The Loop.
One desk holds what a studio splits across a relay.
Work rarely fails at the desk. It fails in the gaps between desks, where a brief passes from one specialist to the next and loses a little every time it changes hands. A studio runs it as a relay of people. I run it as one continuous system, every thread held in one place, so nothing sits waiting on a hand-off that never comes. The loss lives in the seams; I run it without them.
A studio is a relay of specialists, and a brief loses a little at every hand.
Call it a studio, a department, a marketing team. Under the name it is a relay: a brief starts on one desk, gets carried to the next, then the next, each pair of hands doing its part and passing the rest on. It works, mostly. But every pass is a translation, and every translation drops something. Across the field almost half of projects come in late and four in ten come in over budget, and when a project misses its original goals the cause is rarely the work itself. Nearly half the time it is the brief degrading as it moves between hands, arriving at the next desk a little wrong. The desks are fine. The seams between them are where the day goes.
Field: 48% of projects finish late, 43% over budget (PMI, Pulse of the Profession 2018). 47% of failed projects trace to inaccurate requirements management, the brief degrading between hands (PMI, Requirements Management).
The desks do the work. The seams charge for it.
Watch one brief travel the relay: it is scoped, then strategized, then made, then built, then published, and it changes hands at every arrow. Each hand-off is a stop and a restart. The person picking it up has to reload the whole context the last person already held, and the tax on that reload is steep and measured. Meanwhile the target keeps drifting under the work, because a brief left open long enough always does. Add the reload tax to the drift and you have the real cost of the way it is run. None of it shows on any one desk. It shows in the gaps.
Relay proportions are illustrative. The four seam figures are independent, separately-sourced field measures of different hand-off failure modes, each on its own denominator, not a compounding funnel: requirements 47% of failed projects (PMI, Requirements Management); scope creep 52% and late delivery 48% of all projects, waste 9.9¢ per dollar (PMI, Pulse of the Profession 2018); context-switching up to 40% of productive time, directional (APA, citing Meyer et al. 2001).
The stage proportions are illustrative; the seam figures are real field data. Context-switch is an “as much as” estimate, not a measured median (APA, on the switching cost first quantified by Rubinstein, Meyer & Evans, 2001). Scope creep and requirements loss are survey-measured (PMI, Pulse of the Profession 2018).
The field runs it as a relay. Here it runs as one loop.
The geometry is the whole argument. A department splits the work across four to six specialists, and between every two of them sits a seam where a day and a little fidelity can go. Six desks, five seams. One loop has no seam to lose them at. The work does not stop and restart between hands because there is one pair of hands, holding the brief from the first line to the thing that ships. Same functions, fewer joins. The gain is simply the joins you removed.
Every thread held live at once, and none of them dropped.
This is the part that reads as chaos and runs as order: a dozen threads open at the same time, all moving, none waiting on someone else to pick them up. The density is the proof. A relay can only hold a few things live before the hand-offs pile up and the whole queue slows; a single loop can hold the lot because nothing is ever mid-pass. There is a law under this, and it is not an opinion.
In a stable system, the time each thing takes to finish equals the work in progress divided by the throughput. Hold the pace steady and pile on more open threads, and every thread takes longer to get through. Fewer things in flight at once finish sooner. That is the mathematics behind holding the board tight instead of letting it sprawl.
Threads are illustrative of concurrency, not a live task list. The instrument is Little’s Law: in a stable system, average cycle time equals work-in-progress divided by throughput, so half-open work in a relay lengthens every item’s cycle while finishing keeps it short (Little, Operations Research, 1961). Stated as a principle, with no attached figures.
The functions a marketing department normally splits across four to six people, held from one desk.
Caught early it stays small. Left alone it compounds.
The thing about drift is that it never holds still. A thread half-closed today is a small fire next month and a lost month by spring, because the misses feed each other. On a relay the small stuff hides in the seams until it is big. Held in one place it gets caught while it is still small, and the payoff is the pile-up you never see, the bad quarter that never happened. The teams that run this way are the ones that keep the target from moving under them; the field’s own numbers say the discipline is what holds the line.
Curves are illustrative of the compounding shape, not a measured series. The figures anchoring them are real field data: 9.9¢ lost per dollar to poor project performance and scope creep on 52% of projects, markedly lower among higher-performing operations (PMI, Pulse of the Profession 2018); 63% report better management of changing priorities under an agile discipline (State of Agile Marketing, via MarketingCharts); adoption reached 51% of teams in 2021 (AgileSherpas + Forrester).
The system is not a trick, and there is nothing secret in it. The discipline is the edge, and it compounds.
The work you see got run this way.
Every other case on this site sits on top of this one. One system holding all of it, so nothing loses a day in the gaps between the desks. It is the part you never see, and it is the part that makes the rest hold.
No pitch. No deck. Just a straight read on what you’ve got.