HR MIS Reports: Essential Reports Every SMB Should Track

HR MIS Reports: Essential Reports Every SMB Should Track

HR MIS reports are how a growing company stops managing people by anecdote. When headcount is fifteen, the founder knows who joined, who's unhappy, and what payroll costs. At fifty, that knowledge is folklore; at a hundred and fifty, it's fiction. The management information system (MIS) — the disciplined set of recurring HR reports on headcount, attrition, attendance, leave, payroll cost, recruitment, and compliance — is what replaces folklore with facts.

With AI and people analytics dominating the HR conversation, it's easy to feel you need a data-science team before you can be "data-driven." You don't. The unglamorous truth is that a dozen well-defined MIS reports, produced accurately and reviewed on a fixed cadence, deliver most of the decision value — and they are the prerequisite for any analytics ambition anyway, because analytics built on undefined metrics and dirty master data is just folklore with charts.

This guide covers the essential HR MIS reports for SMBs: what each report contains, the exact metric definitions (where most MIS efforts quietly fail), recommended cadences and audiences, how to build the reporting stack in an HRMS, dashboard design, common pitfalls, and how to graduate from reporting to genuine analytics.

What Is an HR MIS Report?

An HR MIS report is a recurring, standardised report on a defined aspect of the workforce, produced from HR systems for a defined audience and decision purpose. Three words in that definition carry the weight:

• Recurring: monthly headcount is a report; a one-off headcount pull is a query. MIS value comes from trend lines, which require identical definitions applied period after period.
• Standardised: the same metric definitions, filters, and layout every period. If "attrition" is computed three ways by three people, the meeting is about arithmetic instead of decisions.
• Decision purpose: every report should have a named audience and an action it informs. Reports nobody acts on are decoration.

MIS vs analytics vs dashboards

• MIS reporting tells you what happened: headcount grew 6%, attrition was 18% annualised, overtime spiked in the warehouse.
• Analytics asks why and what next: which factors predict attrition, what does absenteeism cost, will the hiring plan meet the growth target.
• Dashboards are a delivery mechanism for either — live views instead of monthly PDFs.

The sequence matters: definitions → clean data → reliable MIS → dashboards → analytics. Skipping ahead produces impressive-looking numbers nobody trusts.

The Core Report Set: What Every SMB Should Run

1. Headcount and workforce composition

The foundation report. Contents: total headcount (opening, joiners, exits, closing), by department, location, employment type (permanent/fixed-term/contract/intern), gender, and grade; span-of-control summary; vacancy count against sanctioned positions.

Definitions to fix: who counts (do interns? notice-period employees? contractor-deployed workers — usually reported separately); as-of convention (last day of month is standard).

Cadence: monthly to leadership; the joiners/exits movement table is the reconciliation spine for every other report — payroll headcount, PF member count, and insurance census should all tie back to it.

2. Attrition / turnover report

The most-watched HR number, and the most frequently miscomputed. A defensible standard:

• Monthly attrition % = exits in month ÷ average headcount in month × 100, where average headcount = (opening + closing) ÷ 2.
• Annualised attrition = trailing 12-month exits ÷ average headcount over the same 12 months — resist multiplying one month by twelve, which turns any bad month into a panic.
• Split voluntary vs involuntary (resignations vs terminations/redundancies) — they demand opposite responses.
• Segment by department, manager, tenure band (0–6m, 6–12m, 1–3y, 3y+), grade, and hiring source. Early attrition (first-year exits ÷ first-year population) is its own metric — it indicts hiring and onboarding, not retention.
• Add regretted vs non-regretted flags if your performance data is mature enough to be honest.

Include reasons (from exit interviews) coded to a fixed taxonomy — compensation, manager, growth, personal, commute/relocation — free-text reasons resist trending.

3. Attendance and absenteeism

Contents: average attendance %, absenteeism % (unplanned absence days ÷ scheduled working days), late-coming and early-leaving counts, overtime hours by department, shift coverage gaps for shift-run operations.

Definitions to fix: what counts as unplanned (sick leave without prior approval, no-shows) versus planned leave; how half-days and regularisations are treated. For plants, warehouses, retail, and field teams this is a weekly — sometimes daily — operational report; for offices, monthly usually suffices.

Watch the interaction with payroll: attendance-to-payroll reconciliation (paid days per payroll vs attendance system) is a monthly control, and discrepancies here are the root of most payslip disputes.

4. Leave report

Contents: leave taken and balances by type (earned/casual/sick/comp-off/maternity etc.), leave liability (balance days × per-day wage — a real balance-sheet number your finance team needs), negative balances, encashment paid, lapse projections before year-end, and sandwich/pattern flags if policy uses them.

Two management signals hide here: employees with excessive unused leave (burnout and key-person risk — some teams literally cannot take leave) and leave pattern anomalies (chronic Monday–Friday absences worth a manager conversation, handled with care and privacy).

5. Payroll cost and compensation report

Contents: gross payroll cost, employer statutory cost (PF, ESI, gratuity accrual), overtime cost, arrears, cost per department and per FTE, month-on-month bridge (headcount change vs increment vs overtime vs one-time items), variable pay payouts, and payroll accuracy indicators (off-cycle corrections, error-driven arrears).

The month-on-month bridge is the piece leadership actually reads: payroll rose 4.2% — how much from hiring, how much from increments, how much from overtime? Producing this reliably requires effect-dated compensation records — an HRMS discipline, not a spreadsheet one.

Add annually: compa-ratio distribution against your salary bands and internal-equity outliers, feeding increment planning.

6. Recruitment funnel report

Contents: open positions and ageing, funnel counts by stage (sourced → screened → interviewed → offered → accepted → joined), time to fill (requisition approval to acceptance) and time to join, offer-acceptance rate, cost per hire (agency + advertising + referral payouts ÷ hires), source mix and source quality (early attrition and probation outcomes by source), and interviewer load.

Definitions to fix: when the time-to-fill clock starts (approval date, not "when the manager first mentioned it") and what closes it. Offer-decline reasons deserve the same fixed taxonomy treatment as exit reasons.

7. Onboarding and probation report

Contents: joiners' onboarding checklist completion (documents, statutory enrolments, asset issuance, training), UAN/ESI enrolment timeliness, 30/60/90 review completion rates, probation decisions due in the next 60 days, confirmation rate, and early-attrition flags. This is the report that prevents the two classic administrative failures: statutory enrolment delays and forgotten probation end dates.

8. Compliance calendar report

Contents: statutory filings and payments due/completed for the period — PF ECR and payment, ESI contribution, TDS deposit and quarterly returns, professional tax by state, LWF where due, S&E and factory renewals, POSH annual report, contract-labour returns as applicable — each with due date, completion date, owner, and proof reference.

This report's audience is leadership precisely because its ideal state is boring: all green. The month it isn't boring, you want to know from your own report — not from a notice.

9. Performance and development report (as maturity grows)

Contents: appraisal completion rates, rating distributions by department (and their sanity — a team of forty with no one below "exceeds" is a calibration failure), goal-setting coverage, PIP counts and outcomes, training hours and completion, internal-mobility moves.

10. Employee lifecycle summary for the board

A one-page quarterly roll-up: headcount trajectory vs plan, attrition trend vs benchmark, payroll cost ratio (people cost as % of revenue — the number boards actually track), open-position ageing, compliance status, and two or three narrative callouts. The discipline of compressing to one page is itself a quality filter for the whole MIS.

Metric Definitions: The Failure Point of Most MIS Efforts

If you take one thing from this guide: write a metric dictionary before you build a single report. One page per metric: name, formula, inclusions/exclusions, data source, owner, and refresh cadence. Examples of the ambiguities that destroy trust when left implicit:

• Does headcount include employees serving notice? (Pick one; disclose it.)
• Is attrition based on last working day or resignation date? (Last working day is standard.)
• Do transfers between departments count as exits in the losing department's attrition? (No — track as internal movement.)
• Is overtime cost in "payroll cost per FTE" or separate? (Separate line; blended numbers hide the signal.)
• Are contractor-deployed workers in absenteeism? (Separate report; different management levers.)

The dictionary turns every future "your number is wrong" meeting into a two-minute lookup.

Cadence and Audience: Who Sees What, When

Two cadence principles: reports must land on a fixed day (the 5th working day for monthly packs is a common standard — payroll-dependent numbers are final by then), and every report ends with "so what" — three bullets of interpretation and proposed action. Numbers without narrative train audiences to skim.

Building the MIS Stack in an HRMS

The spreadsheet era ends the month you have two people maintaining parallel trackers. What the HRMS-based stack looks like:

1. Master data first. Every employee record complete and current: department, location, grade, manager, employment type, dates. Every MIS defect traces back to master data eventually — instituting a monthly master-data audit (missing fields, stale managers, orphan departments) is the highest-ROI reporting investment you can make.
2. Transactions captured at source. Attendance from the attendance system (biometric/GPS/web), leave through workflows (not email), recruitment stages in the ATS, exits through offboarding workflows with coded reasons. If a process bypasses the system, its report is fiction.
3. Standard reports configured once to the metric dictionary — saved filters, saved layouts, scheduled generation and distribution (the 5th-working-day pack emails itself).
4. Role-based dashboards: leadership sees trends; managers see their teams (attendance exceptions, leave calendar, probation dates, open positions); HR sees operations (pending workflows, compliance due dates, data-quality flags).
5. Access control and privacy: salary visibility restricted by role; personal data minimised in distributed packs (aggregate where possible); exports logged. MIS distribution is a data-protection surface — treat it like one under India's DPDP regime.
6. An audit trail for corrections: when March's headcount is restated in May (it happens — backdated exits), the restatement is noted, not silent.

Dashboard Design: Less Is a Feature

• One screen, one question. A leadership dashboard answers "are we growing healthily?" — headcount vs plan, attrition trend, cost ratio, compliance status. Manager dashboards answer "what needs my action this week?"
• Trends over snapshots. A single month's attrition is noise; thirteen months of it is information. Default every metric to a 13-month view.
• Exceptions over averages. Average attendance of 94% is soothing and useless; the list of employees below 85% is actionable.
• Annotate events. Mark the increment month, the restructuring, the new attendance policy on the trend lines — future readers (including you) will otherwise re-litigate known causes.
• Kill vanity metrics annually. Any chart that has never changed a decision gets deleted. Dashboard sprawl is how MIS credibility dies the second time.

Common Pitfalls (and Their Fixes)

• Definition drift: attrition computed differently after a team change → metric dictionary, versioned.
• Report sprawl: forty reports, four read → annual cull; every report needs a named consumer who'd complain if it stopped.
• Accuracy theatre: numbers reconciled nowhere → tie headcount to payroll to PF member count monthly; publish the reconciliation.
• Averages hiding pain: company-level attrition fine, one team at 40% → always segment; the company-level number is for the board, the segmented number is for management.
• Manual assembly: two days of copy-paste per month → automate generation; humans should spend their time on the "so what," not the collation.
• Privacy carelessness: salary sheets forwarded, individual health-related absence discussed in open packs → role-based access, aggregation thresholds, and a distribution policy.
• Analytics before hygiene: an attrition-prediction model on top of undefined attrition → walk the sequence: definitions, data, MIS, then analytics.

From MIS to People Analytics

Once the core set runs reliably for a few quarters, graduation looks like:

1. Correlation questions: does early attrition vary by hiring source? Does overtime predict resignations in ops? Does time-to-fill move offer-acceptance? Your own MIS history answers these with simple analysis — no ML required.
2. Cost quantification: put rupee values on attrition (replacement cost per exit), absenteeism (lost days × loaded cost), and overtime — costs convert HR observations into CFO decisions.
3. Leading indicators: engagement pulse scores, internal-mobility rates, review-completion rates — metrics that move before attrition does.
4. Predictive pilots, carefully: attrition-risk scoring can genuinely help retention if used ethically (manager conversations, not automated actions), with transparency about what data feeds it. AI-assisted analytics is only as good as the MIS beneath it — which is why the boring reports were the point all along.

Data Quality: The Ten Checks That Keep MIS Honest

Reliable reports rest on a small set of recurring data-quality controls. Run these monthly, ideally as an automated exception report:

1. Orphan records: employees with no manager, no department, or no grade assigned.
2. Stale managers: reporting lines pointing at exited employees.
3. Date sanity: exits dated before joins; confirmations before joining dates; future-dated transactions older than a week.
4. Movement reconciliation: opening headcount + joiners − exits = closing headcount, exactly — any gap means an uncaptured movement.
5. Payroll tie-out: employees paid this month who aren't active in HR, and active employees who weren't paid — both lists should be explainable (F&F cases, unpaid leave) or empty.
6. Statutory tie-out: PF member count in the month's ECR vs eligible active headcount; ESI-covered count vs wage-band population.
7. Duplicate detection: same PAN or UAN across two employee records — usually a rehire handled as a fresh record.
8. Leave-balance integrity: negative balances beyond policy tolerance; balances that jumped without a transaction.
9. Attendance coverage: employees with zero attendance data but full paid days (device mapping failures for new joiners are the classic cause).
10. Exit-reason coverage: percentage of exits carrying a coded reason — target 100%; "unknown" exits destroy attrition analysis a year later.

Publish the exception counts themselves as a small data-quality scorecard in the HR pack. When the counts trend to zero and stay there, your MIS has earned the trust that analytics will later spend.

A 90-Day Implementation Roadmap

You can stand up a credible MIS in one quarter. A realistic sequence for an SMB starting from spreadsheets:

Days 1–30: Definitions and data

• Write the metric dictionary for the first six reports (headcount, attrition, attendance, leave, payroll cost, compliance calendar). Get finance to co-sign the payroll and headcount definitions — shared definitions end the HR-vs-finance number wars permanently.
• Audit the employee master: export everything, list missing/stale fields (manager, department, grade, employment type, dates), and run a two-week cleanup with department heads confirming their rosters.
• Fix the movement log: reconstruct joiners and exits for the trailing twelve months with dates and coded exit reasons. This backfills your first trend lines.
• Decide the reporting calendar: which day each report lands, who produces it, who receives it.

Days 31–60: First production cycle

• Produce the first monthly pack manually if needed — the point of month two is exercising the definitions, not automation. Expect reconciliation failures (headcount vs payroll vs PF member count); fixing them now is the actual work.
• Configure the reports in the HRMS to match the dictionary; run system-generated and manual versions in parallel once.
• Start the compliance calendar immediately — it needs no history and delivers value from week one.
• Pilot one manager dashboard with a friendly department head; their confusion is your design feedback.

Days 61–90: Automation and cadence

• Switch to system-generated reports with scheduled distribution; retire the manual versions.
• Add the "so what" discipline: every report gets three bullets of interpretation before it ships.
• Present the first full pack to leadership with thirteen months of trend where reconstructable; annotate known events.
• Book the quarterly review: cull unread reports, log definition-change requests, and pick the first analytics question for next quarter.

The trap to avoid in all three phases: building charts before definitions. A beautiful dashboard on disputed numbers sets the program back a year, because the first credibility loss is the expensive one.

Reading a Monthly Pack: A Worked Narrative

What good MIS review actually sounds like, using an illustrative 120-person company:

Headcount closed at 124 against a plan of 130 — the gap is entirely two engineering backfills ageing past 60 days, which the recruitment funnel confirms: both roles have healthy top-of-funnel but a 40% interview no-show rate, so the action is interviewer scheduling, not sourcing spend. Attrition printed 1.6% for the month (≈19% annualised, roughly flat on trend), but the segmentation shows three of the four exits came from one support team with a new manager — a skip-level conversation is scheduled before this becomes a trend. Absenteeism is stable at 2.1% except the warehouse, which jumped to 5% and correlates exactly with the overtime spike in the payroll bridge — the roster is understaffed, and the cheaper fix is two hires, not continued double-rate overtime that is also fatiguing the crew. Leave liability grew ₹3.2 lakh this quarter, concentrated in engineering where utilisation is 40% of accrual — leadership agrees to enforce a minimum-leave nudge before Q4 lapse season. The compliance calendar is green except one state's professional tax filed two days late (bank holiday miss) — the payment calendar now front-runs holidays by three days. Total meeting time: twenty-five minutes, five decisions.

That is the standard to aim for: every number either confirms health in a sentence or produces an owner and an action.

Benchmarks and Context for Indian SMBs

Benchmarks vary widely by sector, city, and labour-market cycle, so treat published figures as orientation, not targets — your own trend is the benchmark that matters most. With that caveat, orientation ranges commonly discussed for India:

• Attrition: IT services and BPO have historically run high (often well into the twenties annualised in hot markets); manufacturing and engineering typically lower; early-stage startups vary wildly. Whatever the level, rising early attrition is the signal that most reliably indicts internal process.
• Absenteeism: low single digits for offices; higher and more seasonal for plants and field forces (harvest, festivals, examination seasons all show up in Indian attendance data — annotate them).
• Offer acceptance: below ~80% sustained usually points at compensation positioning or process speed; measure declines by reason.
• People cost ratio: tracked universally by boards, but sensible only against same-sector peers; the useful discipline is explaining its movement via the payroll bridge, not hitting an abstract number.

Where possible, benchmark against your own history first, sector surveys second, and anecdotes never.

Roles and Responsibilities: A Minimal RACI

Even a two-person HR team benefits from explicit ownership:

• Report owner (HR ops / HR generalist): produces each report on schedule, maintains the metric dictionary, runs reconciliations, flags data-quality issues. Accountable for accuracy.
• HR head: owns the "so what" narrative, presents the leadership pack, arbitrates definition disputes, approves restatements.
• Finance partner: co-owns payroll cost and leave liability; consumes headcount for budgeting; the monthly reconciliation is a joint sign-off.
• Line managers: consume team dashboards, act on exceptions (attendance, probation dates, open positions), and keep their rosters' master data honest — manager confirmation of team lists is the cheapest master-data control there is.
• Leadership: reviews the pack on cadence, makes the decisions the reports exist to inform, and protects the definitions from convenience-driven tampering (the request to "compute attrition the other way this quarter" always comes eventually; the answer is no).

Choosing Reporting Tooling: What Actually Matters in an HRMS

When evaluating HRMS reporting capabilities, the demo dazzle (animated charts, AI summaries) matters far less than five mundane capabilities:

• Effect-dated data. Can the system tell you what an employee's department, salary, and manager were as of last March, not just today? Without effective dating, every historical trend silently rewrites itself after transfers and revisions.
• Custom fields in reports. Your business will have dimensions the vendor didn't anticipate (client account, project, cost centre variant); reports must filter and group by them.
• Scheduling and distribution. Reports that generate and email themselves on the 5th working day, with role-appropriate versions per recipient.
• Export discipline. Clean CSV/Excel exports with stable column schemas — because finance will import them, and a vendor who reshuffles columns in an update breaks month-end downstream.
• Point-in-time snapshots. The ability to freeze and archive the month's pack as issued, so restatements are visible rather than silent.

Ask each vendor to reproduce, live, your attrition number for last quarter from sample data using your dictionary definition. The tools that can, can; the tools that offer their own hard-coded definition instead are selling you a definitions dispute with your own history.

Frequently Asked Questions

1. What are HR MIS reports? Recurring, standardised reports on workforce data — headcount, attrition, attendance, leave, payroll cost, recruitment, compliance — produced from HR systems on a fixed cadence for defined audiences and decisions.

2. What is the correct formula for attrition rate? Monthly: exits ÷ average headcount ((opening + closing)/2) × 100. Annual: trailing 12-month exits ÷ 12-month average headcount. Split voluntary/involuntary and segment by tenure, team, and source — and don't annualise a single month by multiplying by twelve.

3. Which HR reports should an SMB start with? Headcount with movement, attrition, attendance/absenteeism, leave with liability, payroll cost with a month-on-month bridge, and a compliance calendar. Add recruitment funnel and probation tracking as hiring scales.

4. How often should HR MIS reports be produced? Monthly for most (landing on a fixed day, e.g. the 5th working day), weekly for operational attendance and active-hiring funnels, quarterly for performance and the board one-pager. Consistency of cadence matters more than frequency.

5. What is leave liability and why does finance care? The monetary value of accumulated encashable leave balances (days × per-day wage). It is a genuine liability for the balance sheet, and it grows silently unless leave utilisation is managed.

6. Can we run MIS from spreadsheets? At very small headcounts, yes — briefly. Spreadsheets fail at exactly the things MIS needs: consistent definitions, source-captured transactions, access control, and audit trails. The transition point is typically 30–50 employees or the first time two trackers disagree in a leadership meeting.

7. How do we keep MIS reports trustworthy? A written metric dictionary, monthly reconciliation (headcount ↔ payroll ↔ PF member count), master-data audits, automated generation, and annotated restatements. Trust is lost through silent definition changes, not honest corrections.

8. What's the difference between MIS reporting and people analytics? MIS describes what happened with standardised metrics; analytics explains why and predicts what's next. Analytics is only as reliable as the MIS layer beneath it — build the reports first.

Conclusion

An HR MIS is not a technology project; it is a definitions-and-discipline project that technology then automates. Fix the metric dictionary, capture transactions at source, run a dozen reports on an unbreakable cadence, reconcile them against payroll and statutory filings, and end every report with a recommendation. Do that for four quarters and your company will make visibly better people decisions — and be ready for the analytics layer everyone is excited about.

CozyHR gives SMBs the whole stack out of the box: clean employee master data, attendance and leave captured through workflows, payroll that reconciles to the rupee, and scheduled MIS reports and dashboards for leadership, managers, and HR. Try CozyHR and turn your monthly reporting scramble into a five-minute review.