The Indian Founder's Handbook

It covers the full arc: proving there is a real problem worth solving, picking and registering the right legal entity, getting DPIIT recognition under Startup India, staying compliant without drowning in filings, raising money through the routes actually available to Indian founders, getting your first customers, and building the metrics discipline that carries you from ₹0 to ₹10 crore and beyond.

Three things to knowBefore you start

How to read it

If you are pre-idea or pre-revenue, read Parts 1 and 2, then stop. Come back for Part 3 once you are incorporated. If you are already registered and running, skip to Part 4 and audit yourself against the compliance calendar — that is where most early companies are quietly out of order.

The goal of this part is not to "confirm your idea." That framing guarantees you will find confirmation. The goal is to try to kill your idea cheaply, and to keep going only if it survives.

1.1Start with the problem, not the product

Write your idea as a single sentence in this shape, and be ruthless about specificity:

If you cannot fill in the current workaround, that is a red flag, not a green one. A problem with no workaround is usually a problem nobody is actually experiencing. People route around real pain — badly, expensively, with spreadsheets and interns and WhatsApp groups. Find the workaround and you have found the demand.

The three questions your idea must answer

1.2Talk to twenty customers before you build anything

Twenty is not arbitrary. Around the twelfth or fifteenth conversation you stop hearing new things — the patterns repeat, and you know something real. Fewer than ten and you are reading noise.

How to run the conversation

Do not pitch. The moment you describe your solution, the other person switches from informant to critic-or-flatterer, and both are useless. Ask about their past, not your future.

Where to find twenty people in India

For consumer products, go where the behaviour already happens rather than where your friends are.

Reading the signal correctly

1.3Define your ICP and your beachhead

Your Ideal Customer Profile is not "SMEs in India." It is narrow enough that you could list two hundred specific names. A usable ICP names the segment, the size, the geography, the trigger event, and the person who signs.

Founders resist narrowing because it feels like shrinking the opportunity. It does the opposite: a narrow beachhead makes your messaging sharp, your product simple, your first sales fast, and your referrals compounding. You expand outward from a won segment, never from an unwon one.

1.4Test willingness to pay before you write code

The only validated demand is money or a signed commitment. Rank your tests by strength:

1.5Scope the MVP honestly

An MVP is the smallest thing that lets one real customer get the real outcome and pay you. It is not a smaller version of your vision — it is the single most valuable slice, delivered end to end.

1. List every feature you imagine. Be exhaustive.
2. For each, ask: can the first paying customer get their outcome without this? If yes, cut it.
3. Of what remains, ask: can I do this manually for the first ten customers? If yes, do it manually.
4. Build only what survives both cuts.

Typical honest MVP scope for an Indian software startup: one core workflow, manual onboarding, a payment link instead of a billing system, WhatsApp instead of in-app support, a Google Sheet instead of an admin dashboard, and no mobile app. Four to eight weeks, not six months.

TrapsCommon validation mistakes

1.6The validation gate

Do not move to Part 2 until you can tick most of these honestly. This is the single highest-leverage checklist in the handbook.

Changing it later is possible but expensive and disruptive — so choose deliberately, once.

2.1The five options, compared

The short decision rule

• You intend to raise angel or VC money, or issue ESOPs → Private Limited Company. There is no realistic alternative. Investors will not buy into an LLP, and foreign investment rules still do not treat startup LLPs the same way as companies.
• You are bootstrapping a services or consulting business → LLP. Lower compliance, no dividend distribution complexity, partners taxed once.
• You are testing, earning under a few lakh, and want zero overhead → Proprietorship. Convert later. Do not incorporate to feel legitimate.

2.2Registering a Pvt Ltd: the SPICe+ route

Incorporation runs entirely through the MCA portal (mca.gov.in) using the integrated SPICe+ (INC-32) form. One filing now covers name reservation, incorporation, DIN allotment, PAN, TAN, EPFO, ESIC, professional tax where applicable, and a bank account opening request.

1. Get Digital Signature Certificates (DSC) for all proposed directors and subscribers. Issued by licensed Certifying Authorities; requires video KYC. Allow 1–2 days, roughly ₹1,000–₹2,000 each.
2. Reserve the name (SPICe+ Part A). Propose up to two names, with one resubmission allowed. Check availability against the MCA name database and the trademark register at ipindia.gov.in before filing — a name that clears MCA can still be a trademark infringement.
3. File SPICe+ Part B with director details, registered office address, capital structure and share subscription.
4. Attach e-MoA (INC-33) and e-AoA (INC-34). The Articles are where founder protections, share transfer restrictions and board rules live. Do not accept a generic template blindly if you plan to raise money.
5. File AGILE-PRO-S (INC-35) alongside, for GST, EPFO, ESIC, professional tax and bank account.
6. Receive the Certificate of Incorporation with your CIN, plus PAN and TAN. Typical end-to-end timeline: 7–15 working days when documents are clean.

Documents you will need

2.3Registering an LLP

The LLP route uses FiLLiP on the MCA portal: obtain DSCs, reserve the name via RUN-LLP, file FiLLiP with partner details and the registered office, then file the LLP Agreement in Form 3 within 30 days of incorporation. Missing that 30-day window is the single most common LLP mistake and attracts penalties. The LLP Agreement is your real governing document — it sets profit sharing, contribution, decision rights and exit terms, and it deserves genuine drafting attention.

2.4The first 30 days after incorporation

Incorporation is not the finish line; a specific set of obligations starts immediately, and several carry hard deadlines.

Tax holidays, patent and trademark rebates, self-certification relief, government funding schemes and public procurement exemptions all run through it. There is no good reason for an eligible entity not to have it.

3.1Eligibility under the 2026 framework

Deep Tech classification is not automatic — you must apply for it and submit supporting documentation (patents or applications, R&D spend, technical publications, prototype evidence, team credentials). The reward is a materially longer runway of recognition, which matters enormously for space tech, biotech, semiconductors, advanced materials, robotics and similar long-gestation fields.

3.2How to apply

1. Register on the Startup India portal at startupindia.gov.in (applications also route through the National Single Window System at nsws.gov.in).
2. Complete the recognition form: entity details, CIN or LLPIN, PAN, registered address, directors or partners, sector and NIC code, and your website or app link.
3. Upload the incorporation certificate and supporting evidence — pitch deck, product screenshots, patent filings, website, awards, letters from customers or incubators.
4. Write the innovation statement. This is the part that decides your outcome.
5. Self-certify that the entity meets each condition and submit. There is no government fee.
6. Outcome in roughly 2–14 working days. On approval you receive a recognition certificate with a DPIIT recognition number. If DPIIT raises a query, respond within the window given.

3.3Writing the innovation statement

A weak or generic innovation description is by a wide margin the most common cause of rejection. The reviewer is asking one question: is this genuinely doing something new or meaningfully better, or is it an ordinary trading or reselling business?

Structure to follow

• Problem — specific, quantified, with the affected group named
• Solution — what your product actually does, concretely
• Novelty — how it differs from what already exists, named alternatives included
• Evidence — pilots, users, revenue, patents, technical results
• Scalability and impact — why this can grow, and what it creates

Avoid: "revolutionary," "disruptive," "one-stop solution," "leveraging cutting-edge AI/ML/blockchain" with no specifics, and any claim you cannot substantiate. Concrete beats grand every time.

3.4What recognition actually unlocks

3.5The tax holiday: getting the IMB certificate

This is where most founders stop, and it is the most valuable step. DPIIT recognition is largely self-certified; the tax holiday requires a substantive application to the Inter-Ministerial Board.

Conditions

• Entity must be a Private Limited Company or LLP — partnership firms do not qualify for this deduction even if DPIIT-recognised
• Incorporated on or after 1 April 2016 and before 1 April 2030 (the Finance Act, 2025 extended the earlier 2025 sunset by five years)
• Turnover not exceeding ₹100 crore in the relevant financial year per the tax statute — a lower ceiling than the ₹200 crore DPIIT recognition threshold, and the two are frequently confused
• Valid DPIIT recognition, plus IMB certification of eligible business

What to submit

• Board resolution, and the memorandum of association or LLP deed
• Audited balance sheets and profit and loss accounts for the preceding three financial years (or since incorporation), CA-certified
• Income tax returns for the preceding three years or since incorporation
• Shareholding pattern and details of any funding raised
• A detailed innovation write-up, pitch deck and a short product video

3.6Why applications get rejected — and what to do

A rejection is not permanent. Address the stated reason and reapply. Most rejections are drafting failures, not eligibility failures.

This part is a map of every obligation you are likely to have in the first three years. Not all of it applies to you — work through it once, mark what does, and put the dates in a shared calendar.

4.1ROC / MCA compliance (companies)

4.2Income tax

4.3GST

Do you need to register?

• Goods: aggregate turnover above ₹40 lakh (₹20 lakh in special category states)
• Services: aggregate turnover above ₹20 lakh (₹10 lakh in special category states)
• Mandatory regardless of turnover if you make inter-state supplies of goods, sell through an e-commerce operator, are liable under reverse charge, are a non-resident taxable person, or are an e-commerce operator yourself
• Many B2B startups register voluntarily well below the threshold, because business customers want input tax credit and will otherwise treat your invoice as a cost rather than a pass-through

Rates after the September 2025 rationalisation

The slab structure was simplified so that most goods and services now fall into 5% or 18%, with the earlier 12% slab largely absorbed and a higher special rate retained for a small set of demerit and luxury items. Confirm the exact rate for your HSN or SAC code on the CBIC site before you print an invoice — misclassification is expensive to unwind.

E-invoicing applies above prescribed turnover thresholds that have been progressively lowered — check whether you have crossed the current limit. If you are under ₹5 crore, the QRMP scheme reduces your filing burden considerably.

4.4Other registrations you may need

4.5The four labour codes

All four — the Code on Wages 2019, the Industrial Relations Code 2020, the Code on Social Security 2020, and the Occupational Safety, Health and Working Conditions Code 2020 — came into force on 21 November 2025, consolidating 29 central statutes. Central and state rules are still being notified in stages, so existing rules continue to apply during transition. Confirm your state's position before restructuring anything.

What matters most to a startup

• The 50% wage rule. "Wages" is now uniformly defined, and basic pay plus dearness allowance must generally be at least 50% of total remuneration. Salary structures that pushed most of the package into allowances to minimise PF and gratuity no longer work. Expect PF contributions and gratuity liability to rise when you restructure.
• Gig and platform workers are recognised as a distinct category with social security entitlements for the first time in Indian law — directly relevant if you run a marketplace or delivery model.
• Registration thresholds broadly continue: EPF at 20 or more employees, ESI at 10 or more. Contract labour regulation now triggers at 50 rather than 20 contract workers.
• Retrenchment and standing orders thresholds rose to 300 workers, removing a set of obligations for smaller establishments.
• Appointment letters are now expected for all employees — a good discipline regardless.

4.6Data protection: the DPDP Act

The Digital Personal Data Protection Act, 2023, with its rules entering force in phases from late 2025, applies to essentially any startup handling personal data of Indian users — which is essentially every consumer or SaaS startup. Full obligations phase in through 2027, so build for it now rather than retrofitting.

• Collect only what you need, for a stated purpose, with clear and separate consent
• Publish a genuine privacy notice — not a copied template naming another company
• Build a mechanism for users to access, correct and erase their data
• Maintain a record of what you collect, where it is stored, and who processes it
• Take extra care with children's data, where the Act is strict
• Have a breach notification process before you need it

4.7Founder agreements, equity and IP hygiene

These are not statutory filings, which is precisely why they get skipped — and why they cause more startup deaths than any regulator does.

The founders' agreement

Sign it while everyone still likes each other. It should cover: equity split and the reasoning; vesting — typically four years with a one-year cliff, applied to founders too; roles and decision rights; what happens if a founder leaves, voluntarily or otherwise; IP assignment to the company; non-compete and confidentiality; deadlock resolution; and how future equity decisions get made.

IP assignment

By default, work done by a founder before incorporation, or by a freelancer you paid, may not belong to the company. Every founder, employee and contractor should sign an agreement assigning all work-product IP to the company. Investors will diligence this, and a gap here can delay or kill a round.

Protecting what you have built

4.8A compliance rhythm that actually works

Over 2.25 lakh DPIIT-recognised startups, a deep angel network, and a government funding stack that expanded materially in 2025–26.

5.1The funding ladder

5.2Bootstrapping — the underrated default

Most Indian businesses that end up substantial never raise institutional equity. Before you spend six months fundraising, check whether you actually need to. Revenue from customers is the only non-dilutive, non-repayable capital there is, and it comes with the strongest possible validation attached.

Practical levers: take advances or annual prepayments instead of monthly billing; offer a discount for upfront payment; sell services alongside the product to fund the product; keep the team small and senior; use free and open-source tooling plus the credits available through Startup India partner programmes; work from an incubator or coworking space rather than leasing; and negotiate genuine payment terms with vendors.

5.3Government funding: the schemes that matter

Startup India Seed Fund Scheme (SISFS)

The most accessible non-dilutive money for early startups in India, delivered through a network of approved incubators rather than centrally.

• Up to ₹20 lakh as a grant for proof of concept, prototype development and product trials — disbursed in milestone-linked tranches
• Up to ₹50 lakh for market entry, commercialisation and scaling, through convertible debentures, debentures or debt-linked instruments
• Eligibility: DPIIT-recognised; incorporated not more than 2 years before application; not having received more than ₹10 lakh of other monetary government support (some categories excluded); Indian promoters holding at least 51%; a business idea with a clear commercialisation path
• Apply at seedfund.startupindia.gov.in — select up to three incubators; you will be evaluated by the incubator's committee

Credit Guarantee Scheme for Startups (CGSS)

Lets DPIIT-recognised startups borrow without collateral, by having the government guarantee the lender against default. Administered by NCGTC through member institutions — scheduled commercial banks, NBFCs and SEBI-registered venture debt funds.

• Guarantee cover up to ₹20 crore per borrower, following the scheme's expansion
• 85% cover on the amount in default for loans up to ₹10 crore; 75% for the portion above ₹10 crore
• Annual Guarantee Fee reduced to 1% p.a. for startups in the 27 identified Champion Sectors, otherwise 2%
• You do not apply to NCGTC — you approach a member institution for the loan, and the lender applies for guarantee cover

Fund of Funds for Startups (FFS)

A ₹10,000 crore corpus managed by SIDBI which does not invest in startups directly. It commits capital to SEBI-registered Alternative Investment Funds, which then invest downstream — with an obligation to deploy at least twice the committed amount into startups. FFS 1.0 committed to around 145 AIFs, catalysing well over ₹25,000 crore into more than 1,300 startups. Fund of Funds 2.0, with a further ₹10,000 crore approved in the Union Budget 2025-26 and operationalised in 2026, is aimed particularly at deep tech, AI, clean energy and advanced manufacturing.

Debt schemes worth knowing

• Pradhan Mantri MUDRA Yojana — collateral-free loans up to ₹20 lakh, across Shishu (up to ₹50,000), Kishore (up to ₹5 lakh), Tarun (up to ₹10 lakh) and Tarun Plus (₹10–20 lakh). Apply through banks or jansamarth.in.
• Stand-Up India — bank loans from ₹10 lakh to ₹1 crore for SC/ST and women entrepreneurs setting up greenfield enterprises.
• CGTMSE — the broader credit guarantee scheme for micro and small enterprises, useful if you are Udyam-registered but not DPIIT-recognised.
• Sector and ministry schemes — BIRAC (biotech), TDB and DST (technology development), NIDHI-PRAYAS (prototyping grants), MeitY schemes for electronics and software, agriculture and food processing ministry grants.

State startup policies

Almost every state now runs its own policy, and these stack on top of central benefits. Typical offerings: seed grants of ₹5–50 lakh, reimbursement of patent and trademark costs, rent and internet subsidies, marketing and trade-fair support, interest subvention, and preference in state tenders.

Check yours at the Startup India state policies page — the money is real and competition is far lower than for central schemes.

5.4Angels, syndicates and venture capital

Where Indian angels are: Indian Angel Network, Mumbai Angels, Chennai Angels, Lead Angels, and the syndicate platforms — AngelList India, LetsVenture, Tyke — plus operator angels from established startups who often write the most useful cheques. Accelerators such as Y Combinator, Techstars, Antler, 100X.VC and Axilor bring capital, structure and network together. Startup India Investor Connect and the BHASKAR registry are free discovery tools worth using.

What investors actually evaluate

• Team — why you specifically, and why now. Founder-market fit is the first filter and the hardest to fake.
• Market — is the addressable opportunity large enough to return a fund, computed bottom-up rather than as "1% of a huge number"
• Traction — revenue, retention and growth rate. Retention is scrutinised harder than growth at seed stage.
• Unit economics — contribution margin, CAC, payback period. "We'll fix it at scale" is not persuasive in 2026.
• Defensibility — what stops a well-funded competitor copying this in six months
• Clean structure — cap table, compliance, IP assignment, no unresolved founder disputes

5.5Instruments: what you are actually signing

Term sheet clauses that matter more than valuation

• Liquidation preference. 1x non-participating is standard and fair. Participating preference or multiples above 1x mean the investor takes their money back and shares the remainder — this can wipe out founder proceeds in a modest exit.
• Anti-dilution. Broad-based weighted average is normal. Full ratchet is aggressive and should be resisted.
• Board composition and reserved matters. The list of decisions requiring investor consent is where practical control actually sits — read it line by line.
• Founder vesting and lock-in. Expect a reset of your own vesting at the first institutional round.
• Drag-along and tag-along rights. Determine whether you can be forced into a sale, and on what terms.
• ESOP pool creation. Whether the pool is created pre-money or post-money materially changes who is diluted. Pre-money means founders alone bear it.

5.6Cap table discipline

Your cap table is a permanent record of every decision you have made about ownership. Keep it clean from the first day.

• Maintain it in a real spreadsheet or a cap table tool, updated at every event, with fully-diluted percentages visible
• Never give equity for advice, introductions or vague future help — use a small, vesting advisor grant if you must
• Keep founders holding a clear majority through seed; investors want founders motivated for years
• Set up an ESOP pool of roughly 8–15% before your first institutional round
• Avoid a long tail of tiny shareholders — it complicates every future round and every signature you will need

5.7ESOPs

Employee stock options let you hire people you could not otherwise afford, and align them with the outcome. Structure them properly: a pool of 8–15% fully diluted, four-year vesting with a one-year cliff, an exercise window that does not trap departing employees, and a clear written explanation of what the options are actually worth. Indian employees are taxed at exercise on the difference between fair market value and exercise price, with capital gains on eventual sale — recognised startups have some deferral relief available. Communicate the tax position honestly at grant; a badly explained ESOP creates resentment rather than alignment.

Everything else is decoration.

6.1Positioning before promotion

Before spending a rupee on distribution, write your positioning statement. If you cannot, no amount of media spend will compensate.

6.2Pick two channels, not ten

Early-stage startups fail at marketing by spreading thin. Choose two channels, commit for ninety days, measure honestly, then double down or replace.

Choosing well. Match the channel to the purchase. Ticket size above roughly ₹50,000 with a long consideration cycle → founder-led sales and content. High frequency, low ticket, impulse → paid social and marketplaces. Regional or vernacular audiences → WhatsApp, YouTube and regional creators, not English LinkedIn. And go where your customers already gather rather than where you are comfortable.

6.3Your first 100 customers

They will not come from advertising. They come from doing things that do not scale, and that is correct at this stage — the point of the first hundred is learning, not efficiency.

• The list. Write down 200 named prospects who fit your ICP. Not categories — names, companies, contact details.
• Direct outreach. Personalised, short, referencing something specific about them. Twenty a day. Expect 10–20% replies if genuinely personalised, 1–2% if templated.
• Your existing network. Former colleagues, classmates, family businesses. Ask for introductions, not favours.
• Communities where the problem is discussed. Participate for weeks before mentioning your product. Reddit India, sector Slack and WhatsApp groups, trade associations, industry forums.
• Free pilots with a defined end date and a stated price after. Free forever teaches you nothing about willingness to pay.
• Referrals, asked for explicitly. After a customer expresses satisfaction: "Who else do you know with this problem?" The single most underused growth lever in Indian B2B.
• Physical presence. Trade fairs, industrial estates, wholesale markets, local business associations. Still enormously effective for SME-facing businesses, and largely abandoned by tech founders.

6.4Content and search

Content marketing compounds. It is slow — expect four to eight months before meaningful traffic — but the asset keeps returning long after the spend stops, which is the opposite of paid acquisition.

Write for the questions your customers type when they have the problem but have not yet heard of you: "how to reduce GST filing errors," "best inventory method for small pharmacy," "how to register a startup in India." Answer thoroughly and specifically, publish consistently rather than in bursts, and give away real information. Founders who hoard their expertise get no distribution; founders who publish it become the default authority in their niche. One well-researched, genuinely useful piece outperforms twenty thin ones.

6.5Selling to the government

The largest single buyer in India is the government, and DPIIT-recognised startups get structural advantages that most founders never use.

• Register on GeM (gem.gov.in), the Government e-Marketplace. Recognised startups access the Startup Runway storefront.
• Exemptions: prior turnover and prior experience requirements are waived for recognised startups, as is the Earnest Money Deposit — removing the two barriers that historically locked new companies out of tenders.
• Public Procurement Policy for MSEs: a share of central ministry and PSU procurement is reserved for micro and small enterprises, with sub-quotas for SC/ST and women-owned enterprises. Udyam registration is the key.
• Innovation challenges: ministries, PSUs and state governments run problem statements with paid pilots attached. Listed on the Startup India programmes page and far less contested than they should be.

Government sales are slow — six to eighteen months is normal — but contracts are large, sticky, and function as extraordinary references for private buyers afterwards. Start early and treat it as a parallel track, not your only one.

6.6Brand basics that are worth the money

A trademarked name. A logo that works at 32 pixels. One consistent colour and typeface. A website that loads in under three seconds on a 4G connection and explains what you do above the fold. Real photographs of your team and product rather than stock imagery. Beyond that, brand investment before product-market fit is usually procrastination in expensive clothing.

7.1Know your unit economics

7.2Retention is the real product-market fit signal

Growth without retention is a leaking bucket that you are pouring capital into. Track cohorts — group customers by the month they joined, and measure how many are still active and paying at month 1, 3, 6 and 12. If the curve flattens rather than trending to zero, you have found something real. If every cohort decays to nothing, fix that before spending on acquisition.

Look at retention by segment as well as in aggregate. Very often one narrow segment retains beautifully while the average looks mediocre. That segment is your actual business; the rest is noise you should stop serving.

7.3The metrics investors ask for

• Revenue: MRR/ARR for subscriptions, GMV and net revenue for marketplaces, monthly revenue for D2C — with month-on-month growth rate
• Customers: total, new, churned, net revenue retention
• Engagement: DAU/MAU ratio, activation rate, core action frequency
• Efficiency: CAC, payback, burn multiple, gross margin
• Pipeline: for B2B — qualified leads, conversion rate, sales cycle length, average contract value

Track them in one dashboard, review weekly, and send a monthly update to investors and advisors whether or not anyone asks. Founders who report consistently — including the bad months — raise follow-on rounds far more easily than those who go quiet when things get hard.

7.4Hiring your first ten

Your first ten hires set the culture permanently. Hire for what the company needs in the next twelve months, not the next five years — a VP of Sales before you have a repeatable sale is an expensive mistake.

• Hire slowly and deliberately; the cost of a wrong early hire is measured in quarters, not salary
• Prefer generalists early and specialists later
• Use a paid trial project rather than relying on interviews alone
• Write down the role, the outcomes expected and how success will be measured, before you post it
• Be honest about stage, risk and ESOP value — people who join under a rosy misrepresentation leave when reality arrives
• Get the paperwork right from hire one: appointment letters, IP assignment, PF and ESI where applicable, POSH policy at ten employees

7.5Expanding within India, and beyond

India is not one market. Language, purchasing power, distribution and buying behaviour vary enormously across states, and a playbook that works in Bengaluru frequently fails in Indore. Expand city by city or state by state, re-testing the channel each time, rather than declaring national launch. Regional-language product and support is usually a larger unlock than another feature.

For international expansion, the Startup India Go-to-Market Guide and the international engagement programmes — including bridges to markets across Europe, the Middle East and Asia, plus the BRICS and SCO startup forums — provide structured, low-cost entry points. Software and services can go global early; anything involving physical goods, regulation or trust generally cannot.

7.6What changes at scale

Between roughly ₹1 crore and ₹10 crore of annual revenue, three things break and need rebuilding:

8.1Official links directory

8.2Document checklist

8.3Glossary

8.4The 90-day action plan