A counterweight to the predominantly upbeat case I've made for the four-day week elsewhere on this site. The model works, three years in. The first six months were genuinely difficult, in ways that rarely show up in published case studies.
This is the operator-honest version. Six things that broke, what we did about each, and what I'd have done differently if I were starting again.
1. We underestimated the capacity loss in the first quarter
The headline argument for a four-day week is that capacity loss is smaller than 20% because output goes up. That's true at the asymptote. It wasn't true in our first quarter.
The first three months, we lost roughly 15% of practical capacity. Not 20% — output did improve — but materially more than the 0-5% loss the published advocacy suggested. The team was still figuring out how to compress focused work into a four-day week. Meetings were still happening at the rate they had been. Client communication patterns hadn't adjusted. The improvements that the model produces over time were still ahead of us.
The result: we were noticeably tighter on delivery than we wanted to be, for about a quarter. Some projects ran later than they would have. Some clients were politely confused about why their usual responsiveness had changed.
What I'd have done differently: Communicated the transition arc more clearly to clients. Prepared the team for a quarter of compression before the steady-state benefits kicked in. The transition is a real cost. Pretending it isn't makes the first quarter worse, not better.
2. Senior team load increased before it decreased
The four-day week protects everyone's time, in theory. (The full operating model needs to be in place for the protection to be real.) In practice, the senior team's load went up first.
The reason: junior people noticed the time pressure of the model and naturally responded by escalating decisions to senior people more often. The senior team — which had been carrying decision authority for years — found itself making more decisions per week than it had under the five-day model, in a smaller window.
We dealt with this by explicit redelegation. Senior team members had to actively push decisions back down the org and resist the temptation to just answer the question. This was uncomfortable for both sides for a while. The junior layer had to learn to make calls without verification. The senior layer had to learn to let imperfect calls through that they would have caught and corrected in the previous model.
What I'd have done differently: Started the redelegation work months before the four-day week launched, not as a response to the strain after it launched. The cultural shift toward more autonomous junior decision-making is the bigger operational change. The four-day week is just what forces it to happen.
3. Client communication didn't adjust automatically
We told clients during onboarding that we operated four days a week. We told existing clients at the transition. We assumed that, having communicated the model, communication would adjust to it.
It didn't. For roughly four months, clients continued to send us emails late Thursday afternoon expecting a Friday morning response. They continued to send urgent requests on Friday assuming we were available. They were politely surprised when responses came on Monday morning.
This wasn't bad faith. It was that the communication norms with each client had been established over years. A single message at the transition wasn't enough to change the pattern. We had to actively re-communicate, set new expectations in each interaction, and be willing to firmly but politely hold the line for a while.
What I'd have done differently: A more explicit client-side ritual at the transition. A new section in client onboarding documents about availability. A scripted pattern for responding to Friday requests that re-set the expectation each time it came up. Treating it as a behaviour change campaign, not as an FYI.
4. Meetings re-emerged faster than we expected
The discipline of "fewer meetings" works when everyone is paying attention. Our attention was spread thin in the first six months across too many simultaneous changes — new operating model, new client communication patterns, junior layer re-delegation, fewer hours per week. Meetings re-emerged because nobody was actively defending against them.
Three months in, we noticed the calendar had gradually filled back up. Some of the meetings were genuinely necessary. Most weren't. The team had drifted back toward the comfortable habit of putting things in calendars.
We did a calendar audit and killed about 30% of the standing meetings in a single review. We instituted the rule that any new standing meeting required justification at the point of scheduling. We made meeting-free time blocks visible in everyone's calendar.
What I'd have done differently: Built the meeting discipline into the launch of the four-day week, not retrofitted it three months later. The meeting culture that the four-day week implies has to be set explicitly — it doesn't emerge from the time constraint alone.
5. Some senior people genuinely struggled
The model attracts senior, output-oriented, self-directed people. Most of the team thrived. A small number of senior people, who'd built their professional identity around long hours and high availability, didn't.
The discomfort was real. The signal was that they felt less indispensable, less productive, less valuable, than they had under the five-day model — even though, by every external measure, they were doing as much or more work. The internal narrative they'd built about their professional value was tied to time spent rather than output produced.
For some, this resolved over time. They adjusted to a new self-image. For others, it didn't, and they eventually moved on to roles in five-day-week organisations where their preferred operating mode fitted better.
This wasn't a bad outcome for either side. The people who left went to environments better suited to them. We retained and developed the people whose preferred operating mode fitted our model. But the people who left were senior, often well-liked, and the departures were emotionally significant in a way the operating-model conversation hadn't really anticipated.
What I'd have done differently: Been more explicit at the start about the kind of operating mode we were moving to, and given the team the chance to opt into or out of it before the change rather than discovering their fit after.
6. The financial picture was uncomfortable in year one
Revenue per head dropped slightly in year one. Revenue per hour worked went up but the gross revenue line wasn't materially better than under the previous model. Cost per head — including the operational investments we made to support the new model — was meaningfully higher.
The combination produced a year where margins were tighter than they had been the year before, with a clearer path to better margins in years two and three but not yet evident in the financial statements.
For a leadership team confident in the strategic thesis, this was tolerable. For a less-certain leadership team, or for an agency where the financials were already tight, this would have been a difficult year to defend to a board or to investors.
What I'd have done differently: Built a more explicit year-one financial buffer into the planning. The transition costs are real. They show up as an operating-margin compression that lasts about a year. Plan for it explicitly rather than discover it on the way through.
Closing
I am, on the whole, glad we made the change. The model now is operationally sustainable, financially better than the previous one, and produces a different and stronger team. Three years in, it's not really a "four-day week experiment" anymore — it's just how we operate.
But the case for it isn't that the transition is easy. The transition is hard, and the things that go wrong in the first six months are predictable — and most of them are the same reasons four-day-week experiments fail outright. The agencies and operators considering the move should plan for the transition costs explicitly, not assume them away on the basis of optimistic case studies.
The book on the four-day week mostly tells the steady-state story. The starting story is more uncomfortable. Worth knowing about before you commit.