Define every metric before you benchmark it
A number is only comparable when its numerator, denominator, time window, exclusions, list source, dialing mode, and campaign goal are the same.
Answered contacts ÷ attempted contactsDefine whether voicemail, wrong party, and automated answers count.
Qualified conversations ÷ answered contactsWrite a minimum duration or disposition rule so classification stays consistent.
Target outcomes ÷ qualified conversationsName the outcome: appointment, sale, transfer, pledge, renewal, or another goal.
Talk + wrap time ÷ logged-in available timeReview wait time, burnout, and quality alongside occupancy.
Calls disconnected before agent connection ÷ live answersUse the legal definition and threshold that applies to the campaign.
Completed due follow-ups ÷ follow-ups dueMeasure whether outcomes become completed next actions.
Segment before drawing conclusions
List source and age
New inquiries, aged data, customers, and first-party referrals should not share one baseline.
Campaign and dialing mode
Separate inbound, predictive, power, progressive, manual, and callback workflows.
Time and staffing context
Record time zone, calling window, agents, carrier, number pool, season, and process changes.
A repeatable weekly review
Validate the denominator
Confirm attempts, live answers, conversations, and due follow-ups use the agreed definitions.
Compare like with like
Review changes inside the same campaign segment before comparing agents or external ranges.
Inspect the workflow
Pair rates with recordings, dispositions, wait time, follow-up completion, and list quality.
Document the decision
Record what changed, who owns it, the expected effect, and the next review date.