The Quiet Price War Behind the AI Boom in IT Services

Note: This story was written with AI assistance and edited by me.

Infosys reported its Q2 FY27 results on October 23. AI revenue reached 10.5% of total revenue, up from 8.2% the quarter before, and grew 29% sequentially. Total revenue grew 1.5% year over year in constant currency. The company lowered its full-year growth guidance to 1% to 2%, from 1.5% to 3% earlier. CFO Jayesh Sanghrajka said Infosys has seen "pricing pressure on the back of competitive intensity and higher ask on productivity," and expects that pressure to continue into the second half of the year.

Those are the facts as reported. Everything below is my opinion.

My view: clients have figured out that AI makes delivery cheaper, and they want their share of the savings. Selling hours is getting harder every quarter. The firms that win will price the outcome and staff it with senior people who know how to use the tools.

The productivity dividend belongs to someone

When a team ships the same work in fewer hours, somebody keeps the difference. For years, services firms kept it by billing the same rate card and quietly improving margins. Buyers have caught up. They read the same headlines about AI coding assistants and test automation, and they now walk into renewals expecting a lower number.

That is what "higher ask on productivity" means in plain words. The client is saying: you got faster, so I should pay less.

Why hourly billing is the weak spot

If you sell time, every efficiency gain is a revenue cut. You can fight that by hiding the gains, but buyers are not naive, and the trust cost is high. The alternative is to stop selling time.

Fixed-scope projects, outcome-based milestones and managed capacity with clear deliverables all let a vendor keep part of the productivity gain while the client still pays less than before. Both sides win, and nobody has to pretend AI did not happen.

What smaller firms can take from this

Large IT services companies are built on headcount, so this shift hurts them most. Smaller firms have an advantage if they move first.

Price the result. Agree on what done looks like and what it is worth to the client, then price against that.

Staff senior. One experienced engineer with good AI tooling often replaces a pyramid of junior billable hours. Clients notice the difference in quality and speed.

Show the math. Tell the client where AI saved time and what you did with it. Transparency turns a pricing fight into a partnership conversation.

A fair counterpoint

To be fair, Infosys also signed $5.2 billion in large deals this quarter, including a net-new mega deal of over $2 billion, so demand for big transformation work is real. Pricing pressure is not the same as shrinking demand. My view is only that the old way of charging for that demand is under strain.

What I learned selling services through a price squeeze

I have sat across the table from buyers who wanted the same work for less, and the instinct is always to defend the rate. It almost never works. What worked was changing the conversation from hours to results.

The first time we offered a fixed price for a clearly defined outcome, the client stopped asking how many people were on the project and started asking when it would be done. That one shift made room for better tools, smaller senior teams and healthier margins, because we were no longer punished for getting faster.

It also changed hiring. When you are paid for outcomes, one strong engineer is worth more than three average ones, so you get pickier, and the work gets better.

Takeaway

AI is pushing the price of effort down. That is not bad news for anyone who can define and deliver a result. I took a firm from $6M to $60M in under three years, and the turning point was learning to sell outcomes instead of time.

If you run a services team, pick one client this quarter and propose a fixed-scope milestone instead of a time-and-materials renewal. See how the conversation changes.