Company Registration in India: the Route, the Timeline, the Alternative

Every incorporation guide quotes the MCA's processing time and stops there. The certificate of incorporation genuinely can arrive in two to four weeks. But a certificate is not a company you can operate. You cannot pay a salary with it, invoice with it, or receive your own capital into it. The distance between certificate and functioning employer is where the six months goes, and nobody puts that in the brochure.

⏰ Why the bank account is the bottleneck

For a foreign owned subsidiary, the bank account is where timelines die. Indian banks apply enhanced diligence to foreign shareholding: apostilled documents, notarised board resolutions, KYC on every director in person or through consular routes. Six to ten weeks is normal. And until the account exists, nothing downstream can start, because the capital cannot land and the FC-GPR clock cannot even begin.

⚠️ The sequencing trap

PF and ESI registration need the bank account. GST wants the registered office lease. The lease wants the bank account to pay the deposit. The bank wants the certificate plus the directors in a branch. Founders who try to parallelise everything discover the dependencies one rejection email at a time. A good company secretary sequences it properly and still lands around month four at best.

The reason this matters for hiring: candidates do not wait six months. If you are incorporating because you want to employ people in India, the standard play is to run the first hires through an employer of record while the entity is in flight, then migrate them once your PF code is live. We cover that hand off in the entity vs EOR chapter, and the mechanics live on our EOR services in India page.

India offers five structures a foreign company might plausibly use. In practice the decision collapses to one default and four special cases.

📇 Why private limited is the default

The private limited company takes 100 percent foreign ownership under the automatic route in most sectors, needs no prior government approval, supports ESOPs your Indian hires will actually value, and is the structure every investor, bank and acquirer in India already understands. Corporate tax sits at roughly 25 percent effective for most new companies. When people say set up an entity in India, this is the entity they mean.

🚧 The two director rule

A private limited company needs a minimum of two directors, and at least one must be a resident of India, meaning 182 days or more in the country in the preceding year. Foreign founders solve this with a trusted local hire, a professional nominee director service, or a co-founder relocating. Budget for the nominee if you have no local principal: it is a real ongoing cost and a real governance decision, not a formality, because that person carries statutory liability.

🤔 What about just using an LLP?

LLPs look attractive for their lighter compliance, and for a services practice with two partners they are genuinely fine. For anything that might raise venture money, grant options, or be acquired, the LLP becomes a conversion project at exactly the moment you can least afford one. If there is any version of your future involving a term sheet, start with the private limited company.

For the hiring specific comparison, including what each structure means for payroll registrations, the EOR vs entity breakdown walks the same ground from the employment angle.

Incorporation costs are almost a distraction: government fees, stamp duty, digital signatures and professional fees for a clean private limited setup land between 1,500 and 4,000 dollars depending on the state and how much hand holding you buy. The number that matters is the annual run rate of simply existing, before you hire anyone.

💰 The honest annual total

Stack it and a dormant but compliant subsidiary costs 10,000 to 20,000 dollars a year before a single salary. Founders consistently underestimate this because each line looks small. The transfer pricing study is the one that surprises everyone: the moment your Indian entity bills the parent for services, which is the standard model for a captive team, you are into documented arm's length pricing with an annual study and a markup the tax office will scrutinise.

💸 Compare the run rates honestly

An EOR at 149 dollars per employee per month costs 1,788 dollars per person per year, all statutory operations included. Ten employees cost 1,490 dollars a month in fees against an entity's 10,000 to 20,000 dollar overhead plus your own hours in board meetings and audit queries. By raw cost the crossover sits around 10 to 15 heads; price in founder attention and it drifts to 20 to 30. Under the line, the entity is a vanity purchase. Past it, the entity genuinely wins, and the migration is a payroll transfer, not a re-hire.

The per employee mechanics behind that comparison, PF at 12 percent, ESI at 3.25 percent, gratuity accruing at 4.81 percent of Basic plus DA, are itemised with a calculator on our EOR services in India page.

Incorporation is a birth certificate, and newborns need feeding. Here is the recurring load for a small operating subsidiary with employees, laid out by frequency rather than statute, because frequency is how you will experience it.

⚠️ The penalties are asymmetric

Miss a GST return and the late fee is capped and boring. Miss PF and the 12 percent annual interest under Section 7Q plus damages up to 25 percent under Section 14B compound quietly. Miss an FC-GPR filing and you are into RBI compounding proceedings, which are slow, discretionary and exactly the kind of thing that surfaces in acquisition diligence three years later. The rule of thumb: money that touches employees or the RBI gets paid first and filed on time, always.

🧾 Director liability is personal

Several of these obligations attach to directors personally, not just the company. TDS deducted but not deposited is the sharpest one: it can trigger prosecution provisions, and being overseas does not make it abstract if you also want to keep visiting India. This is half the argument for professional company secretarial support, and the whole argument against running an Indian entity from a spreadsheet and goodwill.

None of this argues against having an entity. It argues for having one deliberately, at the right headcount, with the machinery budgeted, rather than backing into one because a template guide made incorporation sound like a weekend project.

The entity or not decision reduces to one chart. On one axis, the entity's fixed overhead: 10,000 to 20,000 dollars a year plus the four to six months of setup and the ongoing tax on your attention. On the other, an EOR's linear fee: 149 dollars per employee per month, 129 past 20 employees.

🤔 Why practice lags the spreadsheet

On paper the lines cross near a dozen heads. In practice most companies migrate later, around 20 to 30, and they are not being irrational. The spreadsheet omits the migration project itself, the risk of running payroll badly during the learning year, the hiring freeze you do not want while the PF code is pending, and the value of your own hours. A founder mid Series A has better uses for a week a month than audit queries. The entity wins when the India operation has an owner on the ground, a head of India or a finance lead, who absorbs that machinery as their day job.

🔁 The migration is the easy part

The pleasant surprise: moving people from an EOR onto your new entity is a payroll transfer, not a re-hire. Same person, new employer of record, PF balance rides along on the same UAN, gratuity continuity negotiated into the transfer letters. A competent EOR treats graduation to your own entity as a designed exit, which is why exit fees are the single most revealing line on any EOR rate card. Ours is zero, and that is a deliberate statement about whose interests the structure serves.

If you want this chapter as a full standalone treatment with the sensitivity tables, it exists: EOR vs entity in India is the piece we send investors who ask why their portfolio company should not incorporate on day one.

Here is the sequencing problem in one line: the entity takes four to six months, and the two engineers you want to hire have offers expiring in two weeks. Every quarter, companies lose their first choice India hires to this gap, then blame the market.

🚀 The bridge play

The standard structure now is to run the first wave of hires through an employer of record while the entity paperwork grinds. The people are employed on the EOR's registered Indian entity from day one, with PF, ESI, TDS and gratuity running properly, working entirely on your roadmap. When your own PF code goes live, you migrate them across. No lost candidates, no compliance improvisation in the interim, no pressure to rush the bank account stage, which is the stage banks refuse to be rushed on.

The bridge also de-risks the entity decision itself. A meaningful fraction of companies that start an India pod discover within two quarters that they want twice the headcount they planned, or half. Committing to entity overhead before the experiment has data is backwards. Run the pod on an India native EOR first, and let the entity decision follow the evidence at the 20 to 30 head crossover.

🚧 What to line up during the bridge

Use the entity's gestation months deliberately. Register the trademarks. Choose the payroll software your entity will run and have the EOR mirror its salary structures so migration is a data export. Decide the ESOP pool and get the plan drafted, because Indian hires increasingly negotiate options, and granting them needs the entity live. And interview company secretaries the way you would interview an engineer: ask each one to walk you through last year's worst filing miss and what changed after it.

⏰ The handover checklist

When the entity is ready, the migration runs in one payroll cycle: transfer letters signed, PF transferred on existing UANs, gratuity continuity recorded, insurance novated or re-issued, and the EOR relationship closed without exit fees. The employees experience a change of letterhead. If any vendor describes this handover as complicated, they are describing their own lock in, not the law.

That is the full loop: bridge on an EOR, build the entity deliberately, migrate on your own timeline. It is the least dramatic way to enter the Indian market, which is exactly the compliment infrastructure should aim for.

Nobody incorporates planning the funeral, but the exit terms belong in the entry decision. Closing an Indian private limited company properly takes longer than opening one, and the fast options only exist for companies that stayed clean throughout.

🧾 The three exits

Strike off under Section 248 is the clean path: available once the company has been inactive long enough and owes nothing, typically wrapping in six to twelve months after you stop operating. Voluntary liquidation under the insolvency code is the thorough path for companies with real assets and history, running one to two years with a liquidator appointed. And dormancy is the parking option: keep the entity alive at minimal compliance cost while you decide, filing skeleton returns. What does not exist is abandonment. An ignored entity keeps accruing filing defaults and director penalties, and India does not forget corporate registrations.

⚠️ Employees make exits humane or ugly

If the entity employed people, the wind down inherits every employment obligation at once: notice periods honoured or paid, gratuity for anyone past five years computed at 4.81 percent accruals you hopefully funded, full and final settlements inside the 48 hour window, PF accounts closed or transferred properly, and relieving letters that do not sabotage anyone's next job. Companies that shortcut this stage generate the labour disputes that outlive the company itself.

🤔 The asymmetry worth noticing

Notice the shape: an EOR engagement ends with 30 days notice and zero exit fees; an entity ends with a year of liquidation paperwork. Neither fact decides the structure question alone, but the asymmetry belongs in the model. Optionality has a price, and for companies still validating their India operation, the exit terms are a real part of what the EOR fee buys. It is also why we publish the crossover math honestly on the EOR services page instead of pretending every company should stay on an EOR forever. Past 20 to 30 heads, they should not.

The decision hygiene is simple. Before signing the incorporation papers, write down the headcount and timeline at which you would wind down, and what that would cost. If the answer does not scare you, incorporate with confidence. If it does, you have just discovered you wanted a bridge, not a building.

Incorporation is federal, so the certificate looks identical whether your registered office sits in Bengaluru or Bhubaneswar. What changes by state is everything around it: professional tax rates, Shops and Establishments rules, stamp duty, the speed of GST verification, state incentives, and, most importantly, where your people actually are.

📇 Registered office vs where people work

Your registered office fixes which ROC you file with and where official mail lands. It does not constrain where employees sit. A company registered in Karnataka with staff in Pune and Gurgaon simply carries Shops and Establishments registrations and professional tax in each working state. That multi state surface is normal now; it just needs to be run deliberately rather than discovered during an inspection.

🤔 Should incentives drive the choice?

Almost never for a services or product subsidiary. State incentive packages are built for manufacturing scale and headcount commitments a 15 person pod will not meet. Choose the state where your first ten hires want to live, register there, and let the incentives be a pleasant accident rather than a strategy. The talent map matters ten times more than the stamp duty table.

If the honest answer is that your first ten hires are scattered across four states because you hired remote first, that is one more argument for starting on an EOR structure, which already carries the multi state registrations, and incorporating once the map has settled.

Timeline slippage in Indian incorporations is almost never the MCA. It is documents arriving late, notarised wrong, or apostilled in the wrong sequence. Here is the checklist that decides your calendar, split by who has to produce it.

⚠️ The apostille cascade

Every foreign document needs apostille or consular legalisation, and validity windows mean sequencing matters: get documents attested too early and they expire before filing; too late and the whole calendar shifts right. Countries not party to the Hague convention route through consular legalisation, which adds weeks. Build the document pack first, then start the clock, not the other way round.

🧾 Object clauses deserve ten minutes

The MoA's object clauses define what the company may do. Template drafts copied from a services boilerplate have a way of omitting the thing you actually plan to do, and amending objects later is a shareholder resolution plus filings. Read the draft. It is two pages. Ten minutes here saves a filing cycle later.

None of this is difficult; all of it is sequential. Which is the recurring theme of this entire guide: India rewards companies that respect the sequence and quietly punishes the ones who treat process as an afterthought. Get the pack right, and month four payroll is realistic. Improvise, and you will meet month seven.

Incorporation gets the certificate. Banking makes the company real, and this is the stage where the polished timeline meets Indian banking practice. Two tracks run in parallel: getting an account open, and keeping the foreign investment paperwork inside its deadlines.

📇 The account hunt

A private limited company needs a current account before it can receive its own share capital. On paper any scheduled bank opens one in days. In practice a foreign owned subsidiary triggers enhanced due diligence: apostilled documents for every foreign director, beneficial ownership declarations, sometimes an in-person visit from a director who lives nine time zones away. Two to six weeks is the honest range, and the variance depends mostly on which relationship manager picks up your file. Foreign founders consistently report smoother runs with banks that have dedicated startup or inbound investment desks. Pick the bank before you incorporate, not after, and get the document list in writing.

⏰ The FC-GPR clock

Your share capital arriving from abroad is foreign direct investment, and the Reserve Bank of India wants it reported. The sequence: the money lands, the bank issues a foreign inward remittance certificate, the company allots shares within 60 days of receipt, and Form FC-GPR is filed on the RBI's FIRMS portal within 30 days of allotment. Miss the windows and you enter compounding territory, a formal RBI process to regularise the lapse, with fees that scale on the amount and the delay. It is fixable and common, and it is also entirely avoidable with a calendar entry. Your company secretary should own these dates the way payroll owns the 15th.

💸 The ongoing remittance rhythm

After the capital round trip, money movement settles into a rhythm with its own rules. Funding the subsidiary can run as further equity, each round repeating the FC-GPR cycle, or as payments against an intercompany services agreement, the usual pattern for a development centre, where the parent pays a cost plus margin fee and the pricing needs to survive transfer pricing scrutiny. Sending money out is harder than sending it in: dividends carry withholding tax, and royalty or fee structures need documentation drawn before the money moves, not after. The practical takeaway for a founder is one line: every rupee crossing the border needs a named legal basis, in writing, in advance.

All of this is the price of running your own entity, and at real scale it is worth paying. Below the crossover, it is another line in the case for starting on an employer of record, where the only cross border payment is one monthly invoice, at the RBI reference rate with zero spread, and the FC-GPR clock belongs to somebody else.

🤔 Do I really need a resident director?

Yes, at least one director resident in India for 182 or more days in the preceding year, with no workaround. Options in order of preference: a senior local hire you trust, a co-founder who relocates, or a professional nominee service with proper indemnities. The nominee route works but choose carefully: this person signs filings and carries statutory liability, so a cut price nominee is a false economy.

🤔 Can my Indian entity hold IP?

It can, but for most foreign parents it should not. The standard model keeps IP at the parent and has the Indian subsidiary provide development services under an intercompany agreement at a cost plus markup. That triggers the transfer pricing study from the cost chapter but keeps your IP consolidated where your investors expect it. Moving IP out of India later is painful and taxed; not putting it there is one decision, made once.

🤔 Can I grant ESOPs to Indian employees before the entity exists?

You can grant options in the parent company to India based employees regardless of local structure, and if your hires are employed via an EOR that is the standard route. Once your Indian subsidiary exists you can also run a local plan or continue parent level grants; each has tax quirks for the employee at exercise. What matters in hiring: Indian senior candidates in 2026 negotiate equity, so have an answer before the offer stage, not during it.

🤔 Does my subsidiary charge GST on services to the parent?

Services exported to the foreign parent are zero rated for GST provided payment arrives in convertible foreign exchange and the paperwork, LUT filing, invoicing discipline, matches. Zero rated is not exempt: you still register, file and claim refunds on input credits, which is why the GST line appears in the annual cost table even for pure captives.

🤔 Is there a way to test India without any of this?

That is the entire premise of the employer of record model, and the honest answer to this question is what this page's parent service exists for. Employ the first hires on an India native EOR, learn what your India operation actually is, and let incorporation be a decision you make with data at the crossover, not a leap you make from a blog post. When you do incorporate, the migration is one payroll cycle.

🤔 Can we convert the EOR team onto our entity later?

Yes, and this is the standard playbook rather than the exception. The employees resign from the employer of record and join your new entity the next working day, with offer letters mirroring their terms, PF accounts moving via UAN transfer, and continuity of service honoured for gratuity. Done properly it is a paperwork weekend, not a re-hiring exercise, and nobody's take home changes. The thing to check before you start: that your EOR has no exit fees and no lock in, because a provider that charges you to leave has priced your future entity into their margin. We put it in writing on day one: no setup fees, no exit fees, leave whenever the crossover math says to.

⭐ The position, plainly. An Indian private limited company is excellent infrastructure at the right time and a 15,000 dollar a year distraction before it. The right time is roughly 20 to 30 employees, or the arrival of an on the ground leader who owns the machinery, whichever comes first. Before that point, hire through an EOR, bank the six months and the overhead, and spend both on product. After it, incorporate without nostalgia and migrate your people in one cycle.

Budget honestly: 1,500 to 4,000 dollars to incorporate, 10,000 to 20,000 a year to stay clean, four to six months to first payroll, and one to two years to leave. Any advisor quoting materially rosier numbers is selling the certificate, not the company.

And the disclosure that doubles as the offer. I run Versatile, an India native EOR. We employ your India team on our own registered entity at 149 dollars per employee per month, first month free, no setup or exit fees, payroll settled at the RBI reference rate with zero FX spread, PF, ESI, TDS and gratuity run in house across 28 states. Companies use us as the bridge while their entity is in flight, and as the permanent structure below the crossover, and we hand teams over to their own entities without friction when the time comes, because that is what the model is for. If India is on your map this year, use the form on this page and I will reply personally.

And if you want the honest gut check before any of it, run one number: what your first year of entity overhead buys in actual hires on the EOR model. For most teams under fifteen people the answer settles the question in a minute, and the entity conversation moves to exactly where it belongs, the quarter your India headcount makes it cheap.