Risk severity depends on your use case.
A missing audit log matters differently for a public website tool, an internal knowledge base, or a customer-data workflow.
Software vendor risk assessment
EvidenceOps weighs evidence by confidence, impact, likelihood, and decision relevance so your team can see which risks actually change the recommendation.
The hidden problem
A vendor can have twenty possible risks. Only a few usually matter for the actual decision, budget, compliance posture, or rollout plan.
A missing audit log matters differently for a public website tool, an internal knowledge base, or a customer-data workflow.
The problem is not always a bad vendor. Sometimes the problem is that no one can prove the claim yet.
A tool can be technically strong and still fail on pricing, lock-in, export quality, or internal approval.
Assessment layer
Each risk is tied to the evidence behind it and the decision it affects. This makes the assessment useful in a procurement, finance, or founder review.
Seat expansion, tier gates, renewal exposure, add-on ambiguity
Trust center quality, SOC 2 evidence, access controls
DPA availability, subprocessors, data locations, terms
Admin effort, migration cost, owner dependency, adoption friction
Export quality, data portability, API limits, replacement cost
What still needs verification before commitment
Decision use
If a risk cannot change the decision, pilot terms, vendor questions, or rollout conditions, it probably does not belong in the executive view.
Example: not 'pricing unclear', but '60-seat rollout may cross the target budget if SSO requires enterprise pricing'.
Every major risk gets a next check: vendor question, document request, pilot test, or internal acceptance decision.
The final brief explains whether to proceed, pause, reject, or proceed under conditions.
Risk model
The goal is not a fake precise score. The goal is a shared language for what matters.
EvidenceOps