Innovate UK Smart Grants: the real success rate nobody puts on their homepage

If you have looked into UK innovation funding, someone has probably told you about Innovate UK Smart Grants: up to £500,000 of non-dilutive funding for game-changing ideas, open to any sector. What they probably did not tell you is the number that should drive your decision about whether to apply.

The number

In its September 2024 round, Innovate UK's own statistical breakdown showed 46 funded projects out of 1,645 applications — a 2.8% success rate (reported by Venturenomix). That is down from industry estimates of 10–15% in earlier years. Roughly 35 out of every 36 applicants — many of whom spent 40+ hours on their application, or paid a bid writer thousands of pounds — got nothing.

Since then, the Smart Grants scheme has been paused for evaluation.

We are publishing this because almost nobody else in our market will. A consultancy charging a success fee has no incentive to lead with a 2.8% base rate. We charge a subscription, not a percentage of your award — so we can afford to tell you the truth before you spend a month writing.

What a 2.8% base rate actually means

A few things follow directly from the arithmetic, and they are worth sitting with:

• A bid writer advertising a "70% win rate" deserves scrutiny. When the base rate is 2.8%, an extraordinary claim needs an extraordinary denominator. Ask how many applications they worked on in the round, not just how many won. Ask to see assessor scores from their last submissions — Innovate UK returns them.
• Quality of fit matters more than quality of prose. Most rejected applications were not badly written; they were a weak fit for the competition's scope. The cheapest improvement to your odds is not a better writer — it is not applying to competitions you were never going to win.
• The application is not the end of the work. Innovate UK runs an interview stage for shortlisted applicants, and post-award, claims are audited. Every year, hundreds of UK businesses have funding delayed, withheld, or clawed back at audit — usually not for fraud, but because they did not understand the cost rules, subcontractor procurement requirements, or reporting conditions that only become visible after the award.

A word on AI-written applications

One more thing the base rate interacts with: UKRI now has an explicit policy on generative AI in applications. It does not ban AI assistance — but it holds you accountable for every claim in your submission, warns that the confidentiality of information entered into AI tools is not guaranteed, and reserves the right to reject applications and reclaim awarded funds where misuse is upheld.

Practically: a generic, AI-padded application into a 2.8% competition is a way of spending hours to lose. If you use AI at all, it should be grounded in your real organisation, your real project, and the funder's actual published criteria — and you should be able to stand behind every sentence in front of an assessor.

What we think you should do instead

Our honest advice, which applies whether or not you ever use GrantIQ:

1. Check eligibility and fit before you write a word. Read the scope document against your project, harshly. If you would not score yourself in the top decile of fit, your time is better spent elsewhere.
2. Widen the funnel. Innovate UK is one funder. There are over a hundred UK and EU funders running live programmes at any time — many with materially better odds for the right applicant. The best response to a 2.8% competition is often a different competition.
3. If you win, treat the award letter as a contract, not a celebration. The reporting deadlines, spend conditions, and procurement rules in it are where funding gets withheld. Extract them, diary them, and keep evidence of compliance from day one.

That is the workflow GrantIQ is built around — eligibility checked before a grant is ever shown to you, discovery across 100+ UK and EU funders, and post-award obligation tracking that turns the award letter into a deadline list with a source excerpt for every obligation. But the advice stands on its own. The grant landscape does not need more optimism. It needs better arithmetic.

Figures cited are from published sources linked above, current as of June 2026. Success rates vary by round and scheme; always check the live competition page before relying on any number — including ours.