Almost every category of litigation on a company's docket arrives from outside. A customer sues, a regulator issues a notice, a competitor moves for an injunction, a landlord changes its mind. Employment litigation is different. It is generated inside the building, by decisions the company made, in processes the company controlled, on timelines the company set. That is uncomfortable, and it is also the good news, because a dispute you manufacture is a dispute you can decline to manufacture.
In-house teams tend to encounter employment matters one at a time and late. A separation goes wrong in March, a notice arrives in September, and by the time counsel is briefed, the facts are fixed and the only question left is how much the company will pay to make it stop. The lawyer is being asked to defend a decision that was made without legal input, recorded by someone who did not expect to be cross-examined on it, and documented in a way that reads worse than the underlying conduct actually was. The merits are often defensible. The record almost never is.
That gap between what happened and what can be shown to have happened is the whole subject of this article. It is written for in-house counsel, heads of HR and legal operations leads at Indian companies, and it makes one argument throughout: employment litigation is the most preventable category on your docket, and it is preventable through documentation habits and process discipline rather than through better lawyering after the fact.
The argument in one paragraph
Employment disputes are not idiosyncratic. They arrive in a small number of repeating fact patterns, and each pattern turns on whether a contemporaneous record exists. The company almost never loses because the decision was wrong. It loses because the decision cannot be shown to have been taken the way the company now says it was taken. Prevention is therefore not about being kinder or slower. It is about creating, at the time, the ordinary paper that a fair process naturally produces.
Why This Category Is Preventable and Others Are Not
Consider what a commercial dispute looks like from the inside. Two sophisticated parties negotiated terms, both had counsel, the contract is the contract, and the disagreement is usually about interpretation or about money that was always going to be contested. There is not much a general counsel can do in advance beyond drafting well, and even excellent drafting does not stop a counterparty who has decided to fight.
Employment matters do not work that way. The employee did not negotiate the appointment letter. The company chose the policy, ran the appraisal, decided the separation, drafted the letter, calculated the settlement and picked the date. Every material fact in the eventual dispute was authored by one side, and that side was you. This asymmetry cuts both ways. It means the company is exposed when it is careless, and it means the company holds nearly every lever that would have prevented the exposure.
There is a second structural feature worth naming. Employment adjudication in India is protective by design and heavily procedural in practice. Forums hearing these matters look closely at whether the employee was told what was alleged, given a chance to respond, and dealt with in the manner the company's own policy promised. A company that followed its policy will usually be fine even where the substantive judgement was harsh. A company that reached a defensible conclusion by an indefensible route frequently loses. Once you internalise that, prevention stops looking like a compliance overhead and starts looking like the cheapest litigation strategy available.
In this category the company writes the evidence before it knows there will be a case. That is the entire opportunity, and the entire risk.
The Fact Patterns That Recur
If you read a year of a mid-sized Indian company's employment matters, you will not find a hundred different stories. You will find seven or eight, repeated. Naming them precisely is useful, because each has a distinct failure point and a distinct fix.
Termination without a defensible record
This is the dominant pattern. A manager loses confidence in someone over months, raises it verbally two or three times, escalates to HR, and the separation happens quickly once the decision is taken. Then the file is opened and there is nothing in it. No written feedback, no note of the conversations, no record of what was expected and by when. The company's position becomes an assertion, made after the fact, by people with an obvious interest in the outcome. The employee's position, that the exit was arbitrary and the performance narrative was constructed afterwards, is not obviously worse supported. Separate the substantive question, was this person underperforming, from the evidential one, can you show it, and you will find the second is where these matters are actually decided.
The resignation that was not really voluntary
A close second, and the one companies most often walk into thinking they have avoided a problem. Someone is told that the company would prefer a resignation, that a resignation looks better on a record, and that the alternative is a termination they will have to explain for years. The letter is signed the same afternoon, sometimes on a form the company supplied, sometimes drafted by HR and typed out by the employee. Months later the position advanced is that the resignation was extracted under pressure and should be disregarded. The company then has to prove the absence of coercion, which is a difficult negative to prove when the meeting was unminuted, the letter was produced within an hour, and the wording matches the company's template. Where a separation is genuinely consensual, treat the paperwork as though it will be challenged: give real time to consider, avoid drafting the letter, and record the terms in a settlement document that recites what was offered and accepted.
The performance improvement plan used as a paper trail rather than a process
A performance improvement plan is a legitimate management instrument and a poor disguise. Companies that use it as a genuine process, with specific and measurable expectations, an honest support obligation on the manager, scheduled reviews that actually occurred, and a real possibility of the employee passing, are in a strong position afterwards. Companies that issue a plan after the exit decision has already been made are usually worse off than if they had issued nothing at all, because the plan documents its own artificiality. A thirty day plan with vague goals, no interim reviews, and a termination on day thirty one is not evidence of fairness. It is evidence that the company knew what fairness required and staged it.
Notice pay and the full and final settlement
A large share of employment litigation is not about dignity at all. It is about money that was calculated wrongly or slowly. Notice pay computed on basic rather than the contractual figure, unused leave encashed on a disputed formula, a variable component withheld on the view that it was discretionary, a recovery for a notice shortfall set off without explanation, and a full and final statement issued as a single net number with no working. The employee cannot reconcile it, asks, gets a delayed or defensive answer, and escalates. These claims are small individually and expensive collectively, and they are almost entirely avoidable by issuing a line by line settlement statement that shows each head, the basis, the calculation and the deduction, within a stated timeline.
Gratuity, statutory dues and the separation checklist
Gratuity under the Payment of Gratuity Act, 1972 becomes payable on qualifying continuous service, and disputes about it cluster around three things: whether the service qualifies, how continuous service is computed where there were breaks or transfers between group entities, and simple delay in payment. Retrenchment compensation and notice, where the separation is a retrenchment of a workman, is a further head, and the entitlement, its computation and the associated procedural requirements are matters to confirm against the framework applicable to your establishment. Statutory dues on separation, provident fund transfers and withdrawals among them, generate a steady stream of grievances driven less by disagreement than by silence. A separation checklist that names each head, the owner, the computation basis and the date by which it will be settled removes most of this category without anyone needing to be persuaded of anything.
Contractors who were employees in substance
Misclassification is the quiet one, and it surfaces at the worst moment, usually when a long serving consultant is disengaged and claims the benefits of employment for the whole period. The label on the agreement is not decisive. What tends to matter is the substance of the arrangement: who controls how and when the work is done, whether the person is integrated into the organisation, whether they are supervised like staff, whether the engagement is exclusive and continuous, and whether they use the company's systems and hold themselves out as part of it. The remedy is not a better clause. It is to run periodic reviews of long running consultant engagements and either align the substance with the label or convert the arrangement deliberately.
Non-compete and confidentiality claims that rarely go anywhere
Every appointment letter in the country contains a post-employment non-compete, and very few of them do any work. Section 27 of the Indian Contract Act, 1872 renders agreements in restraint of trade void, which is why post-employment non-compete restrictions are generally unenforceable in India. Companies that spend money attempting to enforce them usually buy a short interim skirmish and a poor outcome. That is not an argument for having nothing. It is an argument for putting the effort where it survives: well drafted confidentiality obligations, clear ownership of intellectual property, defined treatment of company data and devices on exit, and non-solicitation drafted with care and modest scope. The practical protection at exit is usually operational rather than contractual: revoke access promptly, take back devices, run a documented handover and confirm in writing what the person has returned and what they have retained.
PoSH and the Discipline of the Inquiry
The Sexual Harassment of Women at Workplace (Prevention, Prohibition and Redressal) Act, 2013 sits apart from the rest of this article, because it is the one area where the process itself is statutory rather than a matter of good practice. The Act requires an Internal Committee at workplaces with the prescribed number of employees, with a presiding officer and the composition the Act specifies, including an external member. It sets a limitation period for bringing a complaint with a discretionary extension available in appropriate cases. It requires the inquiry to be completed within a prescribed period. It requires an annual report. Each of those requirements should be checked against the Act and the rules as they stand, and each of them is regularly missed by companies that would describe themselves as compliant.
The single most important practical point is this: procedural defects in an Internal Committee inquiry are the most common ground on which findings are successfully challenged. Not the credibility assessment, not the conclusion, not the proportionality of the consequence. The route. A committee that was improperly constituted, an inquiry where the respondent was not given the allegations in sufficient particularity, a process where evidence relied on was never put to the person it implicated, a report that records a conclusion without recording the reasoning that produced it, an inquiry that drifted far past the prescribed period without explanation: these are the defects that unwind outcomes, and they unwind outcomes that were often substantively correct.
That has a consequence in-house teams sometimes resist. A rigorous process protects the complainant as much as the company. An outcome that is set aside on procedure does not vindicate anyone. It returns both parties to the beginning, with the additional damage of a year spent litigating, and it is a poor answer to a person who came forward in good faith.
Where inquiries actually fail
The recurring defects are constitution, particularity, participation and reasoning. A committee that is not properly constituted taints everything that follows, and no amount of care in the inquiry cures it. Beyond that: allegations must be communicated with enough specificity to be answered, material relied on must be put to the person it affects, both parties must have a genuine opportunity to be heard, and the report must show the reasoning that connects the evidence to the finding. Confidentiality must be maintained throughout, and retaliation against a complainant is its own exposure quite separate from the underlying complaint. Confirm the composition requirements, the limitation period and its extension, the inquiry timeline and the annual reporting obligation against the Act and rules as they currently stand before you rely on any summary, including this one.
Constitute the committee properly, and re-check it every year
Composition drifts. Members resign, move locations, or become unsuitable because they now report to a potential respondent. Verify the constitution and the external member position annually and at every location that requires its own committee, and record the verification. This is the cheapest single item on the list and the one that most often invalidates everything downstream.
Log the complaint the day it arrives, in whatever form it arrives
Complaints do not always arrive as complaints. They arrive as an email to a manager, a remark in an exit interview, or a conversation in a corridor. Record receipt, the date, the form and who received it. The limitation position and the inquiry clock both depend on that date, and a company that cannot say when it first knew is in a poor position on both.
Give the respondent the allegations with enough particularity to answer
A general statement of misconduct is not an allegation a person can meaningfully respond to. Set out what is said to have happened, when, and in what context, subject to the confidentiality the process requires. This is the step most often compressed under time pressure, and it is a common route to a successful challenge.
Run the inquiry as an inquiry, and minute it as one
Record attendance, what each party said, what documents were produced, and what was put to whom. Put material relied on to the person it implicates. Keep the record contemporaneous rather than reconstructed. If the inquiry cannot be completed within the prescribed period, record why and what is being done about it rather than letting it drift silently.
Write a report that shows reasoning, not just an outcome
A finding recorded without the reasoning behind it is difficult to defend and difficult for either party to accept. Set out the evidence considered, the view taken on the contested points and why, and the basis for the recommendation. Then implement the recommendation through the ordinary employment process, properly documented in its own right.
File the annual report and keep the register current
The annual reporting obligation is straightforward, frequently forgotten, and easy to evidence when it has been done. Keep a register of complaints, dates, outcomes and timelines. It supports the annual report, it answers the audit committee question when it comes, and it is the only way to see whether complaints are clustering around a particular manager, function or location.
Where These Disputes Actually Land
Forum matters more in employment than in most categories, because it drives cost, timeline and the character of the proceeding. The threshold question is usually whether the person falls within the statutory concept of a workman, which turns on the nature of the duties actually performed rather than on designation or salary alone. Someone with a managerial or supervisory role performing predominantly managerial functions is generally outside it, while employees performing skilled, technical, clerical or operational work may fall within it even where the title sounds senior. That classification then routes the dispute.
A structural caveat applies to all of this. The four labour codes consolidate much of the older statutory landscape, and the consolidation position, together with the state rules that give effect to it, should be confirmed for the states in which you operate before you rely on any particular framework. What follows describes the substantive shape of the dispute and where it is heard, which is stable, rather than section references, which are not the right thing to take from an article.
| Type of matter | Typical route | What in-house should plan for |
|---|---|---|
| Termination or retrenchment of a workman | Raised as an industrial dispute and heard by a labour court or industrial tribunal, with conciliation machinery involved before adjudication. | Long timelines, a protective posture, and close attention to whether the procedural requirements for the separation were met. Reinstatement and back wages sit in the range of possible outcomes, which changes how you value settlement. |
| Separation of a non-workman | Generally a civil claim for wrongful termination or dues, framed on the contract of employment. | The contract does most of the work. Notice provisions, the definition of cause, and the settlement mechanics in the appointment letter become the battleground, which is an argument for drafting them for litigation rather than for tidiness. |
| Gratuity claims | Raised before the controlling authority under the Payment of Gratuity Act, 1972, with an appellate route from there. | Disputes usually turn on qualifying continuous service, the computation, or delay. Most are avoidable by settling promptly with a working shown. |
| PoSH findings and process | Challenged through the appeal route provided under the Act and the applicable service rules, and in practice frequently through writ proceedings where a public element exists. | Challenges concentrate on the process, not the conclusion. The inquiry record is the case, and it is fixed long before anyone is briefed. |
| Public sector and statutory bodies | Writ jurisdiction before the High Court is commonly available where the employer is the State or an instrumentality of the State. | Service rules, natural justice and the reasoning recorded in the order carry disproportionate weight. Speaking orders matter far more than they do in a private sector separation. |
| Wage, dues and statutory benefit claims | Statutory authorities under the applicable wage and social security framework, alongside the inspectorate for the establishment. | Individually small, collectively steady. These are the claims a good separation checklist eliminates almost entirely. |
Two practical points follow from the table. First, the workman question should be answered before the separation, not after the notice arrives, because it changes the process the company should run and the risk it is accepting. Second, the outcomes available in the workman route, including reinstatement with back wages, are qualitatively different from a damages claim, and they should be reflected in how a matter is provisioned and how a settlement is valued.
The Cost Asymmetry Nobody Puts in a Board Pack
Employment matters are individually small and therefore chronically under-managed. A single claim rarely threatens the business, so it never reaches the board pack, and because it never reaches the board pack, nobody counts the aggregate. That is a mistake of arithmetic rather than of judgement, and it is worth doing the arithmetic once.
The direct spend, which is the smallest part
External counsel fees across a multi-year proceeding, court and filing costs, travel for hearings at a forum that may not be where your office is, and the settlement itself. This is the number that shows up in the legal budget, which is exactly why it is the only one anyone manages.
The internal time, which nobody bills
The manager who authored the decision, the HR business partner, the payroll analyst reconstructing a computation from three systems, and in-house counsel supervising the whole thing across years. Reconstruction is the expensive part: rebuilding a record that could have been created in fifteen minutes at the time frequently takes days two years later.
The provisioning and reporting drag
Contingent liabilities require a view, and a view requires work. Auditors ask, the audit committee asks, and a portfolio of matters with no reliable status data turns each cycle into a fire drill. A company that cannot say how many employment matters it has, and where they stand, is answering that question from memory in front of people who will remember the answer.
The reputational and recruitment cost
Employment disputes involve people who talk, in a market where hiring is competitive and public commentary is permanent. A PoSH matter mishandled is a governance question, not an HR question, and it will be read as one by regulators, acquirers and candidates alike.
Set against that, the preventive measures are close to free. A written feedback note takes fifteen minutes. A minuted separation conversation takes ten. A line by line settlement statement is a template exercise once and a five minute exercise thereafter. An annual review of Internal Committee composition is one meeting. There are very few places in a legal function where the ratio between prevention cost and avoided cost is this favourable, and it is one of the easier cases to make to a finance leader who is otherwise sceptical of legal spend.
The Documentation Habits That Prevent Litigation
Every recommendation below is small. That is deliberate. Prevention in this category does not come from a policy refresh or a training day. It comes from a handful of habits practised consistently by managers who will never read an employment statute, which means the habits have to be simple enough to survive contact with a busy quarter.
- Write feedback down when it is given, not when it is needed. A short dated note of what was raised, what was expected and by when, sent to the employee, is worth more than any reconstruction. If a manager will not put feedback in writing, that is information about the feedback.
- Minute the separation conversation the same day. Who was present, what was said, what was offered, what the employee said in response. A contemporaneous note by a participant is ordinary business record keeping and it is the single most useful document in the eventual file.
- Never draft the resignation. Where an exit is agreed, give real time to consider it, put the terms in a settlement document that recites what was offered and accepted, and let the employee write their own letter in their own words.
- Run improvement plans as processes, not as pretexts. Specific expectations, a defined support obligation on the manager, interim reviews that actually happen and are minuted, and a genuine possibility of passing.
- Issue a line by line full and final statement within a stated timeline. Every head, the basis, the calculation, every deduction, and a named person who will answer questions about it.
- Keep a separation checklist with owners and dates. Notice, leave encashment, variable pay, gratuity where applicable, provident fund action, recoveries, access revocation, device return and the handover confirmation.
- Review long running consultant engagements annually. Look at substance, not the label, and either align the arrangement or convert it deliberately.
- Investigate what an exit interview surfaces. An unactioned grievance in an exit interview is a document the company created, retained and ignored, and it will be produced against you.
- Retain the file for as long as a claim can realistically be brought, and keep it findable. A record you cannot locate when you need it is functionally the same as a record that was never made.
- Train managers on the four things that matter most: write feedback down, do not promise what policy does not permit, escalate a harassment complaint the day it is received in whatever form it arrives, and route anything unusual to HR before acting.
A note on informal channels
The most damaging documents in modern employment matters are rarely the formal ones. They are the manager messages: the thread that shows the exit decision was taken weeks before the improvement plan was issued, or the comment that gives away a reason the company would rather not have advanced. You will not stop managers using messaging apps, and pretending otherwise wastes the training slot. The workable instruction is narrower and lands better: decisions about a person's employment are recorded through the formal process, and anything you would not want read out in a hearing does not belong in a thread about an employee. Consistency between the informal record and the formal one is the real test, and it is a test the company either passes at the time or fails permanently.
Manage Employment Matters as a Portfolio
The final argument is the one most in-house teams have not yet made internally. Employment matters are managed one at a time, by whoever caught them, and reported as a list of individual items. That is the wrong unit of analysis. Handled individually, each matter is a nuisance with a settlement value. Read together, they are a diagnostic.
Three claims from the same manager in eighteen months is not three claims. It is a finding about that manager, and it is one you can act on cheaply while it is still a pattern in your own data rather than an allegation in someone else's pleading. A cluster of gratuity or full and final disputes from one plant or one payroll process is a process defect, and fixing the process eliminates the category rather than the instance. A rise in claims following a particular restructuring tells you the restructuring was executed differently from how it was designed. None of this is visible in a matter by matter view, because the pattern lives between the matters rather than inside any one of them.
The portfolio view also changes what you can say to a board or audit committee. The useful report is not a list of pending cases with a total provision attached. It is the shape of the thing: how many matters, of what type, arising from which function or location, how they are trending, what the recurring cause is, and what has been changed as a result. That is a report a board can act on, and it is the difference between a legal function that reports exposure and one that reduces it.
Doing that requires two capabilities that are unglamorous but not optional. The first is knowing what has actually been filed against you, which is harder than it sounds when matters are spread across labour courts, civil courts, statutory authorities and High Courts in every state you operate in, each publishing separately and often under a party name that does not match your registered name. CourtMesh addresses that mechanical problem: unified search across the Supreme Court, 25 High Courts, District Courts and Tribunals over a corpus of roughly 310 million cases drawn from official government portals, with a private, organisation-scoped watchlist so that a new matter naming your entity surfaces when it appears in the source registries rather than when the summons reaches a reception desk. The second capability is internal: a register of employment matters with type, function, location, manager, status and stage, maintained as a matter of routine rather than assembled the week before a board meeting.
Two limits, stated plainly
First, nothing here is legal advice, and CourtMesh does not give legal advice. Employment obligations turn on the establishment, the state, the category of employee and the framework applicable to you, and the consolidation position under the labour codes together with the state rules giving effect to it should be confirmed for each state in which you operate before you act on any general description. Second, a search that returns nothing is not proof that nothing exists. Court publication in India is neither immediate nor complete, party names vary between filings, and a matter may be filed in a forum or under a spelling you did not query. Treat a clear screen as an absence of retrieved information, never as a clean record.
One employment claim is an event. The same claim three times from the same manager is a management finding you were already holding and had not yet read.
The cheapest employment litigation is the one you never generate
This category is preventable in a way that almost nothing else on your docket is, because the company authors the facts before it knows there will be a dispute. Write feedback down when it is given. Minute the separation conversation. Never draft the resignation. Show the working on the full and final settlement. Constitute the Internal Committee properly and run the inquiry so that its procedure survives challenge, because procedure is where these findings are lost. Then look at the matters together rather than one at a time, so the manager, the plant or the process behind the pattern becomes visible while it is still cheap to fix. CourtMesh gives you the external half of that picture: unified search across the Supreme Court, 25 High Courts, District Courts and Tribunals from official government portals, with a private watchlist so a new matter naming your organisation reaches you early rather than late.
Explore CourtMesh
