How to review an RFP answer from a blank page

Review setup

Remove author assumptions without removing buyer context

Blank-page does not mean context-free. The reviewer needs the exact tender question, its numbered subparts, applicable instructions, defined terms, evaluation criterion, weighting and score descriptors. Include buyer documents expressly referenced by the question and the controlled clarification record. If page limits or a prescribed structure affect what can be said, include them too. Without that basis, the reviewer is testing personal expectations rather than the procurement.

Exclude material that an evaluator will not receive: the author’s outline, workshop notes, internal win theme, response plan, verbal explanation and history of earlier drafts. Do not let the author introduce the answer in a meeting. The first reading should happen asynchronously or in silence. A review copy should show the content as the evaluator will encounter it, including headings, tables, captions and valid cross-references, not a collection of disconnected source fragments.

Before opening the response, the reviewer makes a short answer key. It is not a model answer. It lists the decisions the buyer must be able to make and the observable material needed to make them. For an implementation question, that might include sequence, responsibilities, dependencies, acceptance and proof of feasibility. The key prevents impressive prose from changing the question and gives the later comments an explicit basis.

First read

Write down what the answer actually lets a reader conclude

Read once at a realistic evaluation pace. At the end, state the answer in a few sentences: what is being proposed, who does what, when it happens, what conditions it depends on, how success is demonstrated and what the bidder commits to. Record the page or paragraph supporting each point. If a point can be supplied only from memory, mark it absent. If two passages support incompatible conclusions, record the conflict rather than choosing the one the author probably meant.

Separate explicit statements from reasonable interpretation and unsupported inference. An evaluator can connect a defined acronym to its glossary entry. They should not have to infer that “our team” includes a named subcontractor, that “rapid recovery” equals a particular recovery time or that a case study used the proposed configuration. The more material the conclusion, the less the answer should depend on charitable reconstruction.

Test navigation as part of meaning. A correct fact buried in an unrelated annex may not answer the question if the tender restricts cross-references or the evaluator cannot locate it. Follow every citation and cross-reference. Check that the destination exists in the submitted package, uses the same term and supports the claimed proposition. The FAR requirement to assess proposals solely on stated factors is one reason bidders should make criterion-relevant evidence visible rather than expecting evaluators to rescue the case with unstated knowledge.

• Summarize the offer without consulting the author.
• Attach a visible location to each material conclusion.
• Mark interpretation separately from explicit commitment.
• Follow references as an evaluator would receive them.
• Record contradictions without privately resolving them.

Diagnosis

Name the missing evaluative link, not the sentence you dislike

Compare the reconstruction with the answer key and classify each material difference. A completeness gap leaves a subquestion unanswered. A causal gap asserts an outcome without connecting method to result. An ownership gap describes activity without accountable actors. An evidence gap makes a claim without applicable support. A consistency gap gives different commitments across sections. An orientation gap makes relevant content unreasonably hard to find. An inference gap asks the reader to import knowledge that is not in the bid.

A useful comment contains four parts: the tender basis, the observable text, the evaluative consequence and the required outcome. For example: “Criterion 2 asks how incidents are escalated. The answer names three forums but gives no trigger or decision authority. A reader cannot determine who acts on a critical incident. Add the approved triggers, authority and recorded handoff.” This gives the author more to work with than “governance unclear” and does not dictate unverified prose.

Triage comments by consequence. Material issues affect compliance, a scored subcriterion, credibility of a central claim or consistency of the offer. Minor issues affect readability without changing what the evaluator can conclude. Do not allow a long list of copy edits to obscure three structural failures. Conversely, do not label every preference material. Public procurement frameworks emphasize evaluation against disclosed criteria; the internal review should preserve the same discipline.

Closure

Keep the reviewer diagnostic and make the revision stand alone

The reviewer should not rewrite a specialist answer merely because editing is faster than explaining the gap. Replacement prose can introduce an unapproved commitment, lose technical precision and teach the author nothing about the evaluator’s decision. Assign the issue to the answer owner and the relevant fact authority. The comment specifies what must become assessable; the owner chooses accurate language and obtains any necessary solution, commercial, legal or evidence approval.

When the author returns the revision, suppress the accompanying explanation. Read the affected answer from the clean brief again and close the comment only when its required outcome is visible in the text. Material structural repairs benefit from a second reviewer who never heard the discussion. If that reader reaches the intended conclusion and can point to its support, the repair no longer depends on inherited context.

Use the method early on a representative set of high-value answers, not only at the final gate. Patterns reveal systemic problems: undefined actor labels, evidence stored outside the response, conclusions before method or inconsistent names for the same service. Feed those patterns into author guidance and templates. The aim is not to create an army of late reviewers. It is to produce answers that a buyer can understand correctly on the first independent reading.

• Return the missing decision and consequence, not polished replacement copy.
• Route factual and commitment changes to their accountable owners.
• Retest without the author’s verbal explanation.
• Use a second clean reader for material structural changes.
• Turn repeated findings into earlier author controls.