The Forwarded Version: your proposal argues your case in rooms you'll never enter

The scene I'm in this week

Last Tuesday I got on a call with the CEO of a $19M Series B healthtech company. They build credentialing and claims software for behavioral health provider networks. He was frustrated in the specific way founders get frustrated when nothing obvious went wrong.

The deal had been good. Discovery was sharp. The demo ran long because the buyer kept asking real questions instead of polite ones. Their champion, a VP of operations, said out loud on the call that this was the fix she'd been hunting for since January. They sent the proposal on a Thursday. Then three weeks of quiet, and a two-line email: we're going to revisit this in Q1.

I asked him to forward me the proposal. Page one was a company overview. Founded in 2016, a paragraph about the leadership team, six customer logos. Page two was a scope table, eleven rows, each one a deliverable with a checkmark next to it. Page three was pricing and terms. Across nine pages the document never said what happens inside a behavioral health network when credentialing takes ninety days, or why this company had built something different because of it.

Then I asked when he'd last read the thing end to end. He hadn't. Not once. The template came from the deal desk, the deal desk got it from an operations hire, and the operations hire brought it over from her last employer. The sharpest story in that building had been told out loud twice a week for two years, and it had never made it into the document the deciders actually read.

Naming what's actually broken

Here's what makes the proposal different from everything else you send. Every other artifact in your funnel has a person attached to it. Somebody narrates the deck. Somebody drives the demo. Somebody answers the follow-up email within the hour. The proposal goes out and then travels on its own, into rooms nobody from your team will ever enter.

Your champion hits forward. It lands with a CFO who was never on a call, an operations leader who heard about you secondhand, a procurement analyst whose entire job is comparison, and sometimes a board observer who reads four of these a month. Call it the Forwarded Version: the account of your company that argues your case in rooms you'll never enter, written by somebody who was never in the room where the story got told.

The person who owns that template usually sits in finance or operations. They inherited it, and they built it to answer the questions their side of the house gets asked, which are questions about scope, liability and price. That's a sound document for a deal desk. It's a poor document for a decision.

Look at the order most templates run in. Who we are, what we'll do, what it costs. Three moves, and the buyer's situation shows up in none of them. That's Solution-Centric Marketing with a signature block on the end. You spent six weeks getting a buyer to see their own problem differently, then handed the deciding room a document that never mentions the problem at all. This is just truth.

It's the same gap that shows up when How do you equip a champion to sell you to the buying committee? goes unanswered. Your champion is willing. She just has nothing in writing that makes your case as well as you would, so she forwards the quote and hopes.

Why this is worse now than ever

There was a stretch when a good proposal was a signal on its own. Producing a clean, well-designed, forty-page response took a team and most of a week, and buyers read that effort as seriousness. If yours looked better than the other two, it meant something real.

That edge went to zero. Any vendor in your category can generate a polished proposal in an afternoon now, with the design, the case studies and the executive summary all rendered on demand. Polish stopped being evidence of care the moment it stopped costing anything.

The second shift is the one most founders haven't caught up to. Your proposal is being read by a machine before a human forms an opinion about it. The buying group drops three PDFs into ChatGPT or Claude and asks the obvious question: compare these, what are we actually choosing between? The model does exactly what it was asked. It finds the dimensions all three documents share and compares on those. Hand it a scope table and a number, and it hands back a price comparison, because price was the only thing you made comparable.

“While AI can write the code, humans must still write the story and sign the contract.”

... Margin of Safety #43, 2026

The person signing still needs a reason, and that reason now has to survive being compressed by software and read by someone who missed every call. What's scarce is a document that says something only your company could say. Everything else in the packet can be produced by anybody, in an afternoon, for nothing.

The diagnostic ... run this on your last three proposals

You don't need to hire anyone to find out what your Forwarded Version is arguing on your behalf. Pull the last three proposals you sent and run these tests this afternoon.

1. 1The First Page Test. Read page one of each proposal out loud, and tally every sentence by what it's about: your company, or the buyer's situation. Most page ones come back nine to nothing. The first thing your deciders learn about you is that when you finally had their full attention, you spent it talking about yourself.
2. 2The Stranger Forward. Send your most recent proposal to somebody with zero context. A board member, an advisor, a friend who runs a different kind of business. Ask two questions and time the answers: what problem does this company end, and why them instead of the other two? If it takes longer than a minute, keep in mind the CFO deciding your deal has less context than your friend and considerably less patience.
3. 3The Machine Read. Paste your proposal and two competitors' into ChatGPT or Claude, then ask which one they should pick and why. Read the answer cold, because it's close to the summary your buying committee is working from. When the model's reasoning turns entirely on price, timeline and implementation risk, those were the only dimensions you gave it.

Three tests, one afternoon. If all three point the same direction, the deal you lost last quarter was decided by a document nobody on your team had read in a year.

What I see across 100+ B2B companies

I've now sat with well over a hundred B2B leadership teams, most of them growth-stage companies between $5M and $75M in revenue, and the proposal is the most consistent blind spot in the whole funnel. Homepages get argued over for months. Decks get rebuilt every year by somebody new. The proposal template gets touched when legal changes a clause.

Ask a room who wrote it and when, and watch what happens. Somebody says operations owns it. Somebody else says it came over from the deal desk. Nobody can name the person who wrote page one or the year it was written. The narrative work stops at the demo, which is exactly one step before the document that decides.

The cost lands where it's hardest to see. eMarketer's 2026 research found 97 percent of B2B buyers say trust in the vendor is a decisive purchase factor. Trust gets built by a person right up until your champion hits forward, and from that point it's built or lost by a document. When that document opens with your founding year and closes with a number, you've handed the most important room in the deal nothing to trust except your price.

Underneath it sits a second pattern. Reps who close well are carrying the story in their mouths, live, a little differently every time. That works beautifully while they're in the room and dies the second the PDF goes out, which is the same mechanism behind Why do our demos go great but deals still stall?. And on the rare occasion somebody does try to write the real story into the template, The Consensus Edit: why your sharpest line never ships takes over. Legal softens the claim, finance adds a disclaimer, and the sentence with a point of view gets sanded back into a scope row.

A real example

A PE-backed company at $38M in revenue, building escrow and closing coordination software for title insurance agencies. Their problem showed up as a proposal-stage graveyard: 41 proposals across two quarters, 11 closed, and 19 that went quiet and never came back with a reason anyone could name.

We ran the Stranger Forward first. Five readers, none of them in the industry. Four couldn't say what problem the company ended. The one who could had guessed it from a customer logo. Then we ran the Machine Read against two competitor proposals we could get our hands on. The model recommended the cheapest of the three and described all three as broadly similar workflow platforms.

The truth was already in the building. The CEO had said it to me in the first hour: closings die in the last seventy-two hours, when the title agent, the lender and the buyer's attorney are reconciling documents by email and nobody owns the clock. That sentence had never appeared in a proposal.

We rebuilt the document off the framework we'd written with them. Page one names what goes wrong in the last seventy-two hours and what a dead closing costs an agency. Page two is the change: what those seventy-two hours look like when one system owns the clock and every party can see it. Page three is proof, two agencies and their numbers. Scope and pricing moved to page four, where they read as the answer to something rather than as the whole document.

Four months on, proposal-to-close moved from 26 percent to 39 percent, and the average time from send to signature dropped from 71 days to 52. The change the CEO liked most never showed up in a dashboard. Procurement started asking him questions about the outcome instead of asking him to justify line items.

What this means for you

If your deals go quiet after the proposal, read your last one out loud before you touch anything else in the funnel. There's a real chance the document carrying your case into the deciding room has never been positioned by anyone, and reads like a quote from a vendor instead of an argument from the company that understands this problem better than anyone else in the running.

Here's where it connects to what we do. Your proposal falls back on scope and price because there's nothing else written down for it to reach for. The version of your story that would win the room lives in your mouth and your best rep's, and a template can't get in there. The Magnetic Messaging Framework is where that gets fixed. It's the documented version of who you're for, the villain you end, the change you deliver, and the proof that it's happened before, written so a stranger can repeat it and an AI can run on it without going generic. Once it exists, the proposal becomes the closing chapter of the same story your homepage and your deck are already telling, and your AI tools can draft it without inventing a new company every time. That matters because the Forwarded Version keeps arguing for you in rooms you'll never enter, whether or not you ever decide what it says.

Three things to do this week:

1. 1Read your last proposal out loud, all the way through. Read it the way a CFO would at six in the evening, with two other vendors' documents open. Most founders have never done this once, and nobody who does it forgets what they found.
2. 2Move the buyer's problem to page one. Before any redesign or new template, take the sentence you say on discovery calls when you're being honest about what goes wrong in their world, and put it where the document opens. Scope and pricing still belong in there. They belong after the reason.
3. 3Give the template an owner who was in the room. Whoever holds it now inherited it from somewhere else. Page one needs an owner who has heard the story told well, and it should get rewritten every time the positioning changes.