A fresh graduate hired into a structured, usually cohort-based programme — most commonly the Graduate Engineer Trainee format in manufacturing, engineering and IT services — with rotations and a defined absorption point.
Not an individual junior hire. The cohort is a feature, not an administrative convenience: the peer group is a substantial part of what makes these programmes retain people.
Graduate trainee programmes are the oldest structured early-career format in Indian industry, and the well-run ones are extremely effective: a cohort that joins together, rotates together and is assessed against a common standard produces both capable engineers and a peer network that keeps them in the company.
The failure mode is equally well established. A cohort recruited in volume, distributed into whatever seats were open, given no curriculum, and assessed by whichever manager they landed with, produces a two-year attrition wave that arrives exactly when the trainees have become useful.
When this is the right hire
Compared with hiring individual freshers into specific teams:
you hire more than about ten graduates a year into similar roles, the domain takes time to learn, and rotations are genuinely possible. The cohort effect is worth real money in retention.
you hire a handful of graduates into very different roles. A cohort of four people doing unrelated jobs gets the overhead of a programme and none of the peer benefit.
How to do it well
- 1Size the cohort so it functions as one
Below about eight, the peer effect that makes these programmes work does not appear. Above about forty, one owner cannot see individuals.
- 2Design rotations around what the business actually needs
Rotations that exist because the format requires them produce tourists. Three to four months each, with a deliverable per rotation.
- 3Assess against one common standard
The whole point of a cohort is comparability. Letting each rotation manager use their own criteria destroys it.
- 4Give the cohort a named programme owner
Someone who sees all of them, holds the calendar, and notices the person who has quietly been parked in a corner.
- 5Plan the placement decision before month one
Where they land after rotations, decided on criteria, communicated on a date. Trainees who do not know where they are going start looking.
- 6Plan for year two, not just year one
The attrition wave is at eighteen to twenty-four months, when they are useful and marketable. Everything you do about it must be in place before then.
What to pay
GET pay in India is usually a published, standardised figure by discipline and location — one of the few early-career categories where the market is genuinely transparent. Compete on the progression rather than the starting number: a programme where pay steps at defined milestones and reaches market rate on absorption retains far better than one that pays slightly more at joining and then stays flat for two years.
How to assess
- Aptitude testing as a defensible first cut at volume, followed by structured interviews — not the reverse.
- Willingness to rotate and to relocate, asked directly and specifically, since both are where GET offers are declined late.
- Evidence of learning under instruction, which is what the first year actually consists of.
- Domain interest over employer brand — trainees who wanted the brand leave once they have it on the CV.
- For manufacturing and plant roles, genuine comfort with the shop floor. This is best surfaced with a site visit, not a question.
What goes wrong
The trainees who land in unplanned seats are the ones who leave, and they are not randomly distributed — they tend to be the ones with options.
A rotation with nothing to finish is a visit. The trainee learns the org chart and nothing else.
Destroys the comparability that is the whole reason to run a cohort.
The attrition wave arrives when trainees are useful and marketable. Retention has to be designed in before it, not responded to during it.
Bonds retain the people who could not leave and antagonise the people who can. Pay steps and a clear absorption path do the actual work.
Questions employers ask
What is a graduate engineer trainee programme?+
A cohort-based programme for fresh engineering graduates, usually twelve to twenty-four months, with rotations across functions or plants, a common assessment standard, and absorption into a permanent role at the end. It is the dominant structured early-career format in Indian manufacturing and engineering.
How large should a GET cohort be?+
Large enough for the peer group to function — roughly eight at the low end — and small enough for one programme owner to see individuals, which in practice means under about forty. Below eight you get the overhead without the cohort effect.
How long should rotations be?+
Three to four months, each with something the trainee has to finish. Shorter rotations produce familiarity with the org chart rather than capability, and a rotation with no deliverable teaches almost nothing.
Why do graduate trainees leave in year two?+
Because that is when they become useful and marketable, and because many programmes go quiet after the rotations end — no clear placement, no pay step, no next thing. The retention design has to be in place before month eighteen, not devised in response to the first resignations.
Are service bonds effective for graduate trainees?+
They retain the people who could not leave anyway and antagonise the people who can, which selects against you. Milestone pay progression and an early, clear absorption decision retain trainees far more reliably, and they cost less than the reputational damage bonds do on campus.
Related
Free tools for this
No signup. Nothing here asks for your email first.