For businesses that send written proposals for work over a few thousand pounds: your last 20 to 60 proposals scored against a checklist from your own files, with repeat clients, competition and pre-proposal meetings shown alongside; a new template, section by section, with a page limit and a writer for each; a seven-part, one-page executive summary and a before/after rewrite of one of yours; a review checklist with a pass/fail gate; and baselines to measure again the same way, led by the hours it takes to write a proposal.
What do our proposals look like?Each one scored against ten points a buyer would check.
What goes with winning?Win rate by score, with the give or take: association, not proof.
What should we send?A template and a one-page executive summary.
Did it work?Writing hours first; the win rate for direction.
Choose this if your proposals are long, start with your history, and look the same whether you win or lose.
Founder pricing. The method, the calculation spreadsheet with 33 built-in checks, and a complete worked example are built and tested. The first client for this product gets the founder price, in return for permission to measure the results and, if they are good, write them up as a case study. After that, the price goes up. Book a call to take the founder place.
The hours it takes to write a proposal are recorded now from your timesheets and checked again after six months or a year, with the win rate, the share with no decision and the days to a decision alongside; at 20 to 60 proposals a year the win rate can only show direction, not size, and the spreadsheet says so, with the give or take on every change and a blind re-scoring as the early read.
A Foundations product: the groundwork the money products depend on. Its value is shown this way rather than as a single headline figure, and every number in it is labelled counted, estimated or assumed.
The last 20–60, lost as well as won, against a 10-point checklist.
With repeat clients, competition and meetings alongside; the give or take shown.
Each section's purpose, content, evidence, page limit and writer, with a pass/fail gate and nine points anyone can apply before a proposal goes out.
Seven parts, a worksheet and a before/after of one of yours.
Writing hours, win rate, no-decision share and days to a decision, honest about what your volume can show.
On guided and done for you: each enquiry drafted to your template in your own accounts, every question the client asked mapped, for your writer to finish, check and send. Nothing leaves your business, and no decision changes, without one of your people approving it.
A fictional 48-person structural and civil engineering consultancy logged its last 52 proposals. They averaged 17 pages and fewer than one in five opened with the client's situation. Each structural or civils proposal took 16.5 hours to write; 28% were won within 90 days (8 of 29) and 38% had no decision by then.
The team agreed an eleven-section template within ten pages and a seven-part, one-page executive summary, and rewrote one of its own. At about thirty such proposals a year the win rate can only show direction, so writing hours are the primary measure, taken again on 16 October 2027 with a blind re-scoring as the early read.
Your last proposals, lost as well as won.
Ten points each, read by the same people.
The template and the executive summary.
Hours, then the win rate, later.
Your AI assistant drafts each proposal to the template from the enquiry for your writer to finish.
The same method and the same spreadsheet in every option. You choose how much of your own time goes in.
For owners comfortable with a spreadsheet and an AI assistant.
For owners who want a second pair of eyes on the data and the analysis.
For owners who would rather hand over the data and get the answer.
It is built on the open Agent Skills standard and designed for Claude (any plan, with code execution switched on). A paste-in version is included for ChatGPT, Gemini and Microsoft Copilot; it has not yet been tested on them.
The calculations live in the spreadsheet, not in the AI, so the numbers don't depend on which assistant reads them.
Book a short call to talk it through.
Book a call