Build a vendor evaluation scorecard that stops the loudest stakeholder from picking the vendor

Build a vendor evaluation scorecard for the options below, structured so the winning vendor is decided by the scorecard, not by whoever argues most persuasively in the room.

VENDOR OPTIONS
VendorA, VendorB, and the current provider VendorC.

MUST-HAVE REQUIREMENTS
Must support SSO integration and must have SOC 2 Type II certification already in place, not in progress.

BUDGET CEILING
$45,000/year all-in, no exceptions without separate executive approval.

WEIGHTING PRIORITY
Reliability and uptime matter most, followed by ease of integration with our existing stack, then price, then support responsiveness.

INCUMBENT VENDOR, IF ANY
VendorC is the incumbent; switching would require about 3 weeks of migration work and retraining 15 staff.

Start by applying the must-have requirements as a hard filter, before any scoring happens — any vendor that fails a genuine must-have gets eliminated outright and does not proceed to the weighted comparison, regardless of how strong it looks elsewhere, since letting a must-have failure get outweighed by other strengths defeats the purpose of calling it a must-have. For the vendors that pass the filter, build scoring criteria directly from the weighting priority given, and assign weights that reflect the stated priority explicitly rather than defaulting to equal weighting across categories, which quietly hides the actual priority instead of expressing it. Score each surviving vendor against the criteria using only the information given — where the input doesn't actually support a confident score on some criterion, mark it as insufficient information rather than guessing a plausible-looking number, since a fabricated score dressed up in a clean table looks more authoritative than the actual state of the evidence justifies. If there's an incumbent vendor, do not give it implicit credit for familiarity or switching-cost-avoidance unless that was named as an actual weighted criterion — an incumbent should win or lose on the same criteria as everyone else, with switching cost scored explicitly as its own line item if it matters, not folded invisibly into every other score as a thumb on the scale.

WHAT NOT TO DO
Do not produce a scorecard where every vendor scores similarly across the board — if the underlying evidence actually differentiates them, the scores should show it; a scorecard that hedges into near-ties on every line isn't doing the job of a scorecard. Do not let budget ceiling function as just another weighted criterion if any vendor actually exceeds it — a hard budget ceiling is a filter, like the must-haves, not a score.

OUTPUT FORMAT
1. Must-have filter results: which vendors are eliminated and on which specific requirement.
2. Budget filter results: which vendors exceed the ceiling.
3. Weighted scorecard table for surviving vendors: criteria, weight, score per vendor, with 'insufficient information' marked explicitly where relevant.
4. Final ranked recommendation with the specific reasoning, including how the incumbent (if any) fared on switching cost as an explicit line.