Wanted: New value metrics for IT giants in AI age

Wipro

14/08/2026 13:00

The author is a Reuters Breakingviews columnist. The opinions expressed are her own. Updates to add graphic.
By Ujjaini  Dutta
BENGALURU, Aug 14 (Reuters Breakingviews) - India's $315 billion IT industry was once easy for investors to read: work was largely billed on human effort, so headcount was a straightforward indicator of revenue growth and earnings potential. That correlation is fraying as AI tools made by Anthropic, OpenAI and others mean IT companies can serve clients with fewer people and increase revenue without hiring. New measures are needed to differentiate between firms adapting fast enough and those who can't keep up.
Traditional contracts paid IT companies for hours worked and engineers deployed. That model is fading fast, especially at smaller and more agile firms. Yet the distinction is not reflected in share prices. Investors have indiscriminately dumped shares in giants led by $90 billion Tata Consultancy Services TCS.NS , Infosys INFY.NS and Wipro WIPR.NS, alongside stakes in smaller providers like Blackstone-backed BX.N Mphasis MBFL.NS. The benchmark Nifty IT index .NIFTYIT is down 17% this year.
Some companies are starting to report "AI revenue". Venu Lambu, the boss of $15 billion LTM LTIM.NS, reckons it will become the industry's defining metric but that seems unlikely. At Tata Consultancy, this newly adopted measure is poorly explained and, worse, lumpy. A more nuanced approach is probably required.
From March next year, $5 billion Mphasis,will report “key operating metrics” such as annual recurring revenue and platform attach rate. This suggests it sees a growing share of future revenue from selling software-like services. The firm already generates less than half of its topline from the traditional "Time and Materials" billable hours model. Similarly, $8.5 billion Coforge is building up a portfolio of AI-related software products and tools including 8 AI platforms, 22 AI assets and over 100 reusable AI agents, per its latest post-earnings call.
AI is not only making engineers more productive; it is also encouraging IT firms to package their expertise into reusable software platforms as clients increasing value speed of deployment. So instead of building a custom customer-service application from scratch, an outsourcer could deploy a pre-built AI platform across dozens of clients and charge subscription fees.. Attach rate would show how often clients purchase additional products or services after adopting a platform, offering a clue about future revenue growth.
This shift to a "services-as-software" model is imminent, says Sanjeev Narsipur, Managing Director and Lead of Digital, AI and Technology Services at Alvarez & Marsal. And as this line blurs globally, AI developers OpenAI and Anthropic are also moving into enterprise services. Whatever the way forward, after decades of falling back on old metrics, IT firms need to find new ways to demonstrate their worth. 
Follow Ujjaini Dutta on LinkedIn and X.
     
CONTEXT NEWS
Blackstone-backed Mphasis sees a shift toward "recurring, platform-attached revenue” the company said on July 24 in a post-earnings call. In May, the company said it would start reporting “key operating metrics” such as annual revenue runrate, platform attach rate and net retention rate for the year ending March 2028.

(Editing by Una Galani; Production by Aditya Srivastav)
((For previous columns by the author, Reuters customers can click on DUTTA/ujjaini.dutta@thomsonreuters.com))