DEV Community

Cover image for Outcome-Based Pricing and AI: Why India's IT Giants Just Changed How They Charge (2026)
Shaam
Shaam

Posted on Originally published at aitecharchive.com

Outcome-Based Pricing and AI: Why India's IT Giants Just Changed How They Charge (2026)

If you pay a consultancy, agency, or development partner by the hour, the ground just moved under your contract. In August 2026, the CEOs of Persistent Systems and Tata Consultancy Services (TCS) told Reuters, on the record, that clients are now demanding 25% to 30% lower prices for the same work, and that roughly 80% of TCS's finance, HR and business-services contracts are priced on outcomes rather than hours. AI productivity is the stated reason on both sides of the table. The billable-hour model that built the $300-billion Indian IT industry is being rewritten in real time, and the same logic is coming for any business that sells knowledge work.

TL;DR · Last verified: 2026-08-22

  • Persistent Systems CEO Sandeep Kalra: clients demand the same work for 25-30% less, delivered faster (Reuters, Aug 2026).
  • TCS CEO K Krithivasan: about 80% of contracts in the finance, HR and business-services segment are now outcome-based (Reuters, Aug 2026).
  • TCS says AI generates 10-15% productivity savings, which it passes to clients in phases to protect margins (Mint, July 2026).
  • Cognizant signed a multi-year AI workplace deal with Daimler Truck in February 2026 built on its WorkNEXT platform (PR Newswire, Feb 2026).
  • For buyers: you can now ask for outcome pricing. For sellers: hours are becoming a liability; outcomes are the moat.

What exactly changed in IT contracts this month?

The pricing model itself changed, not just the price level. For three decades, Indian IT services sold capacity: engineers, hours, headcount. The more people a vendor could throw at a problem, the more it billed.

On August 20, 2026, Reuters reported two on-the-record confirmations that this model is breaking:

  1. Persistent Systems CEO Sandeep Kalra said clients are demanding the same work for 25% to 30% less money, while simultaneously expecting faster delivery and higher productivity (Reuters, Aug 20, 2026).
  2. TCS CEO K Krithivasan said about 80% of the company's contracts within its finance, HR and other business-services segment are now priced on outcomes, not hours worked (Reuters, Aug 20, 2026). That share has roughly doubled since generative AI went mainstream in late 2023.

Cognizant confirmed the same direction from a third angle. In its own statement on the shift, the company said: "With AI, the fundamentals are shifting. Clients now expect more value and measurable outcomes" (Business Standard, Aug 2026).

These are not analysts speculating. They are the CEOs of three of the largest IT services firms in the world confirming the same structural move within days of each other.

What is outcome-based pricing, in plain terms?

Outcome-based pricing means the client pays for a delivered result, not for time spent. Instead of "1,000 engineering hours at $X per hour," the contract says "reduce claims processing errors by 40% within 12 months" or "keep this platform at 99.95% uptime," and the fee is tied to that outcome.

The three models side by side:

Model Client pays for Who keeps AI productivity gains Risk sits with
Time and materials Hours worked Client (indirectly) Client
Fixed price A defined scope Vendor, if delivered efficiently Vendor on scope, client on change
Outcome-based A measurable business result Split per contract, often shared Mostly vendor

The third row is why this matters now. When AI makes a vendor 10-15% more productive, an hourly contract transparently passes every saved hour back to the client as lost revenue for the vendor. An outcome contract lets the vendor keep part of the efficiency gain, because the price is anchored to the result, not the effort. Both sides can win, but only if the measurement is honest.

What are the verified numbers behind the shift?

Here is what is actually on record, with dates:

  • 25-30% lower pricing demand. Persistent Systems CEO Sandeep Kalra, speaking to Reuters (Aug 20, 2026). Clients want the same scope cheaper and faster.
  • ~80% outcome-based contracts at TCS. CEO K Krithivasan on the finance, HR and business-services segment, roughly double the share from late 2023 (Reuters, Aug 20, 2026).
  • 10-15% AI productivity savings at TCS. Krithivasan disclosed this figure on the July 2026 post-earnings call, adding that TCS front-loads part of the savings to clients and smooths the rest over the project term ("rule of thumb ... 10% to 15% productivity savings because of AI," Mint, July 10, 2026).
  • $2.6 billion annualized AI revenue at TCS, up from $2.3 billion the prior quarter (Mint, July 2026). AI work is already a serious revenue line, not a pilot program.
  • Cognizant x Daimler Truck, February 24, 2026. A multi-year deal to modernize Daimler Truck's global workplace services using Cognizant WorkNEXT, an AI- and automation-enabled platform, across factories and offices worldwide (PR Newswire / Cognizant, Feb 24, 2026). Reuters reported the deal structure includes sharing AI-driven cost savings between vendor and client.

One honest caveat: the Reuters reporting covers newer and renegotiated deals first. Multi-year legacy contracts signed on hourly pricing do not renegotiate themselves; they shift as they come up for renewal. The transition is real but measured in contract cycles, not weeks.

Why does AI break hourly billing specifically?

Because hourly billing prices effort, and AI attacks effort. If an AI-assisted engineer ships in 4 hours what used to take 10, an honest hourly vendor loses 60% of the revenue on that task for doing better work. That is an unsustainable incentive, and buyers know it, which is why they are demanding the discount.

The shift we documented in why the companies spending the most on AI are also hiring the most people runs on the same logic: AI changes what a person-hour is worth, so every business model built on person-hours has to be rebuilt. For TCS-scale vendors that means outcome contracts. For a solo consultant or a small dev shop, it means productized services, subscriptions, and results-based fees. The mechanics differ; the direction is identical.

It also lowers the moat of scale. For 30 years, a giant workforce was the winning argument in every pitch meeting: more engineers, more hours, more capacity than anyone else. If AI does a growing share of the work, a smaller rival with better tooling can underbid a giant on price and beat it on speed at the same time. Vendor size stops being a proxy for vendor capability.

What does this mean for jobs and hiring in India's IT sector?

Fewer billable hours per outcome means fewer entry-level hires per contract, even when revenue keeps growing. The volume-hiring engine that built much of the modern urban middle class in Bengaluru, Pune, Hyderabad and Chennai was tied to the hours model, and the hours model is the part being renegotiated.

The nuance, worth stating plainly: the sector is not shrinking. TCS reported $7.62 billion in June-quarter revenue and $2.6 billion in annualized AI revenue (Mint, July 2026). AI work is creating demand for different skills: agent supervision, data engineering, AI platform administration, vendor governance. But a graduate entering the industry in 2026 is not walking into the same ramp their older sibling did five years ago. The career advice that follows is the same advice we have given for automating business work with the best coding LLMs: learn to operate AI, not compete with it on raw output volume.

What this means for you

If you buy knowledge work (agency, consultancy, outsourced development):

  1. Ask your current vendors how much of their workflow is now AI-assisted, and when that productivity shows up in your bill. The big-enterprise answer is 10-15%; there is no reason your vendor's answer should be zero.
  2. On your next renewal, ask for an outcome- or milestone-priced option alongside the hourly quote. The fact that a vendor can price an outcome tells you they understand the work well enough to stand behind it.
  3. Put measurement in the contract. Outcome pricing without agreed metrics is just a fixed price with better marketing.

If you sell knowledge work (freelancer, agency, small consultancy):

  1. Start anchoring proposals to the result the client actually wants (leads, uptime, throughput, hours saved), not the hours you will log. This is how you keep the AI productivity dividend instead of handing it over.
  2. Build repeatable, AI-assisted delivery so your cost per outcome keeps dropping while your price per outcome stays flat. That spread is the new margin, and small teams can now compete for contracts that used to require thousands of seats, as we covered in orchestrating a team of AI agents for your business.
  3. If you hire, hire for AI-orchestration skills, not raw task capacity. The whole industry's contract structure just told you which one is appreciating.

FAQ

Q: What is outcome-based pricing in IT services?
A: A contract model where the client pays for a measurable business result (error-rate reduction, uptime, throughput), not for hours worked. TCS CEO K Krithivasan told Reuters in August 2026 that about 80% of TCS's finance, HR and business-services contracts now use it.

Q: How much cheaper are clients demanding IT work be?
A: Persistent Systems CEO Sandeep Kalra told Reuters (Aug 20, 2026) that clients want the same work for 25% to 30% less, delivered faster, with higher productivity.

Q: How much productivity is AI actually saving these firms?
A: TCS disclosed 10-15% cost savings from AI on its July 2026 earnings call, passed to clients in phases so margins are preserved over the contract term.

Q: Is the Indian IT industry shrinking because of AI?
A: By revenue, no: TCS reported flat-sequential revenue of $7.62B for the June 2026 quarter and $2.6B in annualized AI revenue. The pressure is on the pricing model and entry-level hiring volume, not on top-line demand.

Q: Should small businesses ask their own vendors for outcome pricing?
A: Yes. On the next renewal, request an outcome- or milestone-priced option with agreed metrics. Vendors that can price outcomes prove they understand the work; vendors that refuse tell you something too.

Q: Does outcome-based pricing transfer risk to the vendor?
A: Mostly. The vendor commits to a result, so it carries delivery risk, but it also keeps more of any AI-driven efficiency gain, since the price is tied to the outcome rather than the effort.

Sources

Updates & Corrections

  • 2026-08-22 - Initial publication. All quantitative claims verified against the Reuters, Mint, PR Newswire and Business Standard items listed above.

Researched and drafted with AI agents; reviewed and fact-checked under human editorial oversight. How we work.

Top comments (0)