Indian IT’s real problem isn’t the quarter — It’s the playbook

Every earnings season, the same ritual repeats itself. Analysts model sequential growth to the decimal point, companies hedge their guidance bands, and headlines fixate on whether Tata Consultancy Services or Infosys beats or misses by 50 basis points. This year, as the sector heads into its fourth straight year of low single-digit growth, that ritual is starting to look beside the point. The real story isn’t another soft quarter. It’s that the operating model Indian IT built its identity on for two decades has quietly stopped working, and nobody in the industry has found a convincing replacement yet.

For most of its history, Indian IT sold predictability. Clients handed over large chunks of routine technology work, armies of engineers delivered it at a fraction of onshore cost, and revenue grew in a fairly straight line as headcount grew alongside it. Growth was linear because the model was linear: more bodies meant more billable hours, which meant more revenue. That equation is now broken from both ends. On one side, artificial intelligence is compressing the cost of the very work that used to justify large teams, shaving an estimated two to three percentage points off core services revenue every year simply because tasks that once required ten people now require three. On the other side, the new revenue that AI is supposed to generate, helping enterprises design and run intelligent workflows, remains a promise measured in a five-year horizon rather than a line item on next quarter’s income statement.

Sitting in between those two forces is a workforce and a client base that were both built for the old equation. Vendors are chasing an identical, shrinking pool of large consolidation deals, which explains why competitive intensity and pricing pressure have both risen even as growth has stalled. It also explains why margins are under quiet siege: wage inflation, the cost of standing up new AI capabilities, and the expense of ramping large deals are all rising at the same time that pricing power is not. A weaker rupee is one of the few tailwinds left, and currency depreciation is a strange thing for a sector built on engineering excellence to be leaning on.

What makes this moment different from past downturns is that there is no obvious cyclical trigger to point to and wait out. Previous slowdowns had a clear villain, a global financial crisis, a pandemic, a rate-hiking cycle, with an implicit promise that spending would snap back once conditions normalized. This time, the drag is coming from three sources simultaneously: a genuinely uncertain macro backdrop, geopolitical shocks like the West Asia conflict that are now visibly denting large clients’ top lines, and a technological shift that is deflationary by design rather than by accident. AI does not become less disruptive to services pricing once the macro cycle turns; if anything, adoption curves suggest the opposite. That is a structural condition, not a weather pattern to be waited out.

This has quiet but significant implications that rarely make it into earnings-day headlines. For employees, it likely means the era of steady headcount growth as the default hiring strategy is over, replaced by a smaller, more skilled workforce oriented around fewer, larger, AI-inflected deals. For mid-sized firms, it is oddly a moment of relative advantage: their smaller base makes double-digit-adjacent growth achievable through large-deal ramp-ups in a way that is now nearly impossible for the giants, who need proportionally far more new business just to keep pace. For investors used to treating IT services as a defensive, cash-generative sector immune to disruption, the last four years should prompt a rethink of that assumption altogether, since resilience that shows up mainly in one vertical, banking and financial services, while telecom stays soft and manufacturing turns patchy, is a narrower kind of resilience than the label implies.

None of this means the sector is in terminal decline. Demand for orchestrating AI within large enterprises is real, and Indian providers still have scale, delivery infrastructure and client trust that competitors would take years to replicate. But treating this earnings season as a temporary dip to be endured, rather than as evidence that the entire growth model needs rebuilding, risks missing the actual transition underway. The number to watch this quarter is not the guidance band. It is how much of that AI-related revenue pipeline, still mostly discussed in the future tense, starts showing up as an actual number on the income statement.

CT Bureau