Deliveries over deployment: IT firms must reassess AI integration as pricing pressures rise, says BCG

A latest report from the Boston Consulting Group (BCG) titled “Now That (Almost) Everyone Has AI”, argues that as AI becomes more ubiquitous, the question for technology services companies is no longer whether they can deploy AI, but whether they can take responsibility for what the technology delivers.

“Over 90% are crowding around the same handful of themes: agents, orchestration, autonomy, platforms. That is not differentiation, that is noise,” Karan Chaddha, managing director and partner at BCG India, told ET AI.

BCG analysed more than 40 technology services providers and over 100 brand positioning claims across their websites, earnings calls and investor presentations and found more than 90% were talking about the same core themes.


Across its CXO interviews, BCG found six recurring priorities – measurable impact on business P&Ls (profit and loss statements), fit with existing processes, integration with technology stacks, data control and governance, flexibility for companies to retain in-house control, and proof that a provider can stand behind the outcome.

“Clients are asking for outcomes. Providers are still selling capabilities. That gap is the whole problem,” Chaddha added.

Also Read: India’s banks could turn legacy tech into an AI advantage: BCG

Pricing pressures surface

BCG estimates the global technology services market could reach $2.2 trillion by 2030, while the business process outsourcing (BPO) market is expected to be worth $435 billion in 2026. However, the economics of these businesses are changing. The cost of GPT-4-class inference has fallen roughly 60 times in less than two years, according to BCG, as there’s a new frontier model in roughly every six weeks.

“Until recently, volume expansion offset price compression. That buffer is running out,” Chaddha said.

In such a scenario, it becomes difficult to charge simply for access to technology or for the number of people deployed to perform a task.

Operations-heavy businesses are likely to see this shift first, according to BCG. BPO, managed services and customer experience have processes where outcomes are relatively easy to measure and labour costs are high.

Software engineering is also facing significant pricing pressure from AI, Chaddha said, although moving towards outcome-based pricing is harder because software quality is more difficult to measure.

India & its IT advantage?

For India's large IT services companies, that raises a bigger question. A model built around scale, large talent pools and relatively lower-cost execution has been one of the industry's defining advantages. AI puts pressure on that model by reducing the amount of human effort required for some tasks.

The cost of GPT-4-class inference falling is one indication of how quickly the economics are changing. Agents also do not work like employees who need a seat, making some traditional per-seat and headcount-based pricing models less relevant.

But BCG does not see India's advantage disappearing. Instead, Chaddha said the basis of the advantage is changing.

“The real advantage” is whether companies can capture that knowledge in their AI layer, he said, through runbooks, process maps and libraries of exceptions, and then reuse and improve that knowledge over time.

That could turn institutional knowledge into a new source of competitive advantage rather than leaving it embedded in people and individual projects.

Also Read: The new AI builder: Skills for the agentic AI era

Look out for the next contract cycle

BCG believes the timing could be particularly important for technology services companies. A large share of enterprise AI contracts signed during the 2023-24 AI boom are expected to come up for renewal by late 2026.

That could give enterprises an opportunity to reassess what they are paying for and how those contracts are structured.

“Some are actually moving. Most are rebranding. There is a real difference between the two,” Chaddha said when asked whether Indian IT services companies are moving fast enough.

In his view, changing the language in an investor presentation or adding “agentic” to a portfolio does not amount to a fundamental shift. Moving means changing pricing, contracts, delivery structures and how performance is measured.

The providers that are changing fastest, he said, are reorganising teams around owning a process rather than simply staffing it, and tying commercial models to results rather than headcount.

BCG's report points to outcome-based pricing, end-to-end accountability and agentic managed services as relatively less crowded areas where providers could differentiate.

It cites the example of a US healthcare payer that traditionally relied on hundreds of analysts for prior authorisation, a labour-intensive process involving manual reviews, errors and cycle times measured in days. An AI operations company embedded its platform into the operation, redesigned the workflow and, within 12 months, delivered $23 million in annual savings. Its commercial model was tied to the outcome rather than the number of FTEs (FDEs?) deployed.