WORKED SOLUTIONS — ATTEMPT YOUR OWN ANSWER FIRST
Fictional teaching case; valid reasoned alternative plans are possible.

LESSON 1: Turn a vague request into a decision brief
  Worked brief:  Audience: operations manager, with the service lead reviewing definitions. Decision: whether to continue the limited pilot and seek stronger evaluation before wider rollout. Question: how did recorded first-response timeliness differ between tickets created on 1–7 and 8–14 June, overall and by channel? Scope: 24 fictional tickets in the supplied general/billing extract, observed to 22 June 2026 00:00 UTC; no customer-level conclusions. Deliverable: a one-page evidence note, cohort/channel table, metric contract and reproducible analysis plan. Limitation: concurrent channel-entry and staffing changes, small teaching data and no assignment/control record prevent an attributable routing-effect estimate. A proposal is to request exposure and staffing records; it is not confirmation that those data exist. Approval of expansion remains a management decision.   Five useful questions:  Which queues were exposed (defines treatment population)? What is the approved response target (defines numerator)? Does the timestamp exclude bots (defines measurement)? What changed in staffing (identifies alternatives)? What evidence would justify expansion or stopping (defines decision requirements)? “Prove routing worked” is the leading phrase. An all-queue cost-effectiveness decision is unsupported without exposure, capacity, cost and stronger outcome evidence. 

LESSON 2: Define scope, grain and fair comparison windows
  Scope:  one ticket per row; key ticket_id. Week 1 is [2026-06-01 00:00, 2026-06-08 00:00) UTC and Week 2 is [2026-06-08 00:00, 2026-06-15 00:00) UTC. Snapshot is 2026-06-22 00:00 UTC. Retain all 24 supplied rows: 12 per cohort. Include general/billing and email/chat only as represented; no unsupported exclusions. Out of scope: causal routing effect, unique-customer outcomes and productivity/cost return.  The 7 June arrival belongs to Week 1 despite its resolution date. The hypothetical 14 June 23:00 arrival has seven days and one hour of follow-up, so it is mature for 72 hours. A ticket arriving two hours before a live snapshot is immature for that outcome and must be excluded from the mature-outcome denominator with a reported count, not marked a failure. Filtering to resolved tickets removes legitimate mature non-resolutions, making a success rate conditional on already succeeding. These hypothetical boundary tickets are reasoning examples, not extra rows to append to the source. 

LESSON 3: Specify metrics, denominators and guardrails
  Primary:  eligible tickets are the specified creation cohort; observed denominator requires valid first-human-response timestamps. Numerator is observed elapsed duration ≤4 hours and ≥0. Report coverage; investigate missing/invalid responses rather than silently treating them as failures. Observed counts are 12/12 per week. Rates are 8/12 = 66.6667% and 10/12 = 83.3333%; difference = 16.6667 percentage points; relative change = 25%.   Segments:  Week 1 email 4/8 and chat 4/4; Week 2 email 2/4 and chat 8/8. Counts reconcile 8+4=12 and 4+8=12. With fixed weights 2/3 email and 1/3 chat, both rates are (2/3 × 0.5) + (1/3 × 1) = 2/3. Do not average the channel rates without weights.   Guardrail:  mature eligible tickets with first resolution recorded within 72 hours divided by all mature eligible tickets, including unresolved: 9/12 = 75% each.  Satisfaction:  sum of valid ratings / observed ratings = 32/8 = 4.0 each; coverage 8/12. A blank rating is not zero and the result describes respondents only. A zero denominator yields unavailable, with count zero, not 0%. 

LESSON 4: Request the evidence and plan quality checks
  Inventory:  cohort → created_at_utc; timeliness → creation/first response; guardrail → creation/first resolution plus snapshot; mix → channel; satisfaction context → optional rating and coverage. Routing exposure, staffing hours and longer baseline are unavailable requests, not imputed columns.   Checks/actions:  unique nonblank key → investigate duplicates; parsed UTC dates → quarantine malformed rows; creation in window → count exclusions; nonnegative response/resolution durations → investigate invalid events; events ≤ snapshot and mature eligibility → separate unobserved/immature outcomes; categories in dictionary and ratings 1–5 → flag unexpected values. For every action keep source rows and record the reason.  Six blank resolutions are mature tickets with no first resolution recorded; retain them in the guardrail denominator. Eight blank ratings are missing observations; retain tickets for other metrics and report respondent coverage. No row deletion follows automatically. Request pseudonymous ticket events, channel/queue, relevant exposure and staffing aggregates via approved access, not names or conversations by default. A bounded descriptive cohort/channel comparison can proceed; an attributable routing-effect or productivity estimate cannot. 

LESSON 5: Write and review a defensible analysis plan
  Plan:  confirm scope/definitions; validate source keys/timestamps/categories; assign creation cohorts and mature outcomes; compute observed timeliness, channel rates/weights and resolution guardrail; reconcile, review and write the evidence note. Acceptance requires 24 rows → 12+12, observed response coverage 12/12 each, channel counts/timely counts reconciling, correct percentage-point labels and explicit uncertainty.   Worked handover:  The supplied fictional extract shows recorded four-hour first-response timeliness rising from 8/12 (66.67%) to 10/12 (83.33%), a 16.67 percentage-point difference. This does not establish that routing caused an improvement. Email timeliness is 50% in each cohort and chat timeliness is 100% in each, while chat's share rises from 4/12 to 8/12. The 72-hour first-resolution guardrail is 9/12 (75%) in both weeks. All tickets have recorded first responses and sufficient follow-up, but the sample is small and deliberately constructed. Concurrent channel-entry and rota changes, absent exposure records and the lack of a control prevent an attributable effect estimate. I propose reviewing the pilot scope with the service lead and requesting ticket-level routing exposure, staffing coverage and a longer consistently recorded comparison period before an expansion decision. Confirm the service target and rollout criteria with the decision owner; this learning plan supplies no actual operational approval.  The fixed-weight result is 66.67% in each period. It answers an arithmetic composition question; it does not observe what the same tickets would have experienced under another routing process. See the separate worked plan for the complete metric contract, inventory and ten-point rubric. 