Functional comparability governs software and IT-enabled services benchmarking, requiring exclusions, segmental inclusion, and fresh verification of distribution tolerance.
Retrospective amendments to the limitation framework in sections 144C and 153, effective from 1 April 2009, place the final assessment order within time. Transfer-pricing benchmarking of software-development and IT-enabled services must rely on contemporaneous functional evidence and reliable segmental data: entities with distinct functions, diversified unsegregated operations, or unreliable related-party transaction information require exclusion, while functionally comparable segregable segments may be included. The distribution transaction's tolerance-band claim requires factual verification of purchase-cost, inventory-to-sales, and arm's-length-price variation conditions; any eligible transaction within the prescribed band requires deletion of the adjustment. Benchmarking must be recomputed using the resulting comparability changes.
Issues: (i) Whether the final assessment order was barred by limitation after the retrospective amendments to sections 144C and 153; (ii) Whether specified comparables were liable to be excluded or retained for benchmarking software development services; (iii) Whether specified comparables were liable to be excluded or included for benchmarking IT-enabled services; (iv) Whether the assessee was eligible for the 3% tolerance band for the distribution-segment transaction.
Issue (i): Whether the final assessment order was barred by limitation after the retrospective amendments to sections 144C and 153.
Analysis: The Finance Act, 2026 amendments to sections 144C and 153, retrospectively effective from 01.04.2009, governed the limitation computation. Under the amended framework, the final assessment order fell within the prescribed period.
Conclusion: The final assessment order was not barred by limitation. This issue was decided against the assessee.
Issue (ii): Whether specified comparables were liable to be excluded or retained for benchmarking software development services.
Analysis: The assessee was a software-development service provider acting on associated enterprises' instructions and assuming limited service-liability and manpower risks. Tata Elxsi and Cybage Software earned revenue from distinct activities without reliable segmental bifurcation of software-development results. Apttus Software's publicly available related-party transaction information was unreliable. Systango's contemporaneous annual-report data and its response under section 133(6) showed that its revenue was entirely from software-development services; unverified website material and decisions relating to different years did not displace that evidence.
Conclusion: Tata Elxsi, Cybage Software and Apttus Software were directed to be excluded, while Systango Technologies was retained as a comparable. This issue was partly decided in favour of the assessee.
Issue (iii): Whether specified comparables were liable to be excluded or included for benchmarking IT-enabled services.
Analysis: The assessee's IT-enabled services comprised regulatory data-sheet work, product and shipment support, label management, order processing and customer-support functions. Savitriya Technologies was retained on contemporaneous records showing an IT-enabled-services segment, and the earlier merger did not affect the relevant year. Sutherland Global's information-technology services, Sagacious Research's intellectual-property consultancy, and TTEC India's BPO activities were functionally distinct. Microland's separately reported IT-enabled-services segment was segregable from its infrastructure-management activities and was functionally comparable.
Conclusion: Sutherland Global Services, Sagacious Research and TTEC India Customer Solutions were directed to be excluded; Savitriya Technologies was retained; and Microland's IT-enabled-services segment was directed to be included. This issue was partly decided in favour of the assessee.
Issue (iv): Whether the assessee was eligible for the 3% tolerance band for the distribution-segment transaction.
Analysis: The applicable notification prescribed a 3% tolerance band for transactions other than wholesale trading. Eligibility depended on factual verification of the purchase-cost and inventory-to-sales conditions and of whether the variation between the arm's length price and transaction value exceeded 3%.
Conclusion: The distribution-segment issue was restored for verification of the assessee's eligibility for the 3% tolerance band; if eligible and the variation does not exceed 3%, the adjustment is to be deleted. This issue was decided in favour of the assessee to that extent.
Final Conclusion: The assessment survived the limitation challenge, while the software-development and IT-enabled-services benchmarking must be recomputed using the directed comparability changes and the distribution-segment tolerance claim must be verified afresh.
Ratio Decidendi: Transfer-pricing comparables must be selected on contemporaneous functional data and reliable segmental information, and entities with materially different functions or unsegregated diverse operations cannot be retained merely because they broadly provide technology-related services.