Where we work
Every market here starts a different first conversation.
We work from India, in Europe and across Asia and Oceania. These are not variations on one another: the way each state funds an early company differs structurally, and that difference decides what a founder should do in their first year. This page is the comparison; each market then has a page of its own.
Why these markets and not others.
We follow the funding architecture, not the market size.
A seed company’s first year is shaped more by how the state funds early work than by how large the home market is. Where those rules force a real decision — scope the project up or down, find the customer first or the money first — there is work for us.
One market at a time, read at source.
Each page here was built by reading the agency rules and, where it mattered, the statute itself. Not by summarising what other advisers say about them. Where two official sources disagree, or where we could not read one at all, the page says so instead of picking the tidier number.
We say plainly what we are not.
We work remotely from Hyderabad, in English. There is no office and no local entity in any of these markets, and every page says so rather than leaving it to be assumed.
One question, a different answer in every market.
What the state offers an early company, what it tests before giving it, and what shape the money takes. The spread first, then every market in full.
- The instrument is not the same kind of thing twice. One state invests through intermediaries, another awards a grant on a project, a third writes a credit into statute that nobody has to award, and a fourth pays its contribution to a research institution rather than to the company at all.
- What is examined ranges from a customer to a calendar. A committed pilot buyer in one market, an accounting ratio in another, a timesheet counted every month in a third, a certifying body’s separate opinion in a fourth — and in several of them a deadline or a company age that has nothing to do with the research.
- The widest difference is whether money reaches a company with no profits. Several of these markets turn their research incentive into cash before there is any tax to offset. One gives a subsidised research partner instead of cash. One lends and invests rather than granting. And one bars the relief from creating even a loss to carry forward.
| Market | The main early instrument | What it tests | The shape of it |
|---|---|---|---|
| India | A fund of funds, invested through registered funds rather than into companies. PIB, 13 Apr 2026 | Recognition, with a twenty-year term for deep tech against a turnover cap. G.S.R. 108(E), Feb 2026 | Capital through intermediaries; the newer scheme lends and invests, and does not grant. PIB, 29 Jul 2026 |
| Europe — The NordicsThree EEA states where an agency decides. The money is a grant or an administered credit, it is applied for, and what it tests differs sharply between them. | |||
| Finland | Sprint, a continuous grant call covering three-quarters of the project. Sprint call | Equity already in the company, and a committed team on the ground. Sprint call | One research grant, after three others were closed in 2025. Business Finland, Sep 2025 |
| Norway | Innovasjonskontrakter, whose minimum is a floor on the grant, not the project. Innovasjon Norge, innovasjonskontrakter | A committed pilot customer, paying a fifth of total project costs. Innovasjon Norge, innovasjonskontrakter | A grant and a statutory credit sharing one state-aid ceiling between them. GBER, Reg. (EU) 651/2014, consolidated |
| Sweden | A research deduction taken straight off employer social contributions. Lag (2023:747), Riksdagen | Each named person, each calendar month: half their working time on research. Lag (2023:747), Riksdagen | A monthly payroll test rather than an application, beside agency grants. Vinnova, Innovativa Startups 2026 |
| Europe — Western EuropeFour markets where the main instrument is written into statute rather than awarded by a panel. Two are outside the EU, and in three of the four a second authority — a devolved agency, a certifying body, a canton — changes the answer. | |||
| United Kingdom | A statutory research credit nobody awards — and no any-sector grant open. HMRC, merged scheme and ERIS | An expenditure ratio, and a notification deadline that ignores the research. HMRC, CIRD183000 | A taxable credit, or an untaxed payable one, capped against local payroll. HMRC, merged scheme and ERIS |
| Germany | The Forschungszulage, a quarter of eligible research costs and more for a smaller company. § 4 FZulG | A separate body certifies that the project is research, before any money is calculated. § 6 FZulG | A cash-settled allowance, with costs another grant pays for excluded from the base. § 7 FZulG |
| France | A research credit at 30%, held together with a young-company status. service-public.gouv.fr, CIR | An annual accounting ratio — research at a fifth of charges — or one of three other routes. BOFiP, BOI-BIC-CHAMP-80-20-20-10 (15 Apr 2026) | Repaid in cash while the company is young, and held against tax otherwise. service-public.gouv.fr, CIR |
| Switzerland | Innosuisse projects, whose federal contribution is paid to the research partner. Innosuisse FAQ | For the one route that pays the company: that it is not yet selling. Innosuisse, start-up innovation projects | Co-funding the company pays into, and a deduction that cannot create a loss. StHG (SR 642.14), Arts. 25a–25b |
| AsiaThree Asian markets beyond our own. The instruments here are timed, intermediated or still waiting on a decree, and none of them behaves like a grant. | |||
| Singapore | A first-time-founder grant, matched one for one by the founders’ own capital. Startup SG Founder, 20 Aug 2026 | Age, not merit: under six months old at the first approach to a mentor partner. Startup SG Founder eligibility, 20 Aug 2026 | A grant, beside an exemption whose years elapse whether or not there is income. IRAS, 7 Aug 2026 |
| Vietnam | Two years free of corporate income tax, then four at half rate. Resolution 198/2025/QH15, 17 May 2025 | A written definition rather than a threshold: intellectual property, technology or a new model. Resolution 198/2025/QH15, 17 May 2025 | Statutory relief, with two headline reliefs still deferring to a decree that must follow. Resolution 198/2025/QH15, 17 May 2025 |
| Japan | A deep tech subsidy at the seed research phase, capped across a whole project. NEDO, DTSU 2026 call notice (Japanese), 26 Jun 2026 | A letter of intent from an operating company, at the point of applying. NEDO, DTSU 2026 call notice (Japanese), 26 Jun 2026 | Up to two thirds, so the company must already be able to raise the rest. NEDO, DTSU 2026 call notice (Japanese), 26 Jun 2026 |
| OceaniaTwo markets that pay for research through the tax system rather than through an agency, and test it in opposite directions. | |||
| Australia | A research offset at the company tax rate plus 18.5 points, refundable while small. ATO, offset rates, 13 May 2026 | The company itself, by self-assessment against a statutory test. business.gov.au, R&DTI eligibility | An entitlement claimed after the year ends, and published by law once claimed. business.gov.au, R&DTI overview |
| New Zealand | A 15% credit above a spending floor, and separately a refund of research losses. Inland Revenue, R&D tax incentive | Approval of the activities before any claim, and a labour-cost share for the refund. Inland Revenue, wage intensity criteria | Cash that comes back: the refund is repaid on profit, and blocks imputation credits. Inland Revenue, R&D loss tax credit |
These are compressed summaries with their sources attached, and every figure behind them is stated in full, with its source and its date, on that market’s own page. Markets appear in the order the list below groups them, so a row can be compared with its neighbours before it is compared with the far side of the world. The rules change. Your accountant and your lawyer advise on them; we plan around what they confirm.
What carries between markets, and what does not.
The commercial question is the same everywhere on this page: who buys, at what price, through a channel that collects on time. That work transfers intact. The funding architecture transfers not at all — a plan built around Indian recognition thresholds does not survive contact with a Norwegian grant floor, and a plan built on an Australian offset claimed after the year ends inverts completely in New Zealand, where the activities have to be approved first.
- It runs in both directions. Indian companies whose first serious buyers are abroad, and companies elsewhere whose first large market is not at home.
- Some corridors have no public route at all. Japan’s co-funded international research call lists its partner countries, and India is not among them — so a Japanese company pairing with an Indian one has no public money for it through that instrument. We would rather say that than leave it to be assumed. NEDO, International Joint R&D (English), 9 Sep 2026
- The public route exists before the commercial one. Business Finland maintains a presence in India, so a Finnish company testing the market has an institutional way in as well as a commercial one. Business Finland in India
- One advisor sees both ends. The same person works on the Indian side and the Nordic side, which is why the comparison above exists at all.
Every market, one page each.
Europe
Worked remotely from Hyderabad in English. Each market funds early companies on its own terms, and they are not interchangeable — which is why they are grouped below by how the money is decided rather than by geography alone.
The Nordics
Three EEA states where an agency decides. The money is a grant or an administered credit, it is applied for, and what it tests differs sharply between them.
Finland
The rebuilt Business Finland ladder, and what a small home market does to a plan.
Norway
The grant floor, the pilot-customer condition, and how the tax credit interacts with it.
Sweden
The monthly payroll test, the Vinnova route, and lending against a personal guarantee.
Western Europe
Four markets where the main instrument is written into statute rather than awarded by a panel. Two are outside the EU, and in three of the four a second authority — a devolved agency, a certifying body, a canton — changes the answer.
United Kingdom
A statutory credit nobody awards, a deadline that ignores merit, and four nations on four calendars.
Germany
A certifying body decides whether it is research. Only then does the tax office decide what it is worth.
France
A status won on an annual ratio, and a research credit repaid in cash while the company is young.
Switzerland
The federal money usually goes to the research partner, and the tax relief is worth nothing until there are profits.
Asia
Three Asian markets beyond our own. The instruments here are timed, intermediated or still waiting on a decree, and none of them behaves like a grant.
Singapore
A six-month grant window, a three-year exemption clock, and the cap-table rule that ends it.
Vietnam
Generous new law, several reliefs still waiting on a decree, and how to plan against that.
Japan
A grant that funds two thirds, and two investor reliefs the company itself has to file for.
Oceania
Two markets that pay for research through the tax system rather than through an agency, and test it in opposite directions.
Australia
One self-assessed offset, refundable in cash, published by law, with a ten-month deadline.
New Zealand
Cash against R&D losses that has to be repaid, a 20% staffing ratio, and an agency mid-move.