How to write a grant application that scores

Grant applications aren't read; they're marked. An assessor — often a practitioner scoring a stack of bids against a published rubric, question by question — is looking for reasons to award points, and reasons they're allowed to. Writing for that reader is a different discipline from pitching an investor or a customer, and most of what makes applications lose is avoidable.

Start from the scoring, not the story

Nearly every UK competition publishes what it assesses — sometimes as explicit scored questions (Innovate UK's application forms), sometimes as criteria in the guidance document. Before writing a word, extract them and treat each one as a contract: this question is worth points; my answer's only job is to earn them.

The most common structural failure is answering the question you wish had been asked. If the question asks about your team's capability to deliver, a paragraph about how exciting the market is earns nothing — however good it is.

Evidence beats adjectives, every time

Assessors are explicitly trained to discount unsupported claims. The difference between a 6 and a 9 on the same question is usually not better ideas but better evidence:

• "We have strong industry interest" → "Two operators covering 3.1GW of installed capacity are in commercial discussions; one paid deployment has run for 18 months."
• "Our team is highly experienced" → "Our CTO led BVLOS safety cases approved by the CAA at two offshore sites."
• "There is a large market opportunity" → a sized, sourced, dated number — with the source named in the text.

A useful edit pass: highlight every adjective and ask what verifiable fact could replace it. If nothing can, the sentence is probably costing you credibility rather than earning points.

Answer the question behind the question

Most scored questions carry an implicit worry the funder is really probing:

• "Describe your project" is really "is this in scope for what we're allowed to fund?" — use the call's own vocabulary for objectives.
• "Why is grant funding needed?" is really "would this happen anyway?" — additionality is a criterion assessors must mark, and "we'd do it slower" is a weak answer. What genuinely changes: scope, timing, risk level, partners?
• "Describe your team" is really "will our money be safe with these people?" — delivery track record beats brilliance.
• "What are the risks?" is really "do they understand their own project?" — a short risk list with named mitigations scores better than claimed fearlessness. "No significant risks" is a red flag, not reassurance.

The mistakes that quietly cap your score

• Recycled text. Assessors read stacks of applications; paragraphs written for a different call (or plainly generated by AI with no specifics) stand out immediately. Every claim should be one only your company could make.
• Ignoring word limits — in either direction. Overrunning gets truncated or penalised; wildly underrunning signals thin thinking. Aim close to the limit with dense, specific content.
• Numbers that don't reconcile. If the budget table, the work-package costs and the prose disagree, assessors notice — and it reads as carelessness with money.
• Unsupported partners. Naming a partner without a letter of support invites the assessor to assume the partnership is aspirational.
• The invisible plan. Prose that never touches work packages, milestones or who does what makes deliverability unscoreable. See our companion guide on building the project plan funders expect.

A drafting process that works

1. Extract the scored questions and the criteria from the guidance document.
2. Collect your evidence first: deployments, figures, letters, track record — the raw material that will separate your answers.
3. Interview yourself per section: what would only we know here? What number, name or event proves it?
4. Draft dense, then cut to the word limit, keeping evidence and cutting throat-clearing.
5. Mark your own bid against the criteria before submitting — coldly, question by question — and fix the weakest answers rather than polishing the strongest.

This is the process GrantSpark automates: it interviews you before drafting so the section is built around your evidence, then reviews the draft the way an assessor would — against the funder's own criteria — and proposes rewrites for the weakest sections.