What GCC and Product Drives Screen For That IT-Services Never Did

In a debrief after the 2025-26 campus drive season, a recruiter from a Global Capability Centre told me something I have since repeated to almost every placement head I speak with. Their team had run drives across fourteen campuses. On every campus, the students cleared the online aptitude round. On most campuses, the technical round produced two or three offers, sometimes none at all. The bottleneck was not reasoning speed or numerical ability. Students were clearing those sections. The bottleneck was a sixty-minute machine-coding task: given a problem statement, build a working system from scratch, write the class design, and explain the thinking as the code takes shape. Most students had never done that in a test room. Several had never done it anywhere.

That recruiter’s debrief describes a structural mismatch that most college placement programs have not yet addressed. The preparation that gets students through IT-services drives does not reliably get them through GCC and product company drives. The gap lives in a category of technical assessment that most placement training was never designed to reach.

What the IT-services trainability screen was built to do

The campus drive format that shaped Indian placement training for two decades was calibrated to IT-services hiring logic. A large services company needed to hire several thousand freshers every year, route them into a shared onboarding and induction track, then deploy them into client projects as work arrived. The campus screen was not designed to find engineers who could already do a specific job. It was designed to find people who could learn one quickly.

The TCS NQT reflects this logic clearly. The Foundation section runs seventy-five minutes across Numerical Ability, Verbal Ability, and Reasoning Ability. The Advanced section adds a ninety-minute coding component. Together they answer one question: is this person sharp enough and disciplined enough to be shaped into a productive delivery team member after six to twelve months of training? The company does not need to know whether the student can already build a working software system. That is the training program’s job.

This is a reasonable screen for that model of hiring. The preparation it calls for is also reasonable: timed aptitude drills, a grounding in data structures and sorting algorithms, enough coding confidence to solve structured problems in a constrained window. Colleges across India have spent two decades building this preparation, and most have gotten good at it.

The challenge arrives when the same preparation is measured against a drive type that does not run on trainability logic at all.

The screen that GCC and product drives run on

A Global Capability Centre is the parent company’s own engineering team, building the company’s own systems in India. A fresher who joins a GCC data team or a GCC platform team is expected to contribute to live work within weeks of joining. There is no common training pool that absorbs a large batch and sorts them into projects later. Before an offer is made, the company must have evidence that this specific student can deliver this specific role.

Times of India reporting from the 2026 drive season named the criteria GCC interviewers evaluate: depth in the relevant technology domain, problem-framing ability, ownership after delivery, and practical use of AI tools in real work. One candidate who joined a Bank of America technology GCC spoke about how his entire technical interview turned on an AI-powered exam preparation tool he had built for his college. The interviewer probed the architecture choices, the edge cases he had discovered and handled, and how he had measured whether the tool actually worked. The degree, the certifications, and the course completions were not what drove that conversation. The project was.

This is the screen that GCC and product company drives are built to run: can this student show us something they built, explain every meaningful decision in it, and discuss what they learned from building it? Infosys, which hired over 20,000 college graduates in FY26 and has framed its fresher hiring explicitly as a strategic bet on GenAI-native talent, has moved its assessment toward applied AI problem-solving alongside code quality. The direction from the largest IT-services recruiters matches the direction from GCCs: proof of what you can build is becoming the primary filter in technical rounds across company types.

NASSCOM and Zinnov’s 2026 GCC ecosystem research documents that India’s GCC sector has grown 32 percent since FY21, with the workforce profile oriented toward building and operating complex engineering systems rather than delivering against a specification handed down by a client project manager. That shift in the nature of GCC work is why the screen has moved in the direction it has.

What the machine-coding round tests

The machine-coding round is the single most consistent point of failure for students who have prepared only for IT-services drives, and it is worth understanding in concrete terms.

A machine-coding task gives the candidate a problem statement and sixty to ninety minutes to build a working small system from scratch. A common version: implement a library management system with defined operations, handle the edge cases, and write the class structure cleanly. Another version: build a rate limiter with a specified API, explain the data structure choice, and describe how your design handles concurrent requests. Some GCC companies now include a task where the candidate is given an AI-generated code block with a bug, and must read it, identify the failure, fix it, and explain the change.

The assessor watches how the candidate interprets a requirement that is not fully specified, how they break the problem into logical units, how they name variables and methods, and how they handle the cases the problem statement does not mention. They also ask questions while the candidate is working, because explaining a design choice under time pressure is part of the evaluation.

The machine-coding round asks one question: can you build a coherent small system from a problem statement and explain every meaningful decision you made while building it? A student who has solved five hundred practice problems on a coding platform and drilled aptitude for months is genuinely prepared for the IT-services screen. That same preparation does not build the skill this round measures.

The problem-framing round

Beyond machine-coding, GCC and product company drives include a problem-framing round that has no direct equivalent in the IT-services format. This round asks a candidate to explain, clearly and logically, how they would approach a business or engineering problem they have not seen before, without necessarily producing a working solution.

A BFSI GCC might present a scenario: customer complaint volumes at a bank’s retail network have risen 40 percent in one quarter, concentrated in three cities. What would you want to know, what data would you look for first, and how would you structure the investigation? A technology GCC might present a production system with unexplained latency spikes and ask the candidate to walk through their diagnostic reasoning.

The interviewer is not looking for the correct answer to the business problem. They are evaluating whether the candidate can handle an unclear problem without freezing, ask useful questions, identify what information would change the approach, and communicate the reasoning structure clearly in English. This lives closer to analytical communication than to technical knowledge, and most campus preparation programs have no structured track for it. Interview preparation that focuses on technical correctness builds something useful and necessary, but it does not build this skill on its own.

Where service-track preparation carries over and where it stops

The aptitude foundation that every IT-services drive tests is the same foundation that GCC and product drives assume is in place before the technical rounds begin. A student who cannot clear a quantitative reasoning filter will not reach the machine-coding stage at any company type. The communications baseline that most college programs now include as part of IT-services preparation carries directly into GCC behavioural and communication rounds. None of that work is wasted.

What requires a separate track is the ceiling. A student who has cleared TCS NQT simulations repeatedly is ready for the IT-services screen. The machine-coding round and the problem-framing conversation sit above that preparation, and they do not improve from aptitude drilling or additional DSA problem sets alone. Building them requires building things.

An engineering college near Pune that closed the machine-coding gap

An engineering college in the Pune region with approximately 2,500 students across CSE, ECE, and IT branches had a well-structured placement program going into FY25. Their aptitude preparation was solid. Their coding fundamentals training consistently moved the cohort above average by the pre-drive baseline they ran every year. They were placing students steadily in IT-services roles.

Their GCC conversion rate was near zero.

When we worked through the data together, the picture was clear. In the two prior years, every student who had attempted a GCC technical round had cleared the aptitude stage and stalled in the technical assessment. The preparation had built the right floor. It had not built the ceiling the GCC screen was looking for.

We designed a two-month addition to their existing CSE preparation, running alongside the aptitude calendar. Weekly machine-coding sessions of ninety minutes each: a problem statement, a blank file, and a time limit. No scaffolding. The session output was assessed on three things: does the solution run, is the class design coherent, and can the student explain one design choice when asked by the observer. The first session was difficult across the board. By the fourth session, a segment of the cohort had stabilised into a consistent working pattern.

Of the approximately 240 students in the CSE batch, around 65 could produce a defensible small system under test conditions by the end of the two months. The placement cell built its GCC shortlist from that group, matched by sector against the companies they had mapped.

Three GCCs ran drives from the shortlist. Twelve offers came through at packages significantly above what the same batch had received from IT-services drives. The students who converted were not the highest scorers on aptitude or on the DSA problem-set tests. They were the ones who had spent eight weeks building things under time pressure and could talk about what they had built.

What a placement cell needs to add to its existing program

Every placement cell that already runs a solid IT-services preparation program has the floor. What it needs to add is a project-and-build track that sits above the existing technical calendar, without replacing any of it.

In practice, this means three additions.

Weekly or fortnightly machine-coding sessions from the third semester onward, where students are given a problem statement and a time limit and asked to produce a working, explainable small system. The output is scored on design quality and the student’s ability to explain one or two choices, not on algorithmic speed. The sessions run alongside the existing coding and aptitude schedule.

A problem-framing practice block, separate from mock-interview preparation, where students handle open-ended business or engineering scenarios and practise communicating their reasoning structure. Three or four such sessions per semester is enough to move a segment of students from freezing on an open question to handling it fluently.

A portfolio checkpoint at the end of each academic year, so that every student in the selective-drive track has at least one documented project they built themselves and can discuss in detail. The checkpoint is a conversation, not a form, and the standard is whether a student can sustain forty minutes of technical questions about the project they claim to have built.

None of this dismantles the existing preparation. The aptitude calendar stays. The company-specific mock drives stay. The coding fundamentals track stays. What changes is the ceiling: students who go through the services preparation plus the build track come to the drive season ready for more doors, not just the ones the IT-services format opens.

Where the preparation gap matters most

For colleges outside the most selective tier, this matters for a concrete reason. GCC campus participation has grown substantially in the 2025-26 drive season, and the companies running those drives are not choosing between famous and less-famous campuses. They are choosing between prepared and unprepared cohorts. A shortlist from a lesser-known college that shows machine-coding-ready students assessed against the company’s specific criteria is often more actionable for a GCC recruiter than a general invitation from a college with a stronger name.

The preparation investment that lifts a cohort from IT-services-ready to GCC-ready is a defined, bounded addition to an existing program. Across the more than 2,000 partnerships we have built over 18 years, the lever that opens the GCC door is almost always the same: a build track above the aptitude floor, a project that a student owns completely, and a way to assess and report readiness before the outreach begins.

The students who cleared GCC rounds from that Pune-area cohort were not the ones with the highest aptitude scores from the prior year’s baseline. They were the ones who had spent eight consecutive weeks building things under test conditions and had something to show for it. For a placement head reading the same pattern in their own debrief data, the question becomes concrete: which students in your batch have something to show? That question has a preparation answer, and preparation is within the placement cell’s control.