RFP Evaluation Criteria, Explained

Every well-run public request for proposal publishes its scoring system with the solicitation: the pass/fail screen, the rated criteria, the weightings, and the price formula. Most losing proposals are written by firms that skimmed it. The grid is not fine print — it is the specification for the document you are about to build.

Mandatory criteria are the gate. Marked with must and shall, listed separately from scored criteria, and applied before any evaluation: fail one and the proposal is out, whatever its brilliance. The discipline is binary verification — for each mandatory, can you point to the specific certificate, year of experience, or named person in your own records? Where the answer is no, the decision is bid/no-bid, not word-smithing. The build for this is the compliance matrix.

Rated criteria and weightings are the effort budget. A grid that scores technical approach at 40%, personnel at 30%, and price at 30% is telling you that two-thirds of the decision is qualitative — and that a thin, cheap submission is out of the money before pricing ends. Allocate drafting hours to the criteria in proportion to their points, not to the sections you enjoy writing. Silence scores zero: an aspect of a criterion you did not address is a row on the evaluator's sheet left empty, not an assumed pass.

The price formula is a scoring machine, not an invoice. Common shapes: price as a fixed percentage of total points; price scored relative to the lowest compliant bid; or a best-value ratio trading technical score against cost. The formula changes your arithmetic completely. Where price is a modest slice of the grid, a technically stronger proposal can afford to be more expensive — work out how many technical points a $10,000 concession would buy before you cut your rate. Where price dominates the weighting, best value in name is lowest price in practice, and the winning skills are estimating and cost control.

Score the way evaluators score. Evaluators work from a sheet with one row per criterion, citing evidence they can defend on review. Make each row easy to fill: mirror the tender's own numbering in your section headers, answer every sub-criterion in order, and attach the proof — the project, the reference, the metric — beside the claim. The registry of bid challenges holds 2,043 documents since 1990, and the factual core of most challenges is criteria applied unevenly or changed after the fact (bulk data: pubsecdata.org); a proposal that maps visibly to the grid is also the proposal hardest to score unfairly. A drafting assistant that structures the response against the published criteria — pubsec.pro's generator is free — keeps that mapping honest while you do the thinking.

Read past the grid for the real buyer. Criteria reveal what the buyer fears — a heavy weighting on transition risk means a painful handover last time; scored Indigenous participation means policy pressure with teeth ($4.6B in flagged federal awards, 7,773 records, 2004–2026). Write to the fear as hard as to the points.

Common questions

Can evaluation criteria change after proposals are submitted?

In public procurement, no — not to rescue a preferred bidder. Criteria are fixed at closing, and material changes or uneven application are the standard grounds for a bid challenge. If an amendment changes the evaluation before closing, it applies to everyone and the clock usually moves.

What happens if I do not answer a rated criterion?

It scores zero, not average. Evaluators cannot infer capability you did not state. Every sub-criterion on the grid deserves an explicit answer with evidence, even when it feels repetitive.

Should I always bid the lowest price?

Only if the formula rewards it. Work the arithmetic first: with price weighted at 20-30%, a cheap-and-thin proposal loses to a strong technical submission routinely. Price to the formula, then to your costs.