The minus-one stage, before you incorporate

What the minus-one stage is, and what you leave it with#

The minus-one stage is the 3–9 months between "this result might be a company" and "we incorporated and took money." Its job is to make the company investable before it exists. You leave it with four things in writing: a clear route to the IP, evidence that a specific buyer has a specific problem, a team that has worked together under pressure, and your own decision to commit.

Stage: minus-one · Lens: Core, with deeptech and India notes · Reading time: about 35 minutes

Who this playbook is for#

• PhD students and postdocs whose thesis work has drawn industry interest.
• Faculty and staff scientists at universities, IITs, IISc, CSIR or national labs who want to found, co-found or advise a company.
• Engineers at corporate R&D labs (pharma, semiconductors, energy) who want to spin a technology out.
• Business founders looking for a technology to build around, who will spend this stage with a lab team.

Why this stage decides so much in deeptech#

Software founders can incorporate on day one because their main risk is the market, and they test it by shipping. Lab founders carry three extra risks that are cheap to fix now and expensive later.

1. The IP may not be yours. If the institution, a sponsor or a co-inventor elsewhere owns part of it, investors will not close until that is sorted out. Sorting it out after incorporation often means renegotiating with a TTO that knows you need the deal.
2. A disclosure can destroy the patent. One preprint, thesis defence or recorded talk before filing can permanently remove patent protection in India and Europe (playbook 2.04).
3. Technical and market risk are tangled together. You can spend two years making a better material for a buyer who will never switch. Minus-one is where you find out which performance threshold actually triggers a purchase.

On top of these come team risk (the most common cause of early startup death) and founder risk: will you really leave a secure position? Each can be reduced without a company, a cap table or outside money. That is the point of the stage.

What "done" looks like#

You are ready to incorporate (playbook 4.01) when you can show all of these:

• A written statement from the TTO or IP cell of who owns the invention, and the standard terms on which a founder startup can license or be assigned it.
• A provisional patent filing, or a written decision that you will rely on trade secrets.
• At least 40 structured interviews, with a named beachhead customer segment, the metric they buy on, and the threshold your technology must hit.
• At least one buying signal from outside your network: a paid pilot, an LOI, a joint development proposal, or a data-sharing agreement.
• A founding team that has finished a 4–8 week trial project, with roles and a draft equity split agreed in principle.
• A personal plan: leave, sabbatical or part-time arrangement approved, and 12–18 months of personal runway or a funded fellowship.
• A one-page thesis and an 18-month milestone plan that shows what the first money buys.

If you can't show these after nine months, the right answer may be to license the technology to an existing company, or return to research. Both are fine outcomes (see the decision framework below).

The four risks to retire before you incorporate#

Every hour at minus-one should reduce one of four risks. Score each one red, amber or green every two weeks. You are ready to incorporate when all four are at least amber and IP is green.

Risk 1: IP. Can the company get clean rights to the technology?#

This is the one risk that can kill a deal outright, and the one founders most often leave until last. An investor's lawyer will ask for the chain of title: who invented what, under which funding, and how the rights reach the company.

What to do

• List every inventor, including students and collaborators at other institutions who contributed to the claims, not just to the paper.
• List every funding source behind the work: government grants, sponsored research, consortium projects, gifts and fellowships. Note what IP terms each one carries (playbook 2.01).
• List every material, dataset and software library from outside the lab, with its MTA or licence. A cell line or dataset under a non-commercial licence can block a product.
• Disclose to the TTO and file a provisional before any public disclosure (2.04).
• Ask the TTO in writing for the standard founder-startup terms (2.07, 2.11).

Tests

Risk 2: Market. Does a specific buyer have a problem your technology solves enough better to switch?#

Lab founders tend to describe markets by size ("batteries are a $100 billion market"). Investors want to know who signs the purchase order, what they use today, what it costs them, and what performance level would make them switch.

What to do

• List 10–20 possible applications and score them (playbook 1.07).
• Run structured interviews: about 40 at minus-one, rising to 100 if you go through I- Corps (1.19).
• For your top segment, write down the buying metric, the current best alternative, the threshold you must beat, and who controls the budget.
• Ask for something that costs the buyer: time, data, samples, a paid pilot or an LOI.

Tests

Risk 3: Team. Can these people build a company together?#

Surveys of failed startups regularly list team problems among the top causes. In lab spinouts the most common pattern is a brilliant technical founder, a professor who will stay in their job, and no one who owns customers, money and hiring.

What to do

• Decide what roles the first 18 months need: usually a technical lead, a commercial lead and someone who owns operations and money. One person can hold two roles.
• If no one on the team wants the commercial role, start looking now (playbooks 3.02, 3.04).
• Run a 4–8 week trial project with a real deliverable, such as an I-Corps application, a grant proposal or a pilot plan (3.03).
• Discuss the hard questions before equity: time commitment, salary needs, who would be CEO, what happens if someone leaves (3.07).
• Agree the professor's role explicitly: co-founder, advisor or scientific lead, and how many hours (2.12).

Tests

Risk 4: Founder. Are you really going to do this?#

The risk investors probe hardest in academic founders is commitment. "I will keep my faculty position and advise" is a reasonable choice, but it means someone else must be the full-time CEO.

What to do

• Work out your personal runway: months of living costs you can cover from savings, a partner's income, or a fellowship (see the money section below).
• Learn your institution's rules for leave, sabbatical, part-time or outside work (see the employment section below).
• Talk to family early, especially in India, where leaving a stable academic or government job carries social cost (playbook 1.17).
• Set a date by which you will decide.

Tests

A 26-week minus-one plan#

Six months is enough for a part-time team of two or three to reach a decision. The order matters: protect the IP before you talk to anyone in detail, find the buyer before you build the demo, and agree the team before you agree terms with the institution.

26-week minus-one plan · 4 phases, 4 gates

Each gate is a short meeting of the founding team with a written yes or no. Don't move to the next phase on a no; fix the gap or stop.

Phase 1: Protect and map (weeks 1–4)#

Goal: Nothing you do in the next five months should put the IP at risk.

• Week 1. Build the IP map: inventors, funders, sponsor agreements, MTAs, outside code and data (risk 1 above). Book a meeting with the TTO or IP cell.
• Week 2. Write the invention disclosure. Include planned publications and dates, prior art you know of, and three possible applications.
• Week 3. Meet the TTO. Ask four questions: who owns this, will you file a provisional and when, will you license to a founder startup, and what are your standard terms? Ask for the answers in writing.
• Week 4. Provisional filed, or a date agreed. Freeze public disclosure until it is filed. Set up a shared folder and a dated lab IP log (template below).

Gate 1: Provisional filed or scheduled, and every funding source and MTA listed. If the TTO says a sponsor owns the IP, stop and read playbook 2.01 before doing anything else.

Phase 2: Discover (weeks 5–12)#

Goal: Name the first buyer, the metric they buy on, and the threshold you must hit.

• Week 5. List 10–20 applications and score them (playbook 1.07). Pick three to test. Build a target list of 150 people across those three segments.
• Weeks 6–11. Run 6–8 interviews a week, aiming for 40 in total. Use the interview guide below. Debrief every Friday: update hypotheses, kill the weakest segment by week 9.
• Week 12. Write a one-page beachhead memo: buyer role, current solution and its cost, buying metric, threshold, budget holder, sales cycle, and regulatory path.

Gate 2: A beachhead with a named buyer role and a written threshold. If no segment shows a real pain after 40 interviews, go back to week 5 with new applications, or consider licensing (see the decision framework).

Phase 3: Prove and team up (weeks 13–20)#

Goal: Show that the technology can reach the buyer's threshold, get one buying signal, and find out whether the team works.

• Week 13. Define the trial project for the founding team. The best ones produce something an outsider can judge, such as a fellowship or grant application, a pilot proposal for an interested company, or a technical demonstration at the buyer's threshold.
• Weeks 14–18. Run the technical work needed to show the threshold is reachable, using existing lab resources within your institution's rules. In parallel, return to the five most interested interviewees with a concrete offer: a paid pilot, a sample test, or a data-sharing agreement.
• Week 16. Apply for minus-one money (see the money section below). Many calls open once or twice a year, so check dates in week 1.
• Week 19. Have the hard team conversations: who is full-time and when, who is CEO, salary needs, the professor's role, and what happens if someone leaves (playbooks 3.05–3.07).
• Week 20. Review the trial project honestly. Did deadlines slip? Did anyone do much less than agreed?

Gate 3: At least one costly buying signal from a stranger, and founders agreed on roles and the CEO. If the team trial went badly, fix the team before you go further; it does not get easier after incorporation.

Phase 4: Decide and prepare (weeks 21–26)#

Goal: Make the decision with evidence, and have everything ready so that incorporation and fundraising can start the week after.

• Week 21. Request the TTO's term sheet or option agreement, and model it against seed and Series A dilution (playbooks 2.07, 2.11, 5.19).
• Week 22. Draft the founder term sheet: equity split, vesting, roles, IP assignment (3.05, 3.06, 4.05).
• Week 23. Confirm personal plans: leave approved, fellowship status, runway.
• Week 24. Write the 18-month milestone plan: what the first money buys, technical and commercial milestones, and the team you need to hire (6.02).
• Week 25. Write the decision memo (template below) and send it to two or three outside advisors for a critical read.
• Week 26. Decide: go, license or stop.

Gate 4: A signed decision memo. If go, incorporate (4.01) and start the TTO negotiation (2.10) and fundraising prep (14.02) the same month.

Staying employed while you build#

Most minus-one work happens while founders still hold a faculty post, a PhD place or a research job. The rule almost everywhere is the same: you can spend about one day a week on outside work, you need permission, and you can't run the company day to day while drawing a full salary. Break these rules and you risk your job and the IP.

Faculty#

The one-day-a-week norm. Indian institutes and US universities mostly use the same limit.

IIT Madras changed its rules in 2016 so that faculty could use sabbaticals to work on their own startups (Business Insider). IIT Delhi lets a startup incubated at its TBI pay the institute up to 3% of turnover during incubation (IIT Delhi).

What this means at minus-one

• File for permission early. IISc requires approval even if the venture uses no institute resources. Most institutes have a form listing role, workload and IP used.
• Keep a time log. If the company becomes valuable, people will ask what was done on institute time with institute equipment.
• Separate lab work from company work. Running company experiments in the lab without a sponsored research or facility-use agreement can give the institution a claim on the results.
• Decide your long-term role now. A professor who stays in post can be co-founder and scientific advisor, but someone else must be CEO. Investors will ask who is full-time.

PhD students#

• Read your fellowship terms. Many government fellowships in India (and most US stipends) forbid other paid work, and some forbid holding a company directorship.
• If your advisor is also a co-founder, you have a conflict of interest. US universities often bar a faculty member with a significant stake in a company from being thesis advisor to a student employed by it, or require an independent oversight member (UC San Diego, UC Irvine). Raise this openly with your department.
• Watch the thesis timeline. Defend before you go full-time if you can. A thesis left unfinished is a common regret, and a defence before filing can destroy patent novelty (2.04).
• Holding equity is usually allowed. Working for the company during funded research hours usually isn't.

Postdocs and research staff#

• Postdoc contracts usually follow staff rules: outside work needs approval and must not use project time.
• Grant-funded posts carry the funder's IP and time rules as well as the employer's. Check both.
• Postdocs often have the most freedom to leave. Many full-time founders of lab spinouts are postdocs, with the PI as co-founder or advisor.

Corporate R&D staff#

• Your employment contract almost certainly assigns inventions made in the course of employment, and often anything related to the employer's business. Some contracts claim inventions made for months after you leave.
• Don't use company equipment, data or time. Don't discuss the idea with colleagues who might later be witnesses.
• Consider asking for a written release of the specific invention, or a spinout agreement in which the employer takes equity. Large companies do this more often than founders expect (playbook 1.15).

Money at minus-one#

Minus-one money pays for your time and your first proof, without giving away equity before the company exists. It comes in three forms: personal stipends that replace a salary, small grants for a prototype or proof of concept, and programmes that pay for customer discovery. The best founders stack them: a stipend for the person, a grant for the work.

The main options#

Other routes to check: your own institute's incubator seed or pre-incubation grants; state startup policies (Kerala, Karnataka, Tamil Nadu and Telangana all run them); corporatesponsored prize challenges; and in the UK and EU, Innovate UK ICURe and European Research Council Proof of Concept grants for ERC holders.

How to sequence them#

1. Weeks 1–4: Make a calendar of every call that fits, with deadlines. Many Indian schemes open once or twice a year, and missing a window can cost six months.
2. Weeks 5–12: Use the customer interviews to strengthen applications. Grant panels increasingly ask for evidence of market need, not just technical novelty.
3. Weeks 13–20: Apply. Use one application as the team's trial project (see Phase 3 above).
4. After incorporation: Move to company-stage grants such as SBIR/STTR, BIRAC's larger schemes or the Startup India Seed Fund (playbook 13.01).

What investors think of minus-one grants#

Non-dilutive money at this stage is almost always a plus. It shows outside validation, and it lets you reach a stronger technical milestone before the first priced round. Two cautions: check whether any grant takes royalties or IP rights that would follow the company, and don't let grant milestones pull the company away from what customers need.

Scripts and templates#

Copy these and adapt them. They are written for a researcher talking to industry for the first time.

1. Cold outreach email for customer interviews#

Keep it under 120 words. Ask for advice, not a sale. Name a specific person or paper that connects you, if you can.

Subject: 20 minutes on [their problem area] – IIT/IISc researcher

Hi [Name],

I'm a researcher at [institution] working on [one-line, plain-language
description – e.g. "a coating that cuts
corrosion in offshore steel"]. Before we decide where to take it, I'm talking
to people who deal with
[problem] every day.

I'm not selling anything. I'd value 20 minutes to understand how you handle
[problem] today and what it
costs you.

Would [two specific slots] work? Happy to fit around you.

Thanks,

[Name] · [Title] · [Lab page link]

Warm introductions get far more replies than cold email, so ask every interviewee for two more names. Send cold emails in batches of 20, track the reply rate, and follow up once after five working days.

2. Interview guide (20–30 minutes)#

The rule: ask about their past and present, never about your future product. People are poor at predicting what they'd buy and good at describing what they did.

1. Context (3 min). What is your role? What are you responsible for? How is your success measured?
2. The problem (8 min). Tell me about the last time [problem] came up. What happened? What did it cost: in time, money, scrap, downtime or risk? How often does it happen?
3. Current solution (6 min). What do you use today? Who supplies it? What does it cost? What do you dislike about it? Have you tried alternatives?
4. Buying (5 min). If you wanted to try something new for this, who else would need to agree? Is there a budget for it? How long did the last purchase like this take? What tests would a new supplier have to pass?
5. Threshold (3 min). What would a new solution need to do for you to switch: cheaper by how much, faster by how much, more reliable by how much?
6. Close (2 min). Who else should I talk to? Can I come back to you in a few months with results?

Only at the very end, and only if they ask, describe what you're working on. Note their reaction, and whether they offer anything that costs them: samples, data, a site visit or a pilot.

After each interview, fill in one row of your tracker within an hour:

Date,Name,Company,Role,Segment,Pain (1-5),Current solution,Cost of
problem,Buying metric,Threshold,Budget owner,Costly signal offered,Next step

3. Lab IP log#

Keep this from day one. It makes the TTO conversation, the patent filing and investor diligence much faster.

Date,Who,What was conceived or done,Where recorded
(notebook/page/commit),Funding source,Equipment used,Outside materials or
data (MTA/licence),Disclosed to anyone? (who/NDA),Planned public disclosure

4. The decision memo (week 25)#

Two pages. Write it as if you were advising a friend, not pitching an investor.

# Decision memo: [technology] – [date]

## Recommendation

Go / license / stop – one sentence why.

## The opportunity

Beachhead customer, their problem, what it costs them, current best solution,
our threshold and whether we can hit it.

## Evidence

Number of interviews, by segment. Costly signals received (who, what).
Technical results against the threshold.

## IP position

Who owns what. Filing status. TTO terms offered. Outstanding issues.

## Team

Who is full-time, when. CEO. Gaps and how we fill them. Result of the trial
project.

## First 18 months

Milestones. Money needed and from where (grants, angels, seed). Hires.

## Biggest risks and what would change our mind
Top three risks. The evidence that would make us stop.

## Personal commitments
Each founder: role, start date, runway, employer approvals.

Case files: how lab founders used the minus-one stage#

Two companies from very different systems show the same pattern: the founders found the right lab and professor, built the team before the company, and used institutional facilities under a formal arrangement rather than informally.

Agnikul Cosmos (IIT Madras, India, 2017)#

The starting point. Srinath Ravichandran was a finance professional with a passion for space and no rocket lab. He and Moin SPM drew up a list of IITs doing aerospace work and cold-called or emailed about 70–80 professors (Gadgets 360 via Newsio). One call reached Prof. Satyanarayanan Chakravarthy, head of IIT Madras's National Centre for Combustion Research and Development (Global Indian).

What they did at minus-one

• Found the lab before the company. The professor agreed to back the idea, and became a co-founder (Entrepreneur India).
• Built a complementary team. Moin, who was moving from manufacturing into aerospace, joined as co-founder. The founders then looked for practical rocket experience (YourStory), and through the professor gained support from retired ISRO scientists (Newsio).
• Used the institution formally. The company was registered in Chennai and incubated at IIT Madras, and it is based at the NCCRD (Wikipedia).

What followed. Agnikul started with about ₹3 crore of seed funding (Wikipedia), became the first Indian firm to sign an agreement with ISRO in December 2020 (Entrepreneur India), and later flew a sub-orbital test of its 3D-printed engine (YourStory).

Lessons for minus-one founders

• A business founder without a technology can find one systematically. A list of 70–80 labs is a normal search, not desperation.
• The professor's network (here, retired ISRO engineers) can be worth as much as the lab.
• Incubating at the institution gave access to facilities on clear terms.

Commonwealth Fusion Systems (MIT, US, 2018)#

The starting point. The concept began in MIT class 22.63 in 2012, when Prof. Dennis Whyte challenged graduate students to design a fusion device around a new kind of hightemperature superconducting magnet (Energy Daily). Students and Plasma Science and Fusion Center (PSFC) staff carried the idea forward for years, and many of them helped found the company (SpaceDaily).

What they did at minus-one

• Formed the team in the lab. The founders included PhD graduates of the PSFC, such as CEO Bob Mumgaard and CSO Brandon Sorbom, alongside Whyte as a scientific cofounder (Komo directory).
• Picked a specific technical threshold. The company's case rested on one enabling breakthrough: high-field HTS magnets that make a much smaller tokamak possible.
• Agreed a clear split with the university. The company would build the machine and commercialise it, and the PSFC would focus on research and education (Copernical), under a formal research collaboration.

What followed. CFS spun out in 2018 with $50 million of initial funding (Wikipedia), demonstrated a 20-tesla HTS magnet in 2021 (Komo directory), and has raised more than $2 billion (Energy Daily).

Lessons for minus-one founders

• A long minus-one stage is normal in deeptech. CFS's idea took about six years to become a company.
• Investors fund a single technical threshold that unlocks the market, not a general research programme.
• A formal division of labour with the lab (who does research, who builds product) prevents IP and conflict-of-interest problems later.

The pattern#

Go, license or stop: the decision#

At week 26 you choose between three good outcomes. Starting a company is only right when the IP is clean, a buyer will pay at a threshold you can reach, and a team will work on it full-time. If any answer is no, one of the other outcomes is better for you and for the technology.

Go, license or stop · 3 questions, 4 outcomes

Ask the questions in this order. Each "no" points to a different outcome, so the order tells you what to do next.

Question 1: Can the company get clean IP rights?#

Yes means: the TTO has confirmed ownership in writing, a provisional is filed, no sponsor or MTA blocks commercial use, and standard founder terms are on the table.

If no: don't incorporate yet. A company that can't show title to its core IP can't raise money. Common fixes: negotiate a release from a sponsor, design around a blocking MTA, or license from a co-owning institution. If none works within three months, stop.

Question 2: Will a buyer pay at your threshold?#

Yes means: you have at least one costly signal from a company you didn't know before (paid pilot, LOI, sample exchange, joint development proposal), and the technical results show you can reach the threshold the buyer named.

If no, decide why:

• The technology isn't there yet. Return to research, publish, and apply for translational grants. Come back when the result moves.
• This market doesn't care, but others might. Go back to Phase 2 with new applications.
• No one cares. Stop, and count the time as well spent: you avoided years on a company without customers.

Question 3: Is a full-time team committed?#

Yes means: at least one founder (ideally two) will work full-time within three months of funding, someone owns the commercial side, the CEO is agreed, and leave or exit from current jobs is approved.

If no, license instead. You have something valuable (clean IP and a buyer) but no one to build the company. Licensing it to an established company gets the technology to market and earns royalties for the institution and inventors. Licences to existing companies are a normal outcome of university technology transfer, and the TTO will run the process (playbook 2.03).

Alternatives before you license:

• Recruit a commercial CEO through incubators, fellowships or entrepreneur-inresidence programmes (playbook 3.04).
• Join a founder-matching programme that pairs scientists with business founders. Keep an option. Ask the TTO for a 6–12 month option while you find a CEO, rather than licensing to a third party straight away.

Go: what happens in the next 30 days#

1. Incorporate (playbook 4.01), and sign the founder agreement and IP assignment (3.07, 4.05).
2. Open the licence negotiation with the TTO, using the term sheet from week 21 (2.10).
3. Apply for company-stage grants (13.01).
4. Start the seed fundraising preparation (14.02), using the decision memo as the first draft of your narrative.

Mistakes and how to recover#

Most minus-one mistakes can be recovered from if you catch them early. The table lists the ones that come up most, roughly in order of how much damage they do.

Warning signs to watch for at each gate#

• Gate 1: "The TTO hasn't replied" for more than three weeks. Escalate politely through your head of department or incubator.
• Gate 2: Interviewees are polite but no one offers anything that costs them. You are describing the technology, not their problem.
• Gate 3: One founder keeps missing trial project deadlines. This is information, not bad luck.
• Gate 4: No one wants to write the decision memo. Often that means no one wants to make the decision; talk about it directly.

Existing playbooks to read, and sources#

Read next#

• Steve Blank and Bob Dorf, The Startup Owner's Manual . The customer development method behind I-Corps. Read the chapters on customer discovery before Phase 2.
• NSF I-Corps curriculum. The seven-week structure and 100-interview standard (NSF). If you are not eligible, follow it on your own.
• Rob Fitzpatrick, The Mom Test . A short book on asking interview questions that people can't lie to you about. Use it with the interview guide above.
• TenU USIT Guide. The UK reference for spinout equity and licence terms, used as a benchmark well beyond the UK (Cambridge Enterprise).
• Your institution's IP policy and startup or faculty-entrepreneurship policy. Read both in full during week 1.

Related playbooks in this library#

1.07 Tech-push vs market-pull · 1.19 Customer discovery from the lab · 2.01 Who owns your invention · 2.03 Working with the TTO · 2.04 Patent or publish first · 2.07 Licence anatomy · 2.11 Spinout equity norms · 3.03 Co-founder dating · 3.04 Bringing in a business CEO · 3.05

Founder equity splits · 4.01 When to incorporate · 13.01 Non-dilutive capital map · 14.02 Fundraising readiness

Sources#

• NSF I-Corps National Teams applicants
• Angels Partners: NSF I-Corps guide
• Cambridge Enterprise: the USIT Guide

• IIT Delhi: outside professional engagement IISc Faculty Entrepreneurship policy v3.0 IIT Kanpur Faculty Entrepreneurship Policy IIT Palakkad Innovation and Startup Policy

• ThePrint: AICTE startup policy
• Business Insider: IIT Madras sabbaticals for startups
• University of Pittsburgh: outside activities
• Weill Cornell start-up guidelines 2026
• UC San Diego: outside professional activities form
• UC Irvine startup guide FAQs
• India Science & Technology: NIDHI-EIR
• NIDHI-EIR guidelines
• Ahmedabad University: NIDHI-EIR
• NGOBox: NIDHI-EIR call
• IndiaFilings: NIDHI-EIR and PRAYAS
• BIRAC BIG Guidelines v8
• Startup Gujarat: BIG scheme
• Activate Fellowship US Cohort 2027 call
• FundsforNGOs: Activate Fellowship 2026 call
• Funds for Individuals: Activate Fellowship
• Wikipedia: AgniKul Cosmos
• Newsio (Gadgets 360 interview): Agnikul
• Global Indian: Agnikul Cosmos
• YourStory: Agnikul and ISRO (2019)
• YourStory: Agnikul vision (2025)
• Entrepreneur India: Agnikul Series B
• Wikipedia: Commonwealth Fusion Systems
• Energy Daily: CFS power plant plans
• SpaceDaily: MIT expands collaboration with CFS
• Copernical: PSFC and CFS roles
• Komo directory: Commonwealth Fusion Systems