The financial case for a four-day week is almost always made in directional terms. "Output goes up." "Attrition drops." "Hiring is easier." All true, none of them numbers.
This post tries to be more specific. It's not a P&L breakdown — those are confidential — but it's the directional shape of the metrics that matter, year one to year three. The pattern is consistent enough across our portfolio that the broad conclusions should generalise to similar service businesses, even if the specific numbers won't match.
A health warning: these are rough metrics from a small-to-mid agency operating in a specific market. Don't take any single number as a target. Take the shape.
Revenue per head
The headline metric. How much revenue does each person on the team generate annually.
Year one (transition year): Roughly flat. The capacity loss in the first quarter offset the productivity gains for most of year one. Some clients trimmed retainers slightly during the transition uncertainty. New business closed at a normal rate.
Year two: Up moderately. The team's output had stabilised in the new model. Senior time was spending more time on revenue-generating work and less on internal admin. Hiring had started to pull in stronger candidates.
Year three: Up substantially. Compounding benefits — the team is now consistently delivering more per person than it did under the five-day model, the senior layer is more productive than it was, and the agentic systems we'd built were absorbing meaningful chunks of low-value work.
The directional shape: a small dip or flat in year one, then rising for years two and three. Worth being clear that this is the shape we observed; another agency might not show the same recovery curve, particularly if the operational investments aren't made.
Gross margin
Margin per project, blended across the portfolio.
Year one: Down a few percentage points. The transition costs hit the cost line — operational redesign, some additional hiring to manage the new operating model, some software and tooling to support async-first work. Revenue per project was broadly stable but the cost of producing each project was temporarily higher.
Year two: Recovered to roughly the previous baseline. The transition costs were behind us. The operational efficiencies were starting to land.
Year three: Better than the original baseline. The agentic systems had absorbed real workflow tax, the senior layer was operating more leveraged, and overall project costs were lower than they had been in the five-day model.
This is the metric where the year-one pain is most visible. An agency with tight starting margins should be honest with itself — see the operator-honest version of what broke for us in year one about whether it can absorb a year of margin compression.
Attrition
The percentage of the team that leaves voluntarily over a 12-month period.
Year one: Higher than baseline. Some of this was the natural shake-out of people whose operating preferences didn't match the new model. Some was uncertainty during the transition. The leavers were mostly senior people whose professional identity was tied to traditional working patterns. The replacements, hired into the new model, fitted better.
Year two: Below baseline. The hiring filter was now consistently bringing in people who matched the new operating model. People who joined under the four-day model rarely left. People who'd survived the transition had self-selected into the model.
Year three: Significantly below baseline. Sustained low attrition, with departures mostly being for genuine career progressions rather than dissatisfaction with the role.
The shape: a year-one spike followed by a dramatic improvement. The total attrition cost over three years is lower than it would have been under the five-day model, but the year-one spike can be uncomfortable.
Cost per hire
Direct hiring costs (recruitment, onboarding) plus the indirect cost of senior time spent on hiring.
Year one: Slightly higher. We were hiring in volume to manage the year-one transition, and some of those hires didn't work out, requiring re-hiring.
Year two: Lower than baseline. Inbound applications had grown materially as the four-day week became known in the market. We were hiring less reactively — most senior roles had multiple strong candidates without active recruitment effort.
Year three: Substantially lower. Most senior hires now come through inbound interest in the model. Recruitment costs are a fraction of what they were under the previous model.
This is, financially, one of the largest single benefits over time. The savings on recruitment, both direct and the senior-time tax, are large enough on their own to offset much of the year-one transition cost.
Time to productivity for new hires
How long it takes a new senior hire to deliver at full capability.
Year one: Worse than baseline. The new operating model required new hires to learn both the agency's domain and the agency's distinctive operating culture simultaneously. The senior people doing the onboarding were stretched.
Year two: Roughly equivalent to the five-day model. The documentation we'd built during the transition was paying off. New hires had clearer artefacts to onboard against.
Year three: Better than the five-day model. The investment in async documentation has produced an onboarding stack that's substantially better than what most agencies offer. New senior hires reach full capability noticeably faster than they did under the previous model.
Cost per billable hour
Fully-loaded cost of producing one billable hour, blended across the team.
Year one: Up. The capacity loss in the first quarter increased the cost per billable hour of senior staff. We were paying for time that wasn't being deployed against client work.
Year two: Roughly stable.
Year three: Down. The combined effect of the agentic systems absorbing low-value work, the senior team being more productive, and the operating model being more efficient produced a measurable improvement in the cost per billable hour.
Client retention and pricing
Two metrics that are harder to put on a clean year-by-year curve but worth flagging.
Client retention has been stronger under the four-day model than under the five-day model. The team being calmer, the senior staff being more available (because they're less burnt out), and the work being higher quality have produced better client relationships. The clients we have, we're keeping for longer.
Pricing has held or improved at the same time. We've been able to charge slightly more for similar work, partly because the model itself is a positioning advantage in a market where clients are weary of agencies whose senior people are clearly running on fumes. The four-day week isn't a discount signal; if anything it signals operational confidence.
What this means for an agency considering the move
A few practical implications.
Year one is more financially uncomfortable than the published case studies suggest. Plan for it. Have the cash position to absorb a year of margin compression. If you're already tight, the transition is genuinely risky.
The operational investments are not optional. If you try to do the four-day week without the operational redesign — async-first defaults, meeting discipline, agentic offloading of internal workflow — the metrics that should improve in years two and three will not improve. The four-day week becomes just a productivity loss.
The hiring benefits compound but take time. Year one feels harder, not easier, on the hiring side. By year two the inbound interest has shifted noticeably. By year three the recruitment cost is meaningfully lower.
The financial case is real but back-loaded. Most of the benefit shows up in years two and three. An agency leader needs the conviction to push through year one, knowing the numbers will be tighter, on the basis of the model showing up properly later.
If your agency's operating model is already strong — strong senior team, good operating discipline, healthy margins — the transition is more comfortable. If your operating model is already stressed, the transition will stress it further before improving anything.
The model works. The economics, in steady state, are better than the alternative. The path to steady state is harder than it usually gets credit for. Both things are true.