Agency vetting scorecard
By Joel Andren · Published by PressFriendly, a PR agency · Reviewed August 21, 2026 · Editorial standards · 3 of 21 in this section
Score every finalist against the same criteria, anchors, evidence, scope, and cost assumptions. Use pass-fail gates for legal, security, ethics, rights, and conflicts; score only candidates that remain.
Freeze criteria, importance, and authority before evaluation
Assign each criterion an importance of 1, 2, or 3 before review. Use 3 only when weakness in that area would threaten the objective. Keep importance values for every candidate and record approved changes before scoring resumes.
Name the decision owner and scorecard custodian. The program owner evaluates strategy, team, work evidence, and operating fit. Finance owns the cost assumptions. Legal, security, privacy, and procurement owners set gates and hold veto authority within their domains. Evaluators should score independently first, then reconcile differences against evidence.
Fixed anchors keep ratings comparable
Use one 0–3 scale for every scored criterion:
- 0: Missing. The response omits the criterion or conflicts with the brief.
- 1: Asserted. The candidate makes a claim without usable evidence, named ownership, or workable plan.
- 2: Supported. Relevant evidence, owners, constraints, and a credible plan support the claim.
- 3: Demonstrated. Verifiable evidence shows the named team has done comparable work and can explain limits, learning, and adaptation.
Multiply rating by importance and add the rows. Keep whole numbers. The total organizes evidence; it is not a probability of success, and a small point difference should not decide the selection by itself.
Each row needs a defined evidence test
| Criterion | Evidence to review |
|---|---|
| Objective and audience fit | Diagnosis, audience rationale, assumptions, exclusions, and proposed decisions |
| Delivery team and capacity | Named roles, actual involvement, availability, review, backup, and escalation |
| Relevant work evidence | Samples, exact contribution, starting point, period, result level, and causal limits |
| Execution and learning | Work plan, quality controls, transparent activity, feedback, and adaptation |
| Measurement | Objective, baseline, activities, outputs, audience response, outcomes, impact, and method limits |
| Scope and company effort | Deliverables, dependencies, approvers, spokesperson time, data access, and service levels |
| Normalized commercial cost | Fees, setup, expenses, tools, paid distribution, overage, taxes, term, exit, and transition |
| Rights, accounts, and data | Ownership or license, company account control, exports, retention, deletion, and handoff |
| Ethics, conflicts, and security | Disclosures, competing clients, subcontractors, access safeguards, and incident duties |
| Working relationship | Responsiveness, clarity, constructive challenge, and access to the delivery team |
Attach the proposal section, sample, report, interview note, reference, or contract term that supports each rating. Record “unknown” when evidence is missing; do not turn an omitted cost or requirement into a zero-dollar assumption.
Normalize cost and company effort before comparing totals
Price the same scope, period, launch or quiet scenario, and exit case. Convert currencies and billing periods using one documented date and assumption. Include internal effort by role and frequency, such as approvals, expert access, spokesperson time, source data, and asset production. A lower fee can require more company capacity.
If proposals define different scopes, create a like-for-like comparison before scoring cost. Keep optional work separate so a bundled proposal does not receive credit for services the company does not need.
Hard gates stay outside the score
Define disqualifiers before selection. Common gates cover material misrepresentation, undisclosed paid placement presented as independent coverage, unresolved direct conflicts, unacceptable security or privacy terms, missing rights or export access, prohibited subcontracting, and exit terms outside company policy.
The PRSA Code of Ethics is a voluntary standard for PRSA members covering accuracy, disclosure, conflicts, confidences, and fair dealing. Company counsel, security, privacy, and procurement owners should set the binding requirements.
Use references to test delivery behavior and sample reports to test measurement. Document unresolved questions, veto decisions, the final rationale, and the evidence that will trigger an early engagement review.