Skip to content
The GrowthCred field guide

Your proposal problem is not writing. It is rebuilding the brief every time.

A blank page makes every proposal feel custom. But most of the work repeats. You hunt for the service description. You rebuild the discovery call from memory. You check the pricing rules. You recreate the same structure. Again. AI helps when it puts approved pieces together. It hurts when you ask it to invent the missing ones.

By Phila Ngwenya · Published

The contrarian view: stop trying to write faster

Most proposal advice starts with a better prompt. That's too late. Your scope lives in meeting notes. Your prices live in a spreadsheet. Your tone lives in your head. So the draft still waits for someone to assemble the truth.

So forget faster typing. What you want is a proposal system. It gives whoever drafts, a person or AI, the same approved inputs every time. And it shows you what's uncertain before anything reaches a prospect.

Build the four proposal assets once

Keep what's reusable apart from what's client-specific. It's easier to maintain. And one customer's confidential detail can't leak into someone else's proposal. You need four things.

First, a business brief. Your positioning, your services, your tone, the proof you're allowed to use and who owns each fact.

Second, a discovery record. The prospect's problem, the outcome they want, their constraints, who's involved and the exact next step.

Third, your commercial rules. Where approved pricing lives. Who can give a discount. Standard exclusions, payment terms and how long a quote stays valid.

Fourth, a proposal structure. Headings, required sections, a review checklist and examples of approved final work. Build these once. Use them every time.

Use a source-bound briefing

Tell the tool to use only the business brief, discovery record and commercial rules you give it. Tell it to mark every missing fact as a question. And to keep assumptions apart from confirmed scope. Say which sections you need and who'll review it.

Here's an instruction I'd use. Prepare a first draft from these sources. Don't invent prices, dates, client facts, proof or commitments. List missing information before the draft. Put the source next to every commercial figure so the reviewer can check it.

Review the proposal where it can actually hurt

Pretty wording isn't the risk. The risk is a polished mistake that changes the deal. The recipient. The scope. Exclusions, quantities, maths and dates. Dependencies, IP and payment terms. The next step.

Work out totals outside the draft. Check every testimonial or result has permission. Strip out internal notes and open questions. Make one person responsible for approval. Then keep that approved version as your next example. Before you send, run this list.

  • Can every price and commercial term be traced to an approved source?
  • Does the scope match what the prospect actually asked for?
  • Are exclusions and client responsibilities explicit?
  • Has information from another customer been removed?
  • Does the proposed next step match the conversation?

Measure turnaround without fooling yourself

Record the full baseline first. Finding inputs. Drafting. Fixing. Review. Sending. Then test the new workflow on a few similar proposals, including one with messy notes. A draft that shows up fast but needs a forensic review hasn't fixed anything.

Track the elapsed time, your hands-on time, the fixing time and how many missing-input questions came up. Keep expected gains apart from measured ones. Faster proposals make you more responsive. They don't guarantee the prospect buys.

What does a measured result look like? TaiAscend reported that a document process that took three days now takes two hours. That's their process, measured by them. Yours will land somewhere else. That's exactly why you measure.

Build it yourself or have it implemented

Want to build it yourself? Take the one-day workshop. You'll create your business brief, map your proposal inputs and practise the first source-bound draft. Bring a made-up or properly redacted example and your approved commercial rules.

Would you rather we build it? Apply for implementation. That fits when proposals pass through several people or tools, pricing needs controlled sources, or you need intake, document generation and approval steps. We scope it, test it and hand it over. Every proposal you send is still yours.

Frequently asked questions

Can AI write the whole proposal for me?
It can prepare a first draft from approved inputs. But a person still confirms the client facts, scope, figures, terms, exclusions and commitments before you send it.
What should go into the business brief?
Your positioning, services, tone, approved proof and who owns each fact. Keep client-specific notes and anything confidential in a separate discovery record.
How do I know whether the workflow is faster?
Measure the whole task before and after: finding inputs, drafting, fixing, review and sending. Test a few similar proposals. Count the missing-input cases as well as the minutes.
Should I attend the workshop or apply for implementation?
Take the workshop to build and understand your first proposal workflow yourself. Apply for implementation if the process crosses people or systems and needs controlled inputs, approvals, testing and handover.