R&D Radar

UK R&D tax relief · screening & claim narrative

51/51

verdicts judged correctly in a 51-scenario stress test — including nine non-qualifiers we planted to catch it out.


Is your project R&D? Find out in 60 seconds — free.

Describe it in plain English. Get a verdict with the exact HMRC guideline paragraphs — and if it qualifies, a drafted claim narrative for your accountant. If it doesn't qualify, we'll tell you straight.

For UK limited companies. No card, no percentage fees. Informational, not a claim.

Three possible verdicts

Likely R&D
It qualifies. You see your biggest gap and what HMRC will still need, free — and can unlock the drafted HMRC narrative and evidence playbook as a pack.
Possibly
It hinges on one thing. The engine asks you a single settling question, then re-runs and updates the verdict for free.
Unlikely
It doesn't qualify — and you're told which of the four statutory tests it failed, so you're not left guessing.

Most screeners only know how to say yes. This one is built to say no — which is what makes the yes worth taking to your accountant.

The claims that never get made

HMRC paid out £7.6 billion in R&D tax relief last year — but the firms this product is built for are exactly the ones who never find out they qualify.

the average SME claim last year
£85k
the average SME claim last year
of small firms are aware the relief exists at all
41%
of small firms are aware the relief exists at all
first-time SME claimants in a single year — eligible firms scared off, not gone
−45%
first-time SME claimants in a single year — eligible firms scared off, not gone
of all claims come from manufacturing — £2.0bn of relief
26%
of all claims come from manufacturing — £2.0bn of relief

There's also a clock. If you haven't claimed in the last three years, you must notify HMRC within six months of your accounting period ending — and more than three quarters of accountants have seen a client lose a claim to that deadline. A free screening tells you in about a minute whether it's worth starting that conversation.

Sources: HMRC R&D tax credit statistics, September 2025 (SME average derived from £3.15bn of SME-scheme relief over 36,885 claims); FSB awareness evidence to the House of Lords; accountant deadline survey reported by the R&D Community, 2025.

How it works

  1. 01

    Describe a project in plain English

    What you were trying to do, what went wrong, what you tried. No tax wording — a paragraph is plenty. It reads pasted engineering reports too.

  2. 02

    Get a cited verdict in about 30 seconds

    Likely, possibly or unlikely — with the exact paragraphs of the UK R&D guidelines it rests on. If it's borderline, it asks you one question and updates the verdict when you answer.

  3. 03

    Take a drafted claim narrative to your accountant

    Qualifying projects can unlock a drafted HMRC Additional Information Form narrative — baseline, advance, uncertainties, how you resolved them — and, with the claim pack, a cost schedule and relief estimate to take to your accountant.

What this normally costs you

Finding out whether your work qualifies — and getting it written up the way HMRC now asks — is the expensive part of an R&D claim. Here's what the alternatives look like.

An R&D tax adviser15–30% of whatever you get back
Specialist firms work on a percentage, with published minimum fees running from £990 to £4,000 — so a headline “10%” with a £4,000 minimum is really 27% of a £15,000 claim. Some contracts run for several years and bill in years you don't claim at all.
Your accountantA separate fee, plus a wait
They file the return, but the technical narrative is the specialist part — many general practices either don't write them or price the R&D work separately. Either way you're paying someone to interview you about work you already did.
Working it out yourselfFree, and a lot of statutory reading
Whether it counts is set by the government's DSIT Guidelines, and before you file HMRC wants an Additional Information Form naming each project, the scientific or technological uncertainty, and how you went about resolving it.

R&D Radar gives you the verdict before you commit to any of them — so you walk into that conversation already knowing whether you have a claim, what it rests on, and where the gaps are.

Verdict
Free
With the DSIT paragraphs it rests on, your biggest gap, and a cost template.
Narrative pack
£349, once
The project narrative for HMRC's form and the evidence playbook, printable.
Claim pack
£1,000, once
Narrative plus cost schedule and relief estimate, for your accountant to check.

Flat prices whatever the claim is worth. It will never be a percentage of your claim.

Fee ranges are typical UK market rates, not quotes — every adviser prices differently, and a good one earns their fee. R&D Radar flags and drafts; it doesn't file.

We're not a claims farm

HMRC is rejecting bad R&D claims and investigating the advisers who filed them. R&D Radar applies the government's own definition of R&D — the DSIT Guidelines — and cites its reasoning paragraph by paragraph. That means it says no a lot. For example, none of these qualify:

  • Unlikely

    A fiddly website build on an established platform

    effort and complexity aren't uncertainty

  • Unlikely

    New recipes, menus or flavours

    culinary craft, not science or technology

  • Unlikely

    An 'innovative' marketing or business method

    commercial novelty isn't a technological advance

  • Unlikely

    Adapting standard techniques to an awkward job

    skilled work a competent professional could plan

A screener that tells you no is a screener whose yes means something — to you, to your accountant, and to HMRC.

Built for the companies that don't realise they do R&D

Most under-claimed R&D isn't in tech startups — it's in firms solving hard production problems who'd never call it “research”:

  • Manufacturers changing materials or processes
  • Food & drink producers solving shelf-life, stability or process problems
  • Engineering and fabrication firms doing what had no established procedure
  • Product and software teams building what off-the-shelf tools couldn't do

Solved a hard technical problem in the last two years?

It takes one paragraph to find out whether it's worth a claim.

Screen a project free