She read the headline — *"Indian employers drop degree requirements"* — the same week her application was filtered out of a campus drive for having 6.7 CGPA against a 7.0 cutoff. Both things were true at the same time, in the same market, and possibly at companies two floors apart in the same building.
There is a genuine shift happening in Indian hiring, and it is being oversold badly. Reporting in 2026 describes roughly 30% of Indian employers having dropped degree requirements for tech roles, notably higher than the global figure, and a 2026 India Employability Report cited widely puts around 68% of Indian startups and GCCs as having removed minimum CGPA cutoffs, replacing them with skills assessments. Those numbers come from industry reports and trade coverage rather than official statistics, so treat them as directional.
But here is the number that gets left out of the headlines, and it is the one that should shape what you do: the same reporting says only about 21% of corporates have fully adopted skills-first hiring, and roughly half still allow CGPA to influence decisions.
So the honest version is not "credentials stopped mattering". It is: the filter did not disappear, it moved — from something you either have or do not, to something you have to build. That is better news than it sounds, and it requires more work than the headlines suggest.
What actually changed
- A second door opened. Startups, GCCs and product companies increasingly run assessments, work samples or "assess-thons" as the first filter instead of a CGPA cutoff. That is a genuinely new route for people who were previously screened out before a human saw them.
- Proof became portable. A deployed project or a merged contribution travels across companies in a way a marksheet never did.
- The interview got heavier. When the credential stops doing the filtering, the assessment moves later — into work samples and into rounds where you have to defend what you built. Our work trials and job simulations guide covers that shift.
What did not change
- Large service companies still filter. CGPA and percentage cutoffs remain standard in mass campus recruitment, because when you are screening a hundred thousand applications, an automatable filter is operationally necessary regardless of what anyone believes about its predictive value.
- Roughly half of employers still weigh it. Not as an absolute cutoff, but as an input.
- Tier-1 college branding still helps, especially for referrals and for first-round attention.
- The bar did not drop. Skills-first hiring is not easier hiring. It is differently-filtered hiring, and the assessment you now face is often harder than the cutoff you avoided.
Which means the correct strategy is not to pick a side. It is to plan for both worlds simultaneously, because you will apply into both in the same month.
What replaces the credential
In rough order of how hard each is to fake, which is the same as how much it is worth:
One deployed thing. Live, with a link, solving something real — even something small and unglamorous. One finished, deployed project beats four half-abandoned repositories by a wide margin, because finishing is the rare part. This is the single highest-return item on the list.
A reviewable work sample. Code a stranger can read, a dashboard, a written analysis, a design document. The point is that someone can evaluate it without taking your word for anything.
An assessment score. The route many startups and GCCs now use in place of a cutoff. Unglamorous, and it genuinely opens doors that were previously closed. Practise the format, because assessment performance is partly a skill in itself.
A verifiable contribution. Two merged pull requests to a real project, a talk, a written post that someone else linked to. Disproportionately persuasive precisely because it is checkable.
The ability to defend all of it, out loud. This is not a fifth item so much as the thing that validates the other four. In a credential-based system nobody asked you to justify your degree. In a skills-based one, every claim is a question you have agreed to answer — which is why our will recruiters know you used AI on your resume guide matters more now than it would have three years ago.
What to do about it, practically
- Build one thing and finish it. Deployed, linked, with a README explaining the decisions. Two months of evenings. Our how to talk about GitHub projects guide covers presenting it.
- Get the proof onto the surfaces that are searched. A project nobody can find does nothing. Our LinkedIn profile tips for software engineers guide covers being retrieved at all.
- Target the companies that have actually moved. Startups, GCCs and product companies are further along than large service firms. Our GCC jobs in India guide covers that channel, which is currently hiring at record volume.
- Do not abandon the credential game entirely. Apply to the CGPA-filtered drives too if you clear them. Two doors beat one.
- Prepare for assessments specifically. If the filter is now an assessment, the assessment is now the thing to practise.
- Rehearse defending your work. The proof only converts if it survives three follow-ups, which is a spoken skill and the one most people skip entirely.
If your CGPA is below the cutoffs
The shift genuinely helps you, and it does not help you everywhere. Our how to explain a low CGPA in an interview guide covers the answer when it comes up in the room. Beyond that: prioritise off-campus and referral routes where the filter is applied by a human rather than a form, target the company types listed above, and put the proof where it is seen before the marksheet is. Our off-campus placement strategy for India guide covers the channels.
A caution about the narrative
Two honest warnings about this trend, because it is being marketed at you hard.
The people telling you degrees no longer matter are frequently selling something — a bootcamp, a certification, a course bundle. The underlying shift is real; the urgency attached to it usually belongs to someone's marketing calendar. A certificate is not proof of skill. A deployed project is.
"Skills-first" also means "prove it, repeatedly." In a credential system you cleared a bar once, in your first year of college. In a skills system you re-prove capability at every application, in an assessment, a work sample, and an interview. For most people that is a better and fairer deal. It is also more work, done more often, and nobody putting the cheerful version of this trend on your feed mentions that part.
Where each option actually helps
- Your own project — the highest-return item available, and the one nobody can sell you.
- Assessment platforms — practise the format if the companies you want use one.
- Free roadmaps and official documentation — the information layer is not the constraint. Our is interview prep worth paying for guide covers what is genuinely worth money.
- Certifications — help at the resume-screen stage, weak on their own, and much stronger paired with something you built.
- ChatGPT — genuinely useful for learning fast. Its trap in a skills-first market is that it makes producing artefacts easy enough that you can end up with proof you cannot defend.
- Greenroom — the spoken layer, which is the step that validates everything else. Ari, the AI interviewer asks about what you built and then asks again. Honest tradeoff: it cannot build the project for you, and the project is the part that matters most.
Frequently asked questions
Are degrees and CGPA no longer required for tech jobs in India?
Partly, and the shift is being oversold. Reporting in 2026 describes around 30% of Indian employers dropping degree requirements for tech roles and an India Employability Report figure of roughly 68% of startups and GCCs removing minimum CGPA cutoffs. The counter-figure that headlines omit is that only about 21% of corporates have fully adopted skills-first hiring and roughly half still let CGPA influence decisions. These are industry reports rather than official statistics, so treat all of them as directional.
What is skills-based hiring?
Hiring that filters on demonstrated capability — an assessment, a work sample, a deployed project or a verifiable contribution — rather than on a credential such as a degree, an institution or a CGPA cutoff. In practice it does not remove the filter, it moves it: instead of clearing a bar once in your first year of college, you re-prove capability at every application through an assessment and an interview where you defend what you built.
What counts as proof of skill without a degree?
In rough order of how hard each is to fake: one deployed project that is live, linked and solves something real, since finishing is the rare part; a reviewable work sample such as readable code, a dashboard or a written analysis; an assessment score, which is the route many startups and GCCs now use in place of a cutoff; and a verifiable contribution such as merged pull requests, a talk or a published post. All of it depends on being able to defend it out loud, which is what validates the rest.
Which companies in India have actually moved to skills-based hiring?
Startups, global capability centres and product companies are furthest along, frequently running assessments or work-sample rounds as the first filter. Large service companies conducting mass campus recruitment still commonly apply CGPA and percentage cutoffs, because an automatable filter is operationally necessary when screening very large application volumes. The practical implication is to apply into both systems rather than choosing a side.
Does a certification count as proof of skill?
It helps at the resume-screening stage and it establishes vocabulary, but it is weak on its own and much stronger when paired with something you actually built. It is also worth noticing that many of the loudest voices claiming degrees no longer matter are selling a course, a bootcamp or a certification bundle — the underlying shift is real, but the urgency attached to it usually belongs to someone's marketing calendar.
Is skills-based hiring easier for candidates?
It is fairer for many people and it is not easier. The bar did not drop; the assessment simply moved later and became harder to shortcut, since a work sample and a defence-under-follow-up are more demanding than clearing a numeric cutoff. It also means proving capability repeatedly at every application rather than once, which is more work done more often — a genuine trade that the cheerful version of this trend rarely mentions.