Recruitment and selection

In short

Recruitment is the process of searching for and attracting a pool of qualified applicants for a job; selection is the process of screening that pool and choosing the person who best fits the job and the organisation. Recruitment is positive (it widens the pool), selection is negative (it rejects to narrow it down) — and the two together deliver the "staffing" function of HRM. It matters because the quality of every later HR activity (training, performance, retention, cost) is capped by who was hired, and because hiring is legally exposed: equal-opportunity rules, the Apprentices Act, 1961, child-labour prohibition and now the four labour codes all touch it. It is a guaranteed question area in BBA/BMS university papers (short notes, process diagrams, source comparisons, reliability-validity), in placement aptitude-plus-HR rounds at TCS, Infosys, Wipro, Accenture, Deloitte, HDFC Bank, Teamlease and Randstad, and in interviews for HR Executive, Talent Acquisition and recruitment-consulting roles.

Explained

Concepts, scope and the foundation in job analysis

• Recruitment: Edwin Flippo — "the process of searching for prospective employees and stimulating them to apply for jobs in the organisation." It generates applications; it does not decide.
• Selection: choosing from among applicants the one(s) whose abilities and attributes best match job requirements; it is a *series of hurdles*, each eliminating some candidates.
• Recruitment is positive, selection is negative: recruitment invites, selection rejects. Recruitment precedes selection; selection cannot be better than the pool recruitment created.
• Placement = assigning the selected person to a specific job/position; induction/orientation = socialising the new hire; onboarding = the whole first 30-90 days. These follow selection and are part of the staffing chain.
• Human resource planning (HRP) supplies the input: demand forecast − internal supply = the manpower gap to be recruited. No gap, no recruitment requisition.
• Job analysis is the base document for both. Its two outputs:
  • Job description (JD) — job-centred: title, reporting line, duties, working conditions, authority. Answers "what the job is."
  • Job specification (JS) — person-centred: qualification, experience, skills, aptitude, physical and behavioural traits. Answers "who can do it." Selection tests and interview criteria must be drawn from here.
• Competency-based hiring: instead of listing degrees, the JS lists observable competencies (e.g. "handles customer objection without escalation"), each assessed by a defined method — the dominant practice in Indian IT and BFSI today.
• Recruitment policy should state: sources permitted, internal-first rule, reservation/diversity commitments, approval authority (usually CHRO or budget holder), and cost ceilings.

Sources and methods of recruitment

Internal sources (filling from within)

• Promotion — vertical move, higher pay and responsibility; transfer — lateral move, same grade; upgradation — same job, higher grade.
• Internal job posting (IJP) — vacancy advertised on the intranet; employees apply themselves. Standard at Infosys, TCS, ICICI Bank.
• Employee referral — current employees refer outsiders; usually paid a referral bonus (commonly ₹10,000-₹1,00,000 by level, payable after the referee completes 3-6 months). Technically an external source routed internally; cheapest high-quality channel.
• Re-employment / rehire — ex-employees ("boomerang" hires), retrenched workers (Industrial Disputes Act, 1947 s.25H gives retrenched workers preference in re-employment), retirees on contract.
• Advantages: cheaper, faster, known performer, improves morale and retention, shorter induction.
• Limitations: no fresh ideas (inbreeding), limited pool, internal politics and "Peter principle" promotions, creates a fresh vacancy lower down.

External sources

1. Campus recruitment / placement — tied to colleges; day-0/day-1 slots, pre-placement talks, pooled campus drives. The main fresher channel for IT services, banks and FMCG management trainees.
2. Advertisement — print (classifieds), trade journals, and increasingly digital. Two forms: open ad (company named) and blind ad (box number, identity hidden).
3. Job portals and e-recruitment — Naukri, LinkedIn, Indeed, Shine, Internshala, Hirect; company career pages with an ATS (applicant tracking system).
4. Placement consultants / recruitment agencies — contingency search for junior-mid roles (fee typically 8.33%-16.67% of annual CTC, i.e. one to two months' salary); executive search / headhunting for senior roles on retainer (often 25-33% of CTC). Players: ABC Consultants, Randstad, Michael Page, Korn Ferry.
5. Staffing / temp firms and RPO — Teamlease, Quess, Adecco supply contract staff; recruitment process outsourcing hands the whole hiring process to a vendor.
6. Employment exchanges — government exchanges under the Employment Exchanges (Compulsory Notification of Vacancies) Act, 1959: notification of vacancies is compulsory for covered establishments, but the employer is *not* bound to hire the candidates sent. The National Career Service (NCS) portal is the digital successor.
7. Walk-ins, write-ins and talk-ins — common in BPO, retail, insurance and healthcare; mass walk-in drives for voice processes.
8. Deputation, labour contractors, gate hiring, badli/casual rolls — manufacturing and construction; contract labour is governed by the Contract Labour (Regulation and Abolition) Act, 1970.
9. Apprentices and trainees — under the Apprentices Act, 1961 (with NAPS incentives); apprentices are not "workmen" for most purposes, and the Act was amended to make engagement largely enabling rather than penal.
10. Social media and talent communities — LinkedIn sourcing, GitHub/Stack Overflow for developers, Telegram/WhatsApp groups for blue-collar; employer branding and poaching/raiding from competitors.

Comparison of recruitment practice: internal sources dominate above mid-management in Indian PSUs and banks; external dominates for freshers and niche digital skills.

The recruitment process, step by step

1. Manpower requisition and approval — raised by the line manager against an approved budget/headcount.
2. Review JD and JS; define the hiring scorecard — update, don't recycle, the old JD.
3. Choose sourcing strategy — internal first (IJP window of 7-15 days is typical), then referral, then external.
4. Create and position the job ad — role, location, CTC band, must-have vs good-to-have, EEO statement.
5. Source and build the pipeline — searches, portals, consultants, campus calendar.
6. Screen applications — against the JS; shortlist ratio fixed in advance.
7. Hand over to selection — shortlisted pool enters the selection funnel.
8. Evaluate the channel — cost per hire, source-of-hire quality, time to fill; feed back into next cycle.

Factors affecting recruitment

• *Internal:* size of firm, employer brand, pay competitiveness, growth and promotion policy, cost, existing recruitment policy and trade-union agreements.
• *External:* labour-market supply-demand, unemployment rate, competitor hiring, legal and reservation requirements, education system output, IT-sector attrition levels.

The selection process and selection tests

Standard sequence (hurdles model)

1. Preliminary screening / CV screening — knock out those failing minimum eligibility; often automated by ATS keyword and GPA filters.
2. Application blank — structured data capture; WAB (weighted application blank) scores predictive items.
3. Written / online tests — see below.
4. Group discussion or group exercise — used in campus hiring, MBA entry and PSU recruitment.
5. Technical interview — domain depth.
6. HR / behavioural interview — fit, stability, salary, relocation.
7. Reference and background verification (BGV) — education, employment, criminal, address, drug test where relevant; done by vendors such as AuthBridge, First Advantage. Also social-media screening (legally sensitive).
8. Medical examination — statutory for hazardous processes under the Factories Act, 1948.
9. Approval and offer letter → acceptance → joining and induction.
10. At each stage rejection is final; a *successive hurdles* model is cheap but can reject someone strong overall, while a compensatory model (totalling scores) allows strengths to offset weaknesses.

Types of selection tests

• Aptitude / cognitive ability tests — general mental ability, numerical, verbal, logical reasoning, data interpretation. Highest general validity of any single predictor.
• Achievement / proficiency / trade tests — what the candidate already knows or can do: coding tests (HackerRank, Codility), typing speed, welding trade test, accounting test.
• Personality tests — self-report inventories measuring traits; Big Five / OCEAN is the research-supported model; MBTI is popular in corporate workshops but weak as a selection tool.
• Interest inventories — occupational preference, used more in career guidance than hiring.
• Projective tests — TAT, Rorschach, Sentence Completion; low reliability, rarely defensible in selection.
• Situational Judgement Test (SJT) — scenario with response options; widely used in BPO and banking.
• Work sample / in-tray exercise — a miniature of the actual job; highest face validity.
• Assessment centre — multiple exercises (case, GD, role play, presentation, in-basket) assessed by multiple trained assessors against defined competencies; used for officer-grade and leadership hiring (SBI, LIC, Hindustan Unilever, armed-forces SSB).
• Polygraph, graphology, astrology-based screening — unscientific; avoid.

Types of interview

• By structure: structured (same questions, anchored rating scale — most valid), semi-structured, unstructured.
• By method: one-to-one, panel/board, sequential, group, telephonic, video (asynchronous one-way video is now common in volume hiring), stress interview, depth interview.
• By question type: behavioural / BEI ("tell me about a time you..."), situational ("what would you do if..."), case, technical, puzzle/brain-teaser (poor predictor).
• STAR (Situation-Task-Action-Result) is the standard answering and scoring framework.
• Interview errors: halo and horn effect, similar-to-me bias, first-impression and primacy effect, contrast error, leniency/severity, central tendency, stereotyping, recency, and the untrained interviewer talking more than the candidate.

Measuring quality: reliability, validity and fairness

• Reliability = consistency of the measure. Types: test-retest, parallel/equivalent forms, split-half and internal consistency (Cronbach's alpha), inter-rater reliability.
• Validity = does it predict job performance? Types:
  • Content validity — the test samples the job's actual content (trade test).
  • Construct validity — it really measures the trait it claims.
  • Criterion validity — statistical link to a performance criterion: predictive (test now, performance later) and concurrent (both measured now).
  • Face validity — looks relevant to candidates; affects acceptance, not accuracy.
• A test can be reliable but invalid; it cannot be valid without being reliable. Reliability is a necessary but not sufficient condition.
• Selection errors: false positive (hired, fails — cost of poor performance and exit) and false negative (rejected, would have succeeded — invisible but real cost).
• Validity coefficient — correlation between test score and performance; typical indicative values: cognitive ability tests and structured interviews around 0.5, work samples around 0.5, unstructured interviews around 0.2-0.3, graphology near zero. Figures vary by study and by meta-analytic correction method, so quote them as approximate.
• Adverse impact — a neutral-looking criterion that disproportionately screens out a protected group. India has no exact equivalent of the US "four-fifths rule", but constitutional Articles 14-16 and reservation policy bind public employment, and the Rights of Persons with Disabilities Act, 2016 mandates 4% reservation in government posts and non-discrimination in private hiring.
• Other legal touchpoints: Equal Remuneration provisions now folded into the Code on Wages, 2019; the Industrial Relations Code, 2020, Social Security Code, 2020 and OSH Code, 2020 complete the four labour codes — their implementation has been phased and state rules differ, so verify current status before quoting. Child Labour (Prohibition and Regulation) Act bars employment below 14. The Maternity Benefit Act bars dismissal for pregnancy; asking marital or pregnancy status in interviews is poor and risky practice. POSH Act, 2013 compliance should be disclosed at onboarding.

Recruitment metrics and current Indian practice

• Key metrics: yield ratio, selection ratio, time to fill, time to hire, cost per hire, offer-acceptance rate, offer-to-join (drop-off) ratio, quality of hire, first-year attrition, source-of-hire mix, applicant satisfaction.
• Offer shopping and dual offers made offer-to-join ratios a headline issue in Indian IT during 2021-22; "buffer hiring" and joining bonuses were the response.
• Notice period and relieving letter — 30-90 days is normal in India; BGV routinely requires the relieving letter and Form 16/payslips. Non-compete clauses are generally unenforceable in India (s.27, Indian Contract Act, 1872, restraint of trade void), though confidentiality and non-solicit clauses are enforceable.
• Technology: ATS (Workday, SuccessFactors, Darwinbox, Zoho Recruit), chatbot screening, AI resume parsing and ranking, proctored online assessments (AMCAT, eLitmus, SHL, Mettl/Mercer, TCS NQT), video interviews (HireVue-style). AI screening raises bias-audit and explainability concerns; keep a human in the loop.
• Gig and contingent hiring: platform work (Swiggy, Zomato, Urban Company) is recognised in the Social Security Code, 2020 with aggregator contributions — a growing share of "hiring" that is not employment.
• Blue-collar and rural hiring: Apna, WorkIndia, Vahan; Skill India/NSDC certification as a screening signal.
• Diversity hiring: returnship programmes for women after a career break, PwD hiring targets, hiring from tier-2/tier-3 campuses.

Key points

• Recruitment - searching for and attracting applicants; a positive, pool-widening process.
• Selection - screening and choosing; a negative, elimination process using successive hurdles.
• Job analysis - the parent activity; yields the job description (the job) and job specification (the person).
• HRP gap - demand forecast minus internal supply; the trigger for any requisition.
• Internal sources - promotion, transfer, IJP, rehire; cheap, fast, motivating, but risk inbreeding.
• External sources - campus, portals, consultants, agencies, walk-ins, employment exchanges, apprentices.
• Employee referral - usually the best cost-to-quality channel; paid as a bonus after the referee's confirmation.
• Employment Exchanges Act, 1959 - compels *notification* of vacancies, not appointment of the candidates sent.
• Yield ratio - output of a stage ÷ input to it; chains together into the hiring funnel.
• Selection ratio - number selected ÷ number of applicants; a smaller ratio means a more selective (usually higher-quality) hire.
• Reliability - consistency of a measure (test-retest, split-half, inter-rater); necessary but not sufficient.
• Validity - predictive power (content, construct, criterion, face); the real test of a selection tool.
• Structured interview - same questions, anchored scale; roughly twice as predictive as an unstructured one.
• Assessment centre - multiple exercises × multiple assessors × defined competencies; costly, high validity.
• Interview errors - halo/horn, similar-to-me, first impression, contrast, leniency, central tendency, stereotyping.
• False positive vs false negative - hiring a failure vs rejecting a success; both are selection errors.
• Cost per hire - (internal + external recruiting costs) ÷ number of hires.
• Non-compete - void under s.27 of the Indian Contract Act; confidentiality and non-solicit survive.

Formulas and examples

Example 1 — Hiring funnel and yield ratios. A BPO in Hyderabad runs a walk-in drive: 2,000 applicants → 800 clear the aptitude test → 300 clear the voice-and-accent round → 120 clear the HR interview → 100 offers made → 80 join.

• Cumulative yield (applicants → joiners) = 80/2,000 = 4%, i.e. 25 applicants per joiner.
• Selection ratio (offers ÷ applicants) = 100/2,000 = 0.05, or 1 : 20.
• Back-calculation: to get 200 joiners next quarter, applicants needed = 200 ÷ 0.04 = 5,000.

Example 2 — Cost per hire, two channels compared. 50 engineers hired: 30 through a consultant, 20 through campus.

• Consultant fee at 8.33% of average CTC ₹6,00,000 = ₹49,980 per hire × 30 = ₹14,99,400.
• Campus cost: travel and logistics ₹1,80,000 + pre-placement talks and branding ₹1,20,000 + assessment vendor ₹60,000 = ₹3,60,000 for 20 hires = ₹18,000 per hire.
• Internal cost (recruiter salaries apportioned, ATS licence) = ₹3,00,000, allocated 30:20 → ₹1,80,000 and ₹1,20,000.
• Consultant channel = (₹14,99,400 + ₹1,80,000)/30 = ₹55,980 per hire; campus channel = (₹3,60,000 + ₹1,20,000)/20 = ₹24,000 per hire.
• Overall cost per hire = (₹16,79,400 + ₹4,80,000)/50 = ₹21,59,400/50 = ₹43,188.
• Reading: campus is ~2.3× cheaper per hire, but it delivers freshers with a 3-6 month productivity lag — compare cost *and* quality of hire, not cost alone.

Example 3 — Manpower gap and recruitment need. A retail chain has 400 store staff, needs 480 next year, expects 18% annual attrition and plans 20 internal promotions out of the cadre. Offer-acceptance rate is 85% and joining ratio 90%.

• Attrition loss = 18% of 400 = 72. Internal supply at year end = 400 − 72 − 20 = 308.
• Manpower gap = 480 − 308 = 172 hires needed.
• Offers required = 172 ÷ (0.85 × 0.90) = 172 ÷ 0.765 = 225 offers (224.8, rounded up).
• At a cumulative yield of 5%, applicants required = 225 ÷ 0.05 = 4,500.

Example 4 — Validity and selection errors. 200 candidates took an aptitude test; 100 scored above the cut-off and were hired. After a year, 70 of the 100 hired were rated successful. A follow-up study on 50 rejected candidates hired elsewhere found 15 had succeeded.

• Success rate of the selected group = 70/100 = 70%; false positives = 30/100 = 30%.
• Estimated success rate among rejects = 15/50 = 30% → false negatives exist; the test is useful but not decisive.
• Base rate (success without the test) ≈ weighted = (70 + 15×(100/50)) ÷ 200 = (70 + 30)/200 = 50%.
• Incremental validity: 70% vs a 50% base rate = +20 percentage points improvement in success probability from using the test. Raising the cut-off would increase the hit rate but reject more good candidates and raise the cost per hire.

Example 5 — Compensatory vs hurdles model. Three finalists; weights: technical 40%, aptitude 30%, interview 30%; stage cut-off (hurdles model) = 50 in every component.

• Compensatory model picks A (72.5); the hurdles model eliminates A and picks C (67.5). The choice of model, not the scores, decides the hire — so the model must be fixed before the process starts, and the hurdle must be justified by the job.

Where it is used

• IT and ITeS: TCS NQT, Infosys InfyTQ, Wipro Elite and Cognizant GenC — pooled campus tests feeding lakhs of fresher offers, plus lateral hiring on Naukri/LinkedIn through an ATS.
• Banking, insurance and finance: IBPS and SBI PO/Clerk exams (prelims-mains-interview hurdles), LIC Development Officers, HDFC and ICICI relationship-manager walk-ins and referral-heavy branch hiring.
• Government and PSUs: UPSC, SSC, railway RRB and NTPC/ONGC/BHEL recruitment through GATE scores; reservation rosters and employment-exchange notification compliance.
• Manufacturing and construction: apprentice intake under the Apprentices Act and NAPS, ITI trade tests, gate hiring and contract labour through registered contractors (L&T, Tata Motors, Maruti).
• Staffing, consulting and HR-tech roles: Teamlease, Quess, Randstad, ABC Consultants; job titles such as Talent Acquisition Executive, Recruiter, Sourcing Specialist, HRBP, Campus Relations Manager, RPO delivery lead.
• Retail, BPO, logistics and gig platforms: mass hiring at scale — Reliance Retail, Flipkart and Amazon seasonal hiring, Swiggy/Zomato/Urban Company partner onboarding through Apna, WorkIndia and Vahan.

Interview questions

• Q: How is recruitment different from selection? A: Recruitment searches for and attracts applicants, widening the pool, so it is a positive process; selection screens and rejects to pick the best fit, so it is negative, and it always follows recruitment.
• Q: What is the difference between a job description and a job specification? A: A job description states the duties, responsibilities and conditions of the job; a job specification states the qualifications, skills and traits the person must have to do it.
• Q: Why do many firms prefer employee referrals? A: Referrals are cheap, fast, and pre-screened by someone who knows both the job and the candidate, and referred hires typically show better retention — the risk is a less diverse pipeline.
• Q: Can a test be reliable but not valid? A: Yes — a test can give consistent scores and still measure the wrong thing; reliability is necessary for validity but not sufficient.
• Q: What is a false negative in selection, and why does it matter? A: Rejecting a candidate who would have performed well; it is invisible in company data but it wastes good talent and often shows up as a competitor's strong hire.
• Q: Why is a structured interview preferred over an unstructured one? A: Every candidate faces the same job-related questions scored on an anchored scale, which cuts interviewer bias and roughly doubles predictive validity while making the decision defensible.
• Q: What does the Employment Exchanges Act, 1959 actually require? A: Covered establishments must notify vacancies to the employment exchange, but they are free to select whoever they want — notification is mandatory, appointment of the exchange's candidates is not.
• Q: Are non-compete clauses in Indian offer letters enforceable? A: Generally no — section 27 of the Indian Contract Act, 1872 voids agreements in restraint of trade after employment ends; confidentiality and non-solicitation clauses are enforceable.
• Q: A manager wants to hire someone "just like the last person who succeeded in this role." What is the risk? A: That is similar-to-me and cloning bias; it narrows diversity, mistakes surface similarity for competence and ignores whether the role itself has changed.
• Q: Your offer-to-join ratio has fallen to 60%. What do you do? A: Diagnose where candidates are dropping out — benchmark the CTC band, shorten the offer-to-joining gap, keep candidates engaged during notice period, and hire a planned buffer while fixing the root cause.
• Q: How would you select for a role with no existing incumbent and no JD? A: Build the JD from the business outcome backwards with the hiring manager, convert it into 4-6 testable competencies, and assess with a work-sample or case exercise rather than a credential filter.
• Q: Is AI-based resume screening safe to use? A: It saves time at scale but learns from past hiring decisions, so it can reproduce historic bias; it needs regular adverse-impact audits, explainable criteria and a human reviewer on every rejection at the margin.

Common mistakes

• Treating recruitment and selection as one process, or calling selection "positive" — examiners look for the positive/negative distinction and the order.
• Swapping job description and job specification in answers; JD describes the job, JS describes the person.
• Listing employee referral or campus placement as an *internal* source — both bring in outsiders.
• Claiming the employment exchange's candidates must be appointed; only the notification of the vacancy is compulsory.
• Confusing reliability with validity, or asserting that a valid test may be unreliable (it cannot be).
• Computing a yield ratio against the original applicant pool instead of the candidates entering that specific stage — and quoting ratios without stating the denominator.
• In interviews, giving opinions instead of evidence: answer competency questions in STAR form with a measurable result, and never claim a selection method is "best" without naming the job it suits.

Related topics

Human resource management · Training and development · Talent management and succession planning · HR analytics · Performance management and appraisal · Employee relations and labour laws