Non-Dilutive R&D Tax Credits Explained
Most startup funding costs you a slice of your company. R&D tax credits do not. If your team is solving real technical problems, the tax system in the US and the UK may give some of that spending back, without a single share changing hands. Here is how these credits work, what usually counts and how to build the records that make a claim stand up.
Key takeaways
- R&D tax credits reward spending on qualifying research and development with lower tax, or in some cases a cash benefit, without giving up equity.
- In the US, qualified small businesses can use the research credit to offset payroll tax, which helps pre-profit startups.
- In the UK, a merged R&D scheme applies from April 2024, with extra support for R&D-intensive loss-making SMEs.
- Software startups often qualify, but only for work that resolves genuine technical uncertainty, not routine development.
- Good records kept during the year are the difference between a confident claim and a stressful one.
February is when many founders sit down with last year's numbers. Tax returns are being prepared, January's investor update is out, and the next raise is on the horizon. It is also the right moment to ask a simple question: did we do work last year that might qualify for R&D tax relief?
For a startup that spends most of its money on engineers, the answer can be meaningful. Non-dilutive funding stretches runway without changing your cap table, and that matters when every percentage point of ownership counts.
What "non-dilutive" means here
Dilutive funding, such as selling shares, SAFEs or convertible notes, reduces your ownership. Non-dilutive funding does not. Grants, revenue, some loans and tax credits all fall into this second group.
R&D tax credits are a particular kind of non-dilutive funding. You do not apply for them in advance like a grant. Instead, you spend money on qualifying work, keep records, and then claim relief through your tax filings. The benefit arrives later, which is why it helps to plan for it rather than treat it as a surprise bonus.
How credits differ from grants
| R&D tax credits | Grants | |
|---|---|---|
| When you apply | After the spending, through tax filings | Usually before the project starts |
| Competition | Not competitive; you qualify or you do not | Often competitive, with limited funds |
| Scope | Defined by tax law and guidance | Defined by each grant program |
| Main risk | Claims challenged if records are weak | Applications rejected or reporting duties missed |
How it works in the US
The US federal research credit lives in Section 41 of the tax code. It rewards qualified research expenses, which typically include wages for people doing or directly supervising research, supplies used in that research and a share of certain contract research costs. The exact rules on what counts are detailed, so check current IRS guidance and work with an adviser.
The payroll tax option for startups
Many early startups have no income tax to reduce, because they are not yet profitable. That is where the payroll tax election matters. Qualified small businesses can use the research credit against payroll tax instead of income tax.
Eligibility depends on factors such as your gross receipts and how long the business has had receipts. Because these tests are technical and the rules can change, confirm your position with a tax professional before you count on the credit in your forecast.
Common mistake
Assuming that because the company lost money, there is nothing to claim. Pre-profit startups are exactly the businesses the payroll tax option was designed to help. Ask your accountant about it before your return is filed, because the election is generally made on the return itself.
How it works in the UK
The UK reformed its R&D relief for accounting periods starting from April 2024. The previous separate schemes for SMEs and larger companies were brought together into a merged scheme, with additional support for R&D-intensive SMEs that are making losses.
The details, including rates, intensity thresholds and the categories of qualifying spending, are set out in HMRC guidance and have changed several times in recent years. Rather than rely on figures you read in a blog post or a sales email, check the current guidance on GOV.UK and speak with an adviser who handles R&D claims regularly.
What HMRC usually wants to see
- A clear description of the scientific or technological advance you were seeking.
- The uncertainty that a competent professional in the field could not easily resolve.
- How your team tried to overcome that uncertainty, including what did not work.
- A breakdown of the costs linked to that work.
Pro tip
Be cautious with advisers who promise large claims for little effort. Tax authorities on both sides of the Atlantic have increased scrutiny of R&D claims. A smaller, well-documented claim is far better than a large one you cannot defend.
What usually counts as R&D
The specific legal tests differ between countries, but the spirit is similar. Qualifying work tries to resolve technical uncertainty through a process of experimentation or systematic investigation. That is narrower than "anything new we built."
Often in scope
- Developing a new algorithm or approach when the right method was not known at the start.
- Solving performance, scaling or reliability problems that existing tools could not handle.
- Building and testing prototypes to see whether a technical approach was possible.
- Engineering new materials, formulas or processes in hardware, food or manufacturing.
Usually out of scope
- Routine bug fixing and maintenance.
- Building standard features with well-known methods.
- Cosmetic design changes, market research and sales activities.
- Work funded by a client who owns the results, depending on the contract and the country.
Example
Imagine Amara, who runs a five-person startup building route planning software for small delivery firms. Last year her engineers spent several months trying to make route calculations fast enough on low-cost phones. Standard libraries were too slow, and the team tested three new approaches before one worked. That project looks like a strong candidate for an R&D claim. The new billing page they built the same year, using familiar tools, does not. (This is an illustrative scenario, not a real company.)
Building records that support a claim
The best time to prepare an R&D claim is while the work is happening. Reconstructing a year of engineering decisions from memory in February is painful and less convincing.
- Name your projects. Group work into clear technical projects, each with a short description of the uncertainty involved.
- Track time. Record roughly how much time each person spends on each project. Time tracking tools such as Toggl Track, Harvest or Clockify can help.
- Keep the evidence. Save design notes, test results, tickets in tools such as Jira, Linear or Shortcut, and records of failed attempts.
- Separate costs. Tag payroll, contractor and supply costs that relate to qualifying projects in your bookkeeping.
- Write a short summary each quarter. A paragraph per project on what you tried and learned saves hours at year end.
- Review with an adviser. Before filing, ask a qualified adviser to check what qualifies and how to present it.
A simple R&D project log template
Copy these columns into a spreadsheet and add one row per project. The figures below are made-up examples.
| Project | Technical uncertainty | Approaches tested | People and time (example) | Related costs (example) |
|---|---|---|---|---|
| Fast route engine | Could routes be calculated quickly on low-cost phones? | Three algorithms, one successful | 2 engineers, about 4 months | $60,000 wages, $2,000 test devices |
| Offline sync | Could data stay consistent with patchy connections? | Two sync models, both partly successful | 1 engineer, about 2 months | $18,000 wages |
| Billing page | None: standard methods | Not applicable | Not claimed | Not claimed |
Common mistake
Counting every engineer's full salary. Only the share of time spent on qualifying R&D usually counts. Overstating it is one of the fastest ways to get a claim questioned.
Using R&D credits in your funding plan
Because the benefit arrives after the spending, treat R&D credits as a timing tool, not a lifeline. A sensible approach is to leave expected credits out of your base runway forecast and show them as an upside case until a claim is confirmed.
When you talk to investors, mentioning a well-supported R&D claim can show financial discipline. Just be clear about what is confirmed and what is expected. Investors appreciate founders who use every non-dilutive source available, as long as the numbers are honest.
Keep the supporting records with your other tax files. The IRS generally expects records to be kept for at least 3 years after filing, and UK companies keep records for 6 years from the end of the financial year. Longer periods apply in some cases, so check with your adviser.
A quick note: this article is general information about R&D tax relief, not tax or legal advice. The rules are detailed, change often and depend on your situation, so check current IRS or HMRC guidance and work with a qualified tax adviser before you file a claim.
Your R&D credit checklist
- List last year's technical projects and the uncertainty in each
- Estimate time spent by each person on qualifying work
- Tag related payroll, contractor and supply costs in your books
- Gather evidence: notes, tickets, test results and failed attempts
- Ask your accountant about the US payroll tax option or the UK merged scheme
- Leave expected credits out of your base runway forecast
- Set up project tracking now for this year's work
Frequently asked questions
Can a pre-revenue startup claim R&D tax credits?
Often, yes. In the US, qualified small businesses can use the research credit against payroll tax. In the UK, the merged scheme includes extra support for loss-making R&D-intensive SMEs. Eligibility tests apply, so check with an adviser.
Does software development count as R&D?
It can, when the work resolves genuine technical uncertainty through experimentation. Routine development with known methods usually does not qualify, so describe each project honestly.
Do I need a specialist adviser?
It is strongly recommended. The rules are detailed and authorities scrutinize claims closely. Choose an adviser who explains their fees clearly and asks hard questions about your evidence.
Do R&D credits dilute my ownership?
No. They reduce tax or provide a tax-based benefit to the company. No shares are issued, so your cap table stays the same.
The bottom line
R&D tax credits are one of the few sources of startup funding that cost you no equity. If your team spends real time solving hard technical problems, it is worth finding out whether you qualify in the US, the UK or both.
Start by naming your projects, tracking time and keeping evidence as you go, then review a claim with a qualified adviser. Treat the credit as a welcome upside in your plan, and you will stretch your runway without stretching your cap table.
