River
Y CombinatorBacked by Y Combinator
FREE TEMPLATE

Consulting Proposal and Criteria Map

Five documents and three sheets built so every section answers something the buyer will compare on, including the criterion they only ever said out loud.

Free download  ·  No account needed

Requirement Response  ·  4 of 11 rows  ·  two plant operations diagnostic, specialty food manufacturer

How we know it is the column that decides the bid

RefWhat they will compare onHow we knowWeightStatus
C-01Whether the work can run without stopping the production linesSaid on the callHeaviestAnswered, approach section 2
C-06Whether their own team can keep it running after we leaveInferredMediumAnswered, half a page
C-04Price against a budget nobody has statedInferred, needs confirmingUnknownClarification asked
C-08Our benchmark data across 40 comparable manufacturersNot stated anywhereNoneCut

C-01 decides this bid and it is in no document. The COO raised it unprompted, raised it again at the end, and described the previous consultant keeping his supervisors in workshops all week. That is why the work is out to tender. A bid that answers the two page brief perfectly and books a week of workshops loses to one offering forty minute slots around the line.

C-08 is the row that will annoy you. Benchmark data across forty comparable manufacturers is the strongest asset this practice owns, and nothing the client said touches it. So it leaves the proposal, where it earns nothing and consumes the pages C-01 needs, and one sentence of it goes in the cover letter instead.

Eleven rows in the full map. Three are stated in the brief, four came out of the call, three are inferred, and one is asked for by nobody.

Search this and every result hands you the same eight sections. Executive summary, problem statement, approach, deliverables, timeline, team, pricing, terms. The better ones tell you to open with the client's problem instead of your bio. All of it is a container, and bids are not lost on containers. They are lost because a proposal answered the written brief beautifully and missed the thing the buyer cared about most, which was said out loud once and written down nowhere.

So the first artifact is a map of what they will compare on, with a column recording how you know each one. Stated in the brief. Said on the call. Inferred and needing confirmation. Not stated anywhere. In the sample the heaviest line is whether the work can run without stopping the production lines, which the COO raised twice in conversation and which appears in no document. A section that maps to nothing gets cut, including the strongest thing the practice owns.

Public procurement is the one place the rule is published. FAR 15.305(a) requires an agency to evaluate solely on the factors the solicitation stated, and FAR 15.304(d) requires it to state their relative importance. A formal solicitation hands you the criteria, so that job is compliance and the RFP response pack owns it. Here you infer them, then price so every day names the condition it rests on and every exclusion appears as a zero line.

The map, the price that names its conditions, and the approach that gives nothing away

Every day in the price carries the assumption it rests on. Every exclusion is a zero line rather than an appendix sentence, because a boundary registers where the reader is thinking about money.

Requirement Response

All eleven rows for a fictional specialty food manufacturer, Marnwood. Two plants, board date fixed in week 8.

RefCompared onHow we knowWeightWhere answeredStatus
C-01Work runs without stopping the linesSaid on the callHeaviestApproach section 2Answered
C-02Numbers survive the finance directorSaid on the callHighApproach section 3Answered
C-03Comparable food manufacturing experienceStated in the briefMediumTeam and credentialsAt risk
C-04Price against a budget nobody has statedInferredUnknownPrice sectionClarification asked
C-05Delivered before the board in week 8Stated in the briefGateDeliverable scheduleAnswered
C-06Their team keeps it running afterwardsInferredMediumApproach section 4Answered
C-07Make-or-buy call on co-packingIn the brief, absent from the call9 days either wayNot yetClarification asked
C-08Benchmark data across 40 manufacturersNot stated anywhereNoneCover letter onlyCut
C-09How quickly we can startSaid on the callLowCover letterAnswered
C-10Whether the sponsor wants to work with usInferred, unwritableNot scored on paperNot applicableNoted
C-11Two other consultants are biddingSaid on the callContextNot applicableNoted

C-07 is the row that stops a guess becoming a price. The brief asks for a make-or-buy recommendation on co-packing in one line. Forty minutes on the call and the COO never raised it once. Nine days of work sit between the two readings, so it becomes a clarification question and a line in the price marked not yet priced, rather than a quiet decision to bid the cheaper reading.

C-03 is marked at risk in our own map. Three references were asked for and this practice holds two. The third goes in as key personnel experience with the labelling made explicit, which is the difference between a reference that gets considered and one that fails at verification.

C-10 and C-11 never become sections. Recording them stops you trying to answer C-10 in prose, which always reads as trying, and points at the twenty minute phone call that would actually move it.

Pricing Build

Nine of fifteen lines. Every day carries the condition it rests on, and exclusions are lines rather than appendix sentences.

LineActivityDaysCostRests onIn the price
P-02Client-side extraction of the six named exports00A-01Client provides
P-03Line-level throughput and yield across both plants108,500A-01Yes
P-04Nine interviews and two shift observations per plant57,250A-02Yes
P-06Reconstructing downtime codes from paper shift logs00X-02No
P-08Two rounds of comment from sponsor and finance director22,900A-04Yes
P-10Make-or-buy recommendation on co-packing00Q-01Not yet priced
P-11Implementation of anything recommended00X-01No
P-13Contingency at 8 percent of labour 2,942A-01Yes
P-15Fixed fee including travel and contingency29.541,697 Yes

P-03 is the largest exposure in this price and the sheet says so. Ten analyst days assume the production export arrives at run level with downtime reason codes attached. At shift summary level the identical analysis takes seventeen. Priced as a condition, the buyer accepts the grain or funds the difference before signing. Written as a disclaimer, it is an argument in week four that you are having from behind.

P-06 and P-11 are zero lines deliberately. An exclusion a reader had to hunt for in an appendix gets treated as something you hid. Implementation at zero, met while they are working through the money, is a boundary that registers. Both carry their exclusion reference, so the sentence behind them is one line away.

Contingency sits against A-01 by name rather than spread across the total. Where the export arrives at run level it is never drawn and the final invoice comes in lower, which is worth saying out loud in the letter.

Deliverable Schedule

Six of eight rows. The acceptance basis is written to be arguable, because a vague one is why a final invoice sits unpaid.

RefDeliverableFinal dueAcceptance basisRounds
D-01Data request and coverage noteWk 1 day 5Marnwood confirms each of the six exports is producible, or names a substitute1
D-02Interview and observation scheduleWk 2 day 118 named people confirmed, with 40 minute slots held around the line1
D-03Throughput and yield packWk 5 day 3Output volumes reconcile to the finance ledger for the same period2
D-04Margin rebuildWk 5 day 5The finance director agrees the standard cost bridge ties to the audited gross margin2
D-06Board sessionWk 8 day 2Delivered with the decision log completed in the roomNot applicable
D-08Weekly progress noteEvery FridayIssued, no acceptance stepNot applicable

D-02 turns the criterion from the call into an acceptance test. Forty minute slots held around the line, eighteen names, no workshops. The heaviest thing in the bid now has a date and a definition of done against it, which is a different order of promise from a sentence in the approach.

D-03 moved because D-01 would not let it stand. The first draft had the pack finalised in week 3. Under D-01 the exports cannot land before week 2, so week 3 was a date we could not have met and would have been held to. Two dependent rows moved with it.

D-06 is the only immovable date here and every upstream date is sequenced backwards from it. D-08 sits outside the acceptance regime on purpose: a status note that needs signing off is a status note nobody reads.

Approach Narrative, the four sentence types

The approach is the heaviest factor in most bids and the section where consultants hand over the engagement trying to win it.

KindThe sentenceWhat it proves
InstrumentCost per work order is calculated from the work order export rather than the maintenance ledger, because the ledger nets off warranty recoveries and the work order file does notKnows both files and why they disagree
Failure modeWhere the exception procedure and the observed practice diverge, the divergence is the finding, so both get documented rather than reconciledHas met the case where the procedure is fiction
ThresholdA theme reported to the board carries the count of interviewees who raised it and the count who were asked, so a view held by two of fourteen is never written as widespreadHas been challenged on a number in a boardroom
Stopping ruleSampling stops when three consecutive files add no new failure mode, which is usually between forty and sixtyHas actually run the sample

Each of those reads as fifteen years of practice and none of them asserts anything about this client. Compare a ranked list of where you expect their costs to be wrong, or a hypothesis drawn from their published accounts. A buyer holding your diagnosis has the option of thanking you and briefing it internally, and some of them take it.

Where the brief demands a worked sample, mark it the way procurement requires. FAR 52.215-1(e) wants a legend on the title page and a second legend on every individual sheet holding the restricted material, with those sheets identified by number. Unmarked pages are unrestricted pages, and applying the rule properly takes ten minutes.

Page budget follows the weights in the map. Where the approach is the heaviest criterion and the bid gives it a page and a half, the bid is losing on arithmetic rather than on judgement.

What's in the pack

01

Requirement Response

One row per thing they will compare on, with how you know it, its weight, the page that answers it, and the evidence offered against it.

02

Pricing Build

Every day attached to the condition it rests on, exclusions carried as zero lines, and contingency held against one named assumption instead of spread across the total. Testing that fee against what similar engagements actually took is the pricing and fee model pack's job, before the number lands here.

03

Deliverable Schedule

Draft date, review window, final date, comment rounds inside the fee, and an acceptance basis specific enough to argue from, matched to what the findings pack actually delivers.

04

What They Are Actually Comparing

The four sources a criterion comes from, the four questions to ask the buyer, and the rule for cutting your best material when it maps to nothing. Where the only channel is a written RFP, splitting those questions between the shared window and your own assumptions replaces the call.

05

Approach Narrative

Instrument, failure mode, threshold and stopping rule: four sentence types that prove method convincingly and assert nothing at all about this client.

06

Assumptions and Exclusions

Each assumption states the number it becomes when the condition fails, so a defensive paragraph turns into a choice the buyer makes before signing.

07

Team and Credentials

Named people against phases with the day split visible, plus the labelling that makes an individual's prior firm experience offerable rather than disqualifying.

08

Proposal

The document itself, with its page budget assigned from the weights in the map rather than eight headings of roughly equal length.

How to use it

  1. 1

    Open in River, or take it blank

    Open the pack in River with the brief and your notes from a qualified discovery call, or download the Word documents and CSV sheets and fill them in yourself.

  2. 2

    Map it before writing

    One row per thing they will compare on, recording how you know each: stated in the brief, said on the call, inferred, or asked for by nobody.

  3. 3

    Price the conditions

    Every day names the assumption it rests on. Every exclusion becomes a zero line, so the boundary lands while the buyer is working through the money.

  4. 4

    Cut what maps to nothing

    Any section answering no criterion comes out, including your strongest material. Then check the page budget against the weights before it goes.

Frequently asked questions

Is this template free?

Yes. The zip is Word documents and CSV sheets, no account and no card. Edit with AI is the optional half: send the brief and your call notes and the criteria map comes back built, with the thin rows flagged. The rest sit in the template library.

What format are the downloaded files?

Word (.docx) for the five documents and CSV (.csv) for the three sheets, zipped together. Excel, Numbers and Google Sheets open the criteria map, the pricing build and the deliverable schedule straight off the download with no conversion step.

The client never told me what they are comparing on. Now what?

Ask. Four questions, ten minutes: what are you comparing bids on, which of those matters most, how does price sit against the rest, and who else is reading this. Asking feels like admitting you do not know, which is why almost nobody does, which is why it works.

How do I show method without doing the work for free?

You show the machine rather than the output. The export you use and why that one, the divergence that is the finding, the threshold at which a theme is reportable, the rule for when sampling stops. Every one reads as years of practice and none names a problem in their operation.

I have no client references. Does that end the bid?

No. Federal evaluation rules score an offeror with no relevant past performance record neutral rather than unfavourably, and they count experience held by key personnel. So the years you spent doing this inside a firm are offerable, labelled as an individual's experience rather than as your practice's contract.

How is this different from a proposal generator?

A generator turns a form into eight headings. This builds the map first: what they will compare on, how you know each one, and what to cut. The proposal is then written to that map, and the deliverables it promises are ones a scored diagnostic genuinely produces.

What happens after they say yes?

The priced scope becomes a statement of work. The SOW and contract pack gives every deliverable an acceptance basis two people who disagree would score the same way, and turns each priced assumption into a mechanism rather than an argument. Then the intake pack turns the client's shared drive into an index and a gap list.

Bid on what they compare

Take the Word documents and CSV sheets blank, or send River the brief and your call notes and get the criteria map back with the weights and the cuts already marked.

Edit with AI