The IT services industry has a new favorite number: 25%. Hexaware's CEO told the press that AI could slice IT project costs by a quarter. The market nodded. The headlines wrote themselves. But I've spent enough time auditing balance sheets and order books to know that when a CEO hands you a clean, round number, it's not data. It's a positioning statement.
Let's cut through the noise. This isn't about AI efficiency. It's about a mid-tier IT services firm trying to survive a structural shift that is dismantling its core business model. The 25% figure is a defensive narrative, carefully calibrated to manage client expectations and investor sentiment simultaneously. We are not predicting the storm; we are shorting the rain.
Context: The Human Arbitrage Business Model Is Dying
The traditional IT services model is simple: hire cheap engineers in India, bill them out at Western rates, and pocket the spread. For three decades, that labor arbitrage was the engine of TCS, Infosys, Wipro, and a hundred smaller players like Hexaware. It worked because the cost of a developer in Bangalore was a fraction of one in Boston, and the quality gap was negligible.
AI breaks this model at its foundation. If a language model can write, test, and document code at a fraction of the cost and time of a junior developer, the arbitrage disappears. You are no longer selling human hours; you are selling a commodity that is rapidly depreciating. The CEO's 25% cost reduction claim is an admission that the floor has fallen out. He is not announcing a technological breakthrough; he is acknowledging that the client already knows the leverage. The client knows they can get the same project done cheaper with AI tools, either through a competitor or by building it in-house.
The global IT services market is roughly $1.2 trillion. If AI delivers a 20-25% efficiency gain across the board, that releases $240-300 billion in potential value. The critical question is not whether that value will be created. It will. The question is who captures it: the client, the IT services firm, or the AI infrastructure providers like OpenAI and Microsoft. The CEO's statement suggests he is preparing to hand a significant chunk of that value to the client to keep them from walking.
Core: The 25% Number Is a Pricing Signal, Not a Cost Report
Let's break down the mechanics. In a typical IT project, labor costs are 60-80% of the total. AI can plausibly reduce the effort required for coding, testing, and documentation by 20-30%. So the 25% headline figure is within the realm of industry estimates from McKinsey and Gartner. But here is the catch: if you cut the cost of delivery by 25% but your billing model is based on time and materials, your revenue also drops by 25%. Your profit margin stays flat. You have just run faster to stay in the same place.
The CEO's mention of balancing productivity gains with pricing strategy to maintain margins is the tell. It's a direct admission that clients are demanding a share of the AI dividend. The IT firm is caught in a vice: if they don't cut prices, they lose the contract to someone who will. If they do cut prices, they need to cut costs even faster to protect the bottom line. The 25% number is the opening bid in a negotiation where the IT firm is already on the back foot.
This is not hypothetical. We have seen this movie in other industries. When algorithmic trading cut the cost of equity execution, brokerage commissions collapsed to zero. When streaming replaced physical media, the margins vanished. The value created by efficiency gains rarely accrues to the intermediary; it flows to the end consumer. The IT services industry is facing its own race to zero. The firms that survive will be those that shift from selling hours to selling outcomes, from providing staff to providing guaranteed results. That is a far harder business, and most are not prepared for it.
Contrarian: The Real Battle Is Between Scale and Agility
Everyone is focused on the tech. I'm focused on the competitive structure. The AI transition is not a level playing field. It is a war of attrition where scale matters more than ever. Accenture has announced $3 billion in AI investments. Infosys has built an entire platform around it. Wipro is spending a billion. These are moats built with capital. Hexaware, with roughly $1 billion in revenue, cannot outspend these giants. It cannot out-recruit them for scarce AI talent either.
But the threat is not just from above. It's from below. AI-native startups like Cursor and Devin are building tools that could make the traditional project-based outsourcing model obsolete. Why hire a team to build an app when you can prompt an AI agent to do it? This is a two-front war. The giants are squeezing Hexaware's price advantage with their AI scale, while the startups are attacking the very concept of the service. This is the classic innovator's dilemma, and the mid-tier players are the most exposed.
The CEO's public stance is a signal to two audiences. To clients: we are AI-fluent, we can deliver the cost savings you demand. To investors: we are not dinosaurs, we are transforming. It is a narrative designed to buy time. But narratives do not pay dividends. The market will eventually demand proof in the form of margin expansion or revenue growth. If the cost savings are all passed to the client, the stock will not react. If the firm can retain some of the savings, they might see a short-term bump. My bet is on the former. Client bargaining power is the dominant force here, and it is only increasing. Leverage doesn't care about feelings.
Takeaway: Watch the Billing Model, Not the Press Releases
The next 18 months will be brutal for the IT services sector. The CEOs will talk about AI-driven transformation, but the data that matters is in the contract structures. Watch for a shift from time-and-materials billing to outcome-based pricing. Watch for the first major price war. Watch the employment numbers in India's tech hubs for signs of a structural contraction. The 25% figure is just a number. The real signal is the underlying admission: the game has changed, and the old rules no longer apply.
The firms that survive will be those that weaponize AI to create proprietary tools and solutions, not just to cut headcount. The ones that simply use AI to do the same work with fewer people will find themselves in a deflationary spiral. I've seen this play out in crypto markets, where protocols that fail to find product-market fit after the incentive subsidies end simply fade to zero. The IT services industry is about to experience its own subsidy cliff. The question is not if the cost cuts will come. They will. The question is whether any of the value will stick to the service provider's bottom line.
I would not be long on the mid-tier IT services basket. The risk-reward is skewed to the downside. The market is still pricing these companies on their historical earnings power, but the forward-looking earnings power is being eroded by a structural deflation in the price of code. We do not predict the storm; we short the rain. The storm is here.
