficult to separate the real from the hype in this space (though I think it will have some adverse effects here)
This is such a well thought-through piece. I really liked the way you explained how this wave is different from previous ones. One of my European colleagues asked me how/why India's large IT services firms aren't in panic mode yet. I didn't have a clear answer to this. Maybe they want to project a more resilient view to their shareholders but the penny will drop very soon (if it hasn't already). Thoughts?
I guess because IT firms typically bill by time and material, there is a perverse disincentive to adopting AI in their day to day work! Since the time savings need to be passed on to the customers !
And so I guess they use AI only in their AI practices ? As an outsider this sounds weird to me!
Like Ravi Shastri says, something’s got to give.
Strong piece. The most important shift you highlight is the inversion of adoption AI winning at the individual user level first and then forcing enterprise scale-up. That is structurally different from cloud/ERP cycles and does compress timelines.
That said, two constraints may slow the “P(AI)N” curve in India more than markets currently assume:
• Wage arbitrage still mattersAI must beat already low offshore costs, not US salaries.
• Enterprise reality (data hygiene, auditability, exception handling) remains the hidden friction layer.
Near term, this looks more like headcount deceleration + margin expansion rather than outright disruption. Medium term is where the real pressure builds especially in commoditized coding, testing, and L1/L2 work.
The signal to watch is simple:
Revenue growth vs. net headcount over the next 4 to 6 quarters.
If that gap keeps widening, the thesis moves from narrative to inevitability.
Interesting piece as always. But I found it bit strange you used SaaS and IT Services inter-changeably. All the big Indian IT ones are IT services not really SaaS. How AI will playout will be very different for these two categories. Well even SaaS is not one category, so much differentiation within them.
Check this out below on why SaaS-ocalypse is over stated as market seems to be running on emotions on short term,
https://www.readtrung.com/p/how-does-docusign-have-7000-employees?
Short term, the IT firms will be fine because these firms will have locked-in contracts for 1-3 years but I think medium term the pressure will build when the clients don’t renew the contracts (Alex Campbell from Campbell Ramble had a good point on this, that software isn’t dead but the service economy built on it is and he was talking specifically about listed Indian IT players)
Unless these firms pivot to idk some hardware business (data centres i guess, which might be labor intensive to build but it is not labor intensive to operate) or migrate into something more high value like AI consulting (pretty sure that no way requires this much head counts), it is difficult to see all the Indian IT players coming out on the other side unscathed.
It’s of course not just a problem for IT companies but the whole knowledge outsourcing economy at large. But the whole AI thing still has some level of Knightian uncertainty to it, so we’ll have to see how it plays out in real time as it’s difficult to separate the real from the hype in this space (though I think it will have some adverse effects here)
Excellent post though as always
This is such a well thought-through piece. I really liked the way you explained how this wave is different from previous ones. One of my European colleagues asked me how/why India's large IT services firms aren't in panic mode yet. I didn't have a clear answer to this. Maybe they want to project a more resilient view to their shareholders but the penny will drop very soon (if it hasn't already). Thoughts?