The Four-Day Week Pilot
You are on a 12-person customer support team evaluating a proposal to adopt a four-day week permanently. The pilot ran for eight weeks in November and December. Participation was voluntary; 9 of 12 staff took part. Tickets closed rose from an average of 420 per week before the pilot to 445 during it. Sick days fell from 18 to 11. Overtime spending rose 18%, and two enterprise clients complained about Friday coverage. The proposal's author writes: "Productivity increased by 6%, so a four-day week will work for the whole company." The company has 300 staff in three time zones. Your director asks for a one-page recommendation. The pilot overlapped with a seasonal dip in ticket volume; the previous year's November-December average was 390.
1. Evaluate the argument's strength: identify the claim, the evidence, and at least three confounders or logical gaps.
Read the worked answer
Claim: productivity rose 6%, so a four-day week will work company-wide. This is both a causal claim and a generalisation from a small, self-selected, single-team pilot.
Evidence: tickets closed rose from 420 to 445 per week; sick days fell from 18 to 11.
Gaps: (1) Selection bias. The 9 volunteers may be the most efficient or most interested staff; the 3 who opted out are excluded. (2) Seasonality. November–December is a seasonal dip, and the prior-year average of 390 suggests volume, not productivity, may explain the change. (3) Measurement. "Tickets closed" can rise if the ticket mix is simpler or a backlog is cleared, not if output per paid hour rises. (4) Cost shifting. Overtime rose 18%, so some output was bought with extra hours. (5) Coverage. Two enterprise clients complained, a cost the productivity metric ignores. (6) Sample size and duration are too small for a 300-person, three-time-zone rollout.
2. Write a counterargument that a reasonable opponent could make, and respond to it without dismissing it.
Read the worked answer
Opponent: "The pilot is too weak to justify a permanent change. A 6% gain in one team during a low season, paid for with 18% more overtime, is not evidence of productivity."
Response: the opponent is right that the pilot cannot support a company-wide mandate. But the pilot does justify a controlled extension: the direction of the result is positive, sick days fell, and the coverage complaints identify a design problem - Friday staffing - rather than a fatal flaw. The reasonable conclusion is "test further with controls," not "reject." Conceding the valid part narrows the dispute to design: what would a credible second pilot need to measure? That is a stronger position than defending the original 6% claim.
3. Recommend a decision rule for the director: under what conditions would a permanent rollout be justified?
Read the worked answer
Approve a permanent rollout only if a second pilot (a) includes the whole team, not volunteers, or reports volunteers and non-volunteers separately; (b) runs at least 12 weeks across a normal and a peak season; (c) measures output per paid hour and overtime-adjusted cost, not raw tickets; (d) meets written coverage service levels for enterprise clients; and (e) shows no net rise in overtime above an agreed threshold. If those conditions are met, phase in by team rather than company-wide, with a review at 90 days. The rule is falsifiable: it specifies in advance what evidence would count against the change, which protects the director from adopting a popular idea on weak data.