Measure the thing you actually observed
A useful technology evaluation defines the workflow, software version, sample, test conditions, and outcome before measuring it. Response latency, submission acceptance, delivery events, replies, and campaign outcomes are different measures. Compare the same units and conditions.
Define it in the engagement
Keep observations, failures, sample size, variation, and limitations with the report. A single demonstration cannot establish a universal performance claim. A numerical Ananasi claim needs an identifiable test and publication approval.
Discuss the actual workflow
Bring the use case to Ananasi. The proposed implementation, access responsibilities, reporting, and support commitments should be specified in the quote before the workflow receives campaign data.