50,000 Candidates,
Zero Pedigree Filtering
How a large IT services company assessed 50,000+ campus applicants on capability instead of college name — and found stronger, more diverse, lower-attrition talent at 26% lower cost.
Company
IT Services & Consulting
Employees
45,000+
Campuses
150+
Annual Hiring
3,000–4,000
Company Profile
| Type | IT Services & Consulting |
| Size | 45,000+ employees |
| Location | Pan-India (12 offices) |
| Annual Campus Hiring | 3,000–4,000 freshers |
| Colleges Visited | 150+ campuses |
| Applicant Volume | 50,000+ per year |
Company name withheld at client request.
The Challenge
The Pedigree Trap
For years, the company’s campus hiring followed a simple formula: Visit Tier-1 colleges, hire as many as possible, ignore everyone else.
The logic seemed sound. IIT and NIT graduates were “proven.” Regional colleges were unknown quantities. Why take risks?
But the formula was breaking:
Tier-1 supply couldn’t meet demand.
They needed 3,500 freshers. Tier-1 colleges produced ~50,000 engineering graduates total, competed for by every major employer. They were fighting for a shrinking slice of a small pie.
Tier-1 costs were escalating.
Starting salaries for IIT graduates had increased 40% in three years. Tier-1 hires expected faster promotions, premium projects, and accelerated growth paths. The economics were straining.
Tier-1 attrition was highest.
Counterintuitively, their Tier-1 hires had the highest attrition. After 18–24 months of training and project experience, they’d jump to product companies or startups offering 50–70% raises.
They were missing talent.
Anecdotally, some of their best performers had come from lesser-known colleges. But the hiring process systematically excluded these candidates before anyone evaluated them.
The Filtering Reality
- All campus applicants
- 90% rejected on college name alone
- Further narrowing within Tier-1
- First actual evaluation of capability
- Group discussion + Interview
90% of applicants rejected based on college name alone — before any evaluation of actual capabilities.
The Question
“What if we’re wrong about college tier? What if there are strong candidates at Tier-2 and Tier-3 colleges that we’re systematically ignoring?”
— VP of Campus Relations
The problem: They had no way to evaluate 50,000 candidates. The current process worked precisely because it filtered 90% before evaluation. Without that filter, the process would collapse. Unless they could automate first-round evaluation.
The Solution
The Experiment
They designed a controlled experiment:
Group A — Control
Traditional process. Tier-1 filter, then aptitude test, then interviews.
500 hires
Group B — Test
No college filter. All applicants take LayersRank assessment. Top scorers advance regardless of college.
500 hires
Each group would hire 500 candidates. After 12 months, they’d compare performance, retention, and trajectory.
LayersRank Implementation
For Group B, the process became:
- All candidates invited to LayersRank assessment
- AI evaluation + scoring
- Interviews (condensed, informed by reports)
- Hired on demonstrated capability
Key Change
College name was not visible to LayersRank models or to interviewers reviewing reports. Evaluation was purely on demonstrated capability.
Assessment Design
The campus assessment measured:
Technical Fundamentals
40%
- Programming logic & problem-solving
- Data structures & algorithms (conceptual)
- Basic system thinking
Learning Orientation
30%
- Response to novel problems
- Reasoning through unfamiliar scenarios
- Intellectual curiosity signals
Communication
30%
- Clarity of expression
- Structured thinking
- Professional presence
Questions were calibrated for fresh graduates — testing potential and fundamentals rather than experience.
The Results
Tier-2/3 Hires
Before
0%
→
After
66%
12-Month Attrition
Before
18%
→
After
11%
Average Salary Cost
Before
₹8.2L
→
After
₹6.4L
Where Top Candidates Came From
Distribution of candidates scoring in the top 15% (advancement threshold):
| College Tier | % of Applicants | % of Top 15% | Representation |
|---|---|---|---|
| Tier 1 (IIT/NIT/BITS) | 12% | 28% | 2.3x |
| Tier 2 (State/Good Private) | 35% | 38% | 1.1x |
| Tier 3 (Regional) | 53% | 34% | 0.6x |
While Tier-1 candidates were over-represented in top scorers (2.3x), 72% of top candidates came from Tier-2 and Tier-3 colleges. Under the old system, these 72% would have been rejected without evaluation.
Hiring Outcomes
| Metric | Group A (Traditional) | Group B (LayersRank) |
|---|---|---|
| Candidates evaluated | 3,500 | 32,000 |
| Hires | 500 | 500 |
| Tier-1 hires | 100% | 34% |
| Tier-2 hires | 0% | 42% |
| Tier-3 hires | 0% | 24% |
| Average starting salary | ₹8.2 lakh | ₹6.4 lakh |
12-Month Performance Comparison
After one year, they compared the two groups:
| Metric | Group A | Group B | Difference |
|---|---|---|---|
| Training completion rate | 94% | 96% | +2% |
| Training assessment scores | 78/100 | 81/100 | +4% |
| Manager satisfaction (1–5) | 3.8 | 4.0 | +5% |
| Project deployment rate | 89% | 93% | +4% |
| Promotion rate (12 mo) | 12% | 14% | +17% |
| Attrition rate (12 mo) | 18% | 11% | -39% |
Key Finding
Group B (LayersRank, no pedigree filter) performed as well or better than Group A (traditional, Tier-1 only) on every metric.
The Attrition Surprise
The most striking result was attrition. Group B’s 11% attrition was dramatically lower than Group A’s 18%.
Hypothesis: Tier-2 and Tier-3 candidates felt they had more to prove. They valued the opportunity more highly. They were less likely to jump ship for incremental salary gains.
Group A Attrition Cost
90 departures × ₹4L
= ₹3.6 crore loss
Group B Attrition Cost
55 departures × ₹4L
= ₹2.2 crore loss
Savings from attrition reduction alone: ₹1.4 crore
Cost Savings
| Cost Category | Group A | Group B | Savings |
|---|---|---|---|
| Total salary (500 hires) | ₹41 crore | ₹32 crore | ₹9 crore |
| Attrition replacement | ₹3.6 crore | ₹2.2 crore | ₹1.4 crore |
| Campus visit costs | ₹85 lakh | ₹40 lakh | ₹45 lakh |
| Assessment/interview costs | ₹60 lakh | ₹75 lakh | -₹15 lakh |
| Net savings | ₹10.7 crore |
Full-Scale Rollout
Based on the experiment results, the company rolled out LayersRank for all campus hiring the following year.
- LayersRank invitations sent to all
- AI evaluation + scoring
- Virtual interviews (informed by reports)
- 3,800 offers accepted
Rollout Results
| Metric | Before | After | Change |
|---|---|---|---|
| Applications evaluated | 5,000 (10%) | 41,000 (66%) | +720% |
| Colleges represented in hires | 45 | 180+ | +300% |
| Tier-1 % of hires | 100% | 32% | -68% |
| Average salary cost | ₹8.2 lakh | ₹6.1 lakh | -26% |
| 12-month attrition | 18% | 10% | -44% |
| Diversity improvement | Baseline | +35% | — |
Diversity Impact
An unexpected benefit: Removing college filters dramatically improved diversity.
Gender Diversity
Female representation in hires. Many strong female candidates came from Tier-2 colleges.
Geographic Diversity
Regional representation improved significantly.
Socioeconomic Diversity
First-generation college students in hires.
Key Learnings
What Worked
Running a controlled experiment first.
The A/B test provided irrefutable data. When Group B outperformed Group A, skeptics had no counter-argument. The experiment was essential for overcoming institutional resistance.
Hiding college information from evaluators.
Identity-blind assessment was critical. When interviewers saw reports without college names, they evaluated candidates on substance. Bias didn’t have a channel to operate.
Measuring what matters.
They defined success criteria upfront: training performance, manager satisfaction, project deployment, attrition. Clear metrics enabled clear conclusions.
Starting with campus hiring.
Fresh graduates have limited work history, so credentials matter more than for experienced hires. If capability-based assessment works for campus (where pedigree signal is strongest), it works everywhere.
What They’d Do Differently
Testimonials
“We found talent we would have filtered out before we ever looked at them. That’s the real win. Not just cost savings — finding people we would have missed.”
— VP, Campus Relations
“I was skeptical that a regional college hire could match an IIT hire. The data proved me wrong. Some of our best performers this year came from colleges I’d never heard of.”
— Delivery Manager
“As a Tier-2 college student, I never thought I’d get a chance at a company like this. The assessment gave me a fair shot. I’m grateful for that.”
— Software Engineer, hired from Tier-2 college
Technical Implementation
Scale Challenges
Processing 40,000+ assessments required:
Configuration
- Duration: 60 minutes
- MCQ: 25% of questions
- Short answer: 35%
- Scenario response: 40%
Proctoring & Access
- Proctoring: Light (browser lockdown)
- Camera: Not required
- Accessibility: Screen reader compatible
- Accommodations: Extra time options
Integration
- ATS: Internal campus portal
- Bulk ops: CSV upload/download
- Reporting: College-wise analytics
- Dashboards: Region-wise views
ROI Summary
Investment
| LayersRank enterprise license | ₹45,00,000 |
| Implementation & customization | ₹12,00,000 |
| Internal team training | ₹3,00,000 |
| Total | ₹60,00,000 |
Returns (Annual)
| Salary savings (3,800 hires) | ₹7,98,00,000 |
| Attrition reduction (300 fewer) | ₹1,20,00,000 |
| Campus visit reduction | ₹45,00,000 |
| Interviewer time savings | ₹30,00,000 |
| Total Annual Value | ₹9,93,00,000 |
Annual ROI
1,555%
Payback period: < 1 month
The Bigger Picture
This case study isn’t just about one company’s campus hiring. It’s about a broken assumption in Indian hiring.
The assumption: College pedigree is the best available proxy for capability.
The reality: College pedigree is a convenient filter that excludes most qualified candidates while providing weak predictive signal.
When you measure actual capability — through structured, identity-blind assessment — you find strong candidates everywhere. The talent isn’t concentrated in 50 colleges. It’s distributed across 5,000.
The companies that figure this out first will have access to talent their competitors ignore. They’ll build more diverse teams. They’ll spend less on salaries. They’ll see lower attrition.
The pedigree era is ending. The capability era is beginning.
Related Resources
This case study is based on a real LayersRank deployment at a large IT services company in India. Metrics are actual client data. Company name and identifying details withheld at client request.
For questions about this case study or to discuss campus hiring at scale, contact info@the-algo.com
© 2025 LayersRank by The Algorithm. All rights reserved.