Sales & PartnershipsFree
RFP Executive Summary Scoring Review
Every sentence assigned to one stated evaluation factor, the word count compared against that factor's published weight, and the gap closed before you submit.
The advice on this page is settled and largely correct. Write it last. Keep it to a page or two. Open with their problem rather than your founding story. Run the four-part shape of needs, outcomes, solution and evidence. Every guide says a version of it, several now add mirror the scorecard, and none of them do the one thing that would tell you whether your draft actually did. River measures it, sentence by sentence, against the weights the buyer published.
The method is arithmetic and takes about a minute. Assign every sentence in the draft to exactly one factor named in the evaluation section, or to nothing. Sum the words per factor. Compare each factor's share against its published weight. What comes back is a ledger with a gap column, and the gap column is usually shocking, because company history is easy to write and a management approach is not.
Then the second pass, which catches the claim nobody can score. Every assertion in the summary gets checked for a pointer into the volume that evidences it, because the evaluator scores the volumes and only reads the summary. Built for the proposal manager assembling the front page at eleven at night, with a round of comments still to fold in. Extracting the requirements themselves is the requirements matrix, the whole submission is the RFP response pack, and the tool index has the rest.
A word spent outside a stated factor earns nothing
Federal procurement is blunt about it. An agency evaluates competitive proposals and assesses their relative qualities solely on the factors and subfactors specified in the solicitation. Not on how well the summary reads, not on the founding story, not on the industry commentary in paragraph two. Commercial buyers running a formal scorecard behave the same way for the same reason, which is that a panel has to defend the award. So the split of words across factors is not a style question. It is a score forecast.
The weights are published, and the buyer is usually required to publish them. All factors and significant subfactors affecting award and their relative importance shall be stated clearly in the solicitation, along with whether the non-price factors combined outrank price, roughly equal it, or fall below it. That is a scoring model handed to you before you write a word. Very few summaries are ever checked against it, and the check is a word count.
The worked example is a 1,180 word draft. Company background takes 402 words, transition logistics another 122, and neither maps to a stated factor. The 40 point technical factor gets 289. Run the arithmetic and the four scored factors are short by 524 words in total, which is precisely the 524 words spent on the two sections that cannot score. Nothing needs cutting for length. The budget is already in the document, in the wrong place.
How it works
Paste their criteria
The evaluation section as written, with factors, subfactors and whatever weights they published.
Paste your draft
The summary if you have one, the volumes if you do not, ideally both together.
River counts and compares
Words per factor against weight per factor, then every claim against the evidence behind it.
Take the rewrite
The summary in their factor order at their weights, plus the claims needing a locator.
What you get
- Every sentence assigned to one stated evaluation factor, or to nothing at all
- Word count per factor against its published weight, with the gap in words
- The rewrite at the same length, in the buyer's own stated factor order
- How far into the summary an evaluator reads before the heaviest factor appears
- Every assertion checked for a pointer into the volume that evidences it
- The claims that exist only in the summary, which are the ones to evidence or drop
Common questions
What if they never published weights?
Then the factors are still listed and their order still carries meaning, because a buyer who lists technical approach first and price last has told you something. River allocates against stated order and any language about relative importance, flags that the weights are inferred, and reports the split so you can argue with it rather than guess at it.
Isn't proportional allocation a bit mechanical?
As a rule to follow blindly, yes. As a diagnostic, no. The point is not that a 40 point factor deserves exactly 40 percent of the words. It is that 34 percent of your summary went to company background, which is worth nothing, while the heaviest factor got a quarter. You can disagree with the target and still act on the gap.
Our summary is over the page limit already.
Then the ledger is more useful, not less. It tells you which words to cut with no score consequence at all, which is normally the company history and the transition narrative. In the worked example that recovers 524 words, and the four scored factors happen to be short by exactly 524. The cut and the fix are the same edit.
Why check whether a claim points at a volume?
Because the evaluation runs against the volumes, and the summary is a reading aid. A claim about accuracy tolerance that appears nowhere but the summary cannot earn a technical point, and it can still be quoted back at you after award. Nine of thirty-one carried a pointer in the worked draft, which is ordinary for a first pass. An experience claim points at a nominated reference.
Can it write the summary from scratch?
Yes, if the volumes exist. It reads them, pulls what each factor is actually evidenced by, and drafts to the weighted allocation with locators already attached. Writing it before the volumes is the one order that does not work, because there is nothing yet for the claims to point at and the allocation has nothing to allocate.
Does this handle subfactors?
It does, and subfactors are where most of the damage is. A 40 point technical factor split into two subfactors scored separately means two allocations, not one, and a summary that spends all 472 words on the first subfactor loses half the factor. The ledger runs at whatever level the solicitation actually scores.
Where does this sit against deciding whether to bid?
After it. Whether the bid is worth writing at all is the bid or no bid assessment, and the answers you reuse across submissions live in the answer library. This runs once the volumes are drafted and the front page is the last thing standing between you and the portal.
RFP Executive Summary Scoring Review
Fill in the form and your workspace opens with the work already underway.