Some architecture firms can, and a four day work week is more plausible in a design office than most principals assume. The catch is that it only works if chargeable hours per person hold steady, which means the firm has to cut the unchargeable 40% rather than ask people to draw faster. Firms on fixed fees with disciplined scope have a real shot. Firms billing hourly with a client base that calls on Fridays mostly do not.
That is the whole argument in one line: this is a utilization problem wearing a benefits costume.
What does the four-day week evidence actually show?
The most cited data is the 4 Day Week Global UK pilot of 2022: about 60 organizations and roughly 2,900 employees moved to a shorter week at full pay. More than 90% of the participating companies continued the schedule afterward. Revenue was broadly flat across the group, and self-reported burnout and turnover both fell.
Read that carefully before quoting it to your partners. The sample was mostly small, mostly service and technology businesses, and self-selected, which means the organizations that signed up already believed it could work. Almost none of them had a professional liability exposure tied to a permit clock.
What the pilot does establish is narrower and still useful: for knowledge work, output is not a linear function of hours in the chair. That matches what anyone who has produced a construction document (CD) set already knows. Hour 46 of a week is not worth what hour 12 is worth.
Does the four day work week math work in an architecture firm?
Run the numbers before the vote. Deltek Clarity 2026, the 47th edition of Deltek’s annual study of architecture and engineering firms, puts firm-wide utilization (the share of total hours that are chargeable to projects) just under 60%.
Take a project architect. PSMJ and Deltek role guidance puts that seat at 75% to 85% chargeable. At 40 hours, 80% is 32 chargeable hours. At 32 hours, holding 80% gives you 25.6, and the firm has lost roughly six billable hours a week per person. Across eight technical staff over 46 working weeks, that is a serious hole in net revenue.
Now run it the way the successful pilots ran it. If a 32-hour week comes with fewer internal meetings, one status meeting instead of three, and no all-hands on Friday, chargeable share can climb. A firm at 60% on 40 hours produces 24 chargeable hours. The same firm at 72% on 32 hours produces 23. That is the trade, and it is tight. There is no version of this where the firm keeps its current meeting habit and its current billings.
So the first question is not “should we try a four-day week.” It is “what is our real firm-wide utilization, and where do the other hours go.” If nobody at your firm can answer that from data, start with a time audit and revisit the four-day question in a quarter.
Which firms can pull it off, and which cannot?
Fit is mostly about how you get paid and who you serve.
- Better odds. Fixed-fee work with scope you control. Residential and small commercial studios. Firms whose clients are institutions with slow decision cycles. Practices with a strong project template and standard details, where the same wall section is not redrawn every time.
- Worse odds. Hourly billing with a not-to-exceed cap, where fewer hours means less revenue directly. Heavy construction administration (CA, the phase after the permit when the building actually gets built) with active sites. Developer clients who expect a same-day answer. Firms of three or four people where one person out means one discipline out.
Size cuts both ways. A five-person studio can decide this over lunch and reverse it in a week. A 40-person firm can stagger coverage so somebody answers the phone every weekday, which a five-person studio cannot.
What breaks first?
Coverage, not production. In every firm we have watched try this, the drawings got done. The site did not stop.
The contractor calls Friday morning because the sub is standing in front of a beam pocket that does not work. The plan checker’s correction letter arrives Thursday afternoon with a resubmittal deadline. A client’s board meets on Fridays. None of that reorganizes itself around your schedule.
Two patterns handle it. Either the firm closes on a fixed day and publishes it to clients and consultants six weeks ahead, with one on-call person rotating and paid for it, or people take staggered days off so the office is covered five days while each individual works four. The staggered version is easier on clients and harder on collaboration, because the whole team is only in the same room three days a week.
The third pattern, the compressed 4/10, is not a four-day week at all. It is 40 hours in four days. It solves commuting and childcare and does nothing for burnout, which is usually the actual goal.
How do you pilot it without betting the firm?
Ninety days, one measured baseline, and a written exit.
Before the pilot, record three months of the numbers you will judge it by: firm-wide utilization, chargeable hours per person per week, overtime hours, projects hitting phase budget, invoices out by the fifth of the month, and client complaints. If you cannot produce those today, that is the finding, and the pilot waits.
During the pilot, protect two things. Keep timesheets honest, because the temptation to log a 32-hour week as if it were 40 is enormous and it will poison the data. And hold scope discipline, since a shorter week magnifies every unpriced revision that used to be absorbed with a Saturday.
The retention case is real but should be argued honestly rather than assumed. Zweig Group’s 2024 Financial Performance Report put AEC staff turnover at 14.1%, up from 12.3% the year before. The American Institute of Architects (AIA) Compensation & Benefits Reports have tracked flexible hours and work-from-home options becoming more common as firms compete to keep people. Replacing a licensed project architect costs months of lost productivity plus recruiting, so a schedule change that keeps two people who would otherwise leave can pay for a few lost billable hours. That is an argument you can make to a skeptical partner. “Everybody is doing it” is not.
If the pilot fails, say so and go back. A firm that tries a four-day week, measures it, and reverses it with an explanation loses nothing. A firm that quietly lets it decay into a four-day week with Friday email loses its credibility on the next change it proposes, and it still has the underlying problem the schedule was meant to fix: too many hours going somewhere nobody is measuring, which is where most project budget overruns start.
AECPulse Project Manager reports chargeable hours by person and week against a role target, so a four-day pilot is judged on utilization data rather than on who appears busiest at 6 pm.