A founder I worked with this spring had tried to automate her whole week three times. Three times she had built out a sprawling stack of agents, prompts, and zaps that were going to give her the working week back. Three times the stack collapsed inside a fortnight, and each time she lost more time to debugging than the automations had ever saved her. By the time she got to me she had concluded AI did not work for solo operators. What had broken was not AI. It was the count.
This is what I run on my own desk every Monday morning. Five workflows, no more. Each one earns its slot, each one is reviewed every week, and the whole stack gets ten minutes on Friday afternoon to keep it honest. None of them is clever in isolation. The discipline is in keeping the count to five and protecting that constraint when you are tempted to add a sixth.
What is a standing Monday workflow, exactly?
A standing Monday workflow is a piece of work that runs every Monday morning, takes inputs from the prior week, and produces an output the founder reviews and acts on before the rest of the week starts. It is not an automation that fires silently in the background. It runs to a schedule the founder owns, and the founder is the last reviewer before anything ships. That review step is what keeps it useful.
The reason Monday is the right anchor is practical, not symbolic. Cognitive attention is highest at the start of the working week, the calendar is mostly empty before meetings stack up, and if a workflow has broken over the weekend you have four days to recover before it causes real damage. A workflow that fires on Thursday afternoon when something has degraded will land in your week at the worst possible moment.
Why does it matter for a founder running a services business?
Founders running services businesses pay the highest possible tax on context switching, because the highest-value output is itself knowledge work. Each Monday spent batching the reactive load into one window buys back the rest of the week for deep work, which is the only kind of work that actually moves a services firm forward.
The Asana Anatomy of Work Index 2024 found founders spend 54 percent of working time on administrative and reactive work and only 18 percent on actual productive output. The Microsoft Work Trend Index 2024 puts the typical founder at around forty interruptions per day, with each one costing 23 minutes of recovery. Inbox, calendar, and client comms are the dominant culprits.
Standing Monday workflows attack exactly those three sources. They do not eliminate interruptions, they batch them into a window the founder controls, and they leave the four days afterwards genuinely free for the work only the founder can do. The Federation of Small Businesses 2025 quarterly survey reports UK SME founders working 48-hour weeks with 41 percent on administration, rising to 53 percent for solo operators. The recovery target is to drag that down to roughly 35 percent, which gives back around eight hours a week.
Where will you actually meet the five workflows?
You meet the five every Monday morning in roughly this order. Each one runs against a defined input set, produces a defined output, and ends with a founder review before anything is acted on. They take between two and a half and three and a half hours together once tuned, which is the price of admission for a much cleaner Tuesday through Friday.
Workflow one is the weekly review summary. Inputs are last week’s calendar, completed tasks, financial actuals, and any captured notes. Output is a two-page document covering completed work, slipped commitments, emerging issues, and one or two strategic observations. Workflow two is inbox triage. Every unread message from the prior week is sorted into immediate, important, routine, and noise, with the second category batched into Wednesday and Thursday review blocks rather than scattered through every day.
Workflow three is client update drafts, one per active engagement, generated from project data and last week’s meeting notes. The founder reviews, personalises, and sends. Workflow four is deep-work calendar protection. The system looks at the coming week and proposes calendar restructuring: consolidated focus blocks, buffers between meetings, and a few meeting times shifted to align with the founder’s cognitive peak. Workflow five is the standing-prompt library refresh, a meta-workflow that keeps the prompts behind the other four aligned with current business reality.
When to add a workflow, and when to leave it alone
The temptation is always to add a sixth. The right time to add one is when you are willing to retire one of the existing five. The wrong time is when you have just discovered a new tool. Five is not an arbitrary number, it is the upper bound at which a solo operator can keep the whole stack honest without spending more time on maintenance than the workflows are saving.
McKeown’s Essentialism research suggests human cognitive systems track between four and seven simultaneous priority threads before performance on all threads degrades. A 2024 Zapier analysis of around 4,000 users found the first automation saves about three hours a month, but by the tenth the figure has dropped to under an hour because maintenance overhead is eating the gains.
Wade Foster, the Zapier CEO, described his own discipline in a Lenny Rachitsky interview: forty-five minutes on Monday morning reviewing which automations are still firing, and killing anything that has not run in three weeks or that fails regularly. The ten-minute Friday sweep on my desk is the same idea, scoped down for solo operation. Friday is the right time for it because the week’s failures are still fresh and the founder has not yet emotionally checked out into the weekend.
Related concepts worth knowing about
The five-workflow Monday discipline sits inside a small constellation of older productivity ideas, each one load-bearing. The weekly review comes from David Allen’s Getting Things Done. Deep-work calendar protection comes from Cal Newport. The library refresh comes from Tiago Forte’s Building a Second Brain. The five-count constraint comes from Greg McKeown’s Essentialism. Together they form the spine of the discipline.
The broader pattern of disciplined AI adoption beating sprawling adoption shows up in McKinsey’s 2025 State of AI report, which found that “targeted architecture” deployments held 12 to 18 percent productivity gains over six months while “spray and pray” adoption flatlined after three. The NBER 2024 founder study reports a 23 percent revenue uplift for founders running AI on a structured weekly review, after controlling for size, sector, and tenure.
If you want to go deeper on any one of the five, the weekly review summary, the inbox triage workflow, and the standing-prompt library refresh each have a dedicated post. The framework that ties them together is the Eat-Automate-Delegate-Outsource cut for AI, which is what tells you whether a piece of work belongs in the standing-Monday set in the first place. If yours has fewer than five workflows running today, that is a feature, not a gap. Add the next one when the existing ones are stable, not before.
