Narayan AmmachchiIndian IT vendors are reportedly struggling to price services contracts, with traditional billing models rapidly losing relevance in the AI era.
Under the conventional model, service providers billed clients based on the number of engineers deployed and the hours worked. This approach offered predictability for both sides.
That framework is now under strain. With IT companies deploying AI assistants such as Copilot in software development, the need for large teams and extended development timelines has reduced sharply.
AI agents can now write code that earlier required five engineers, while several stages of software testing are increasingly automated. As a result, clients are questioning the rationale for paying for large engineering teams when AI can perform a significant share of the work.
Despite recognising the growing role of AI in software development, IT vendors are finding it difficult to price contracts accurately before execution. Analysts say this is because service providers still lack clarity on which functions can be fully automated and where human intervention will continue to be essential.
While some processes lend themselves smoothly to automation, others resist it due to poor data quality, legacy systems, or regulatory constraints. This uncertainty makes it difficult to lock in prices upfront.
“Pricing has become a guessing game,” said a senior executive at a large IT services firm, who declined to be named.
The executive added that AI adoption does not automatically translate into lower costs. Many AI tools used in software development come with high licensing fees, while cloud infrastructure and data management add to expenses.
Meanwhile, some clients are reportedly pushing for outcome-based pricing models rather than traditional effort-based billing.
Add comment