Margin Notes
Languages / Idiomas

Read this page in your language

On this site

Other languages · Google Translate

Other languages open in Google Translate in a new tab. Google receives the public page URL and processes your visit. Machine translations may contain errors. Compare with the English original.

If translation is unavailable, try your browser’s built-in translation. Local previews link to the published site.

Themes & Ideas

The 4-Hour Workweek for Solo Developers: A One-Week Work Audit

An original, modest work audit inspired by Tim Ferriss’s discussion of priorities: measure one recurring task, change one variable, and check what happens to quality.

This is our original experiment proposal inspired by selected ideas in Timothy Ferriss’s The 4-Hour Workweek. It is not a worksheet from the book, and no experiment results are claimed.

Separate a worthwhile task from a quicker task

Chapter 5 questions activity that consumes time without producing a desired outcome; chapters 9–10 discuss testing demand before expanding a business idea. Together, these ideas suggest a question for a small website: what problem does each recurring task actually help a reader solve?

A developer can optimize an image pipeline, revise a paragraph, or refresh a report. All three are activities. Their value depends on the current problem. If an article is hard to understand, a more elaborate deployment process may not improve the reader’s experience. If the site fails to load, reliable delivery becomes urgent. The diagnosis determines the priority.

Observe before changing the routine

For one ordinary working week, record the effort spent on a small set of recurring activities. Keep personal or client information out of the notes. The point is a usable baseline, not an elaborate tracking system. A week with an outage or holiday may be unusual; label it rather than treating it as typical.

That first week is an observation period, not a deadline for proving a result. Before changing anything, choose a follow-up period with comparable working days, task types, and workload. Record the dates and task counts for both periods; extend observation or leave the result inconclusive if the comparison would be misleading. Count setup, checking, and correction time as well as routine work. Comparing minutes per similar task can help, but cannot account for a change in task difficulty.

For a reading website, distinguish correcting an error, preparing a source-backed article, checking search reports, and adjusting decoration. Record the problem each task addresses. Unknown outcomes should remain unknown: a newly published page may not yet have enough evidence to evaluate its search performance.

Use a small audit table

The following fields and examples were designed by Margin Notes. They are not adapted from the book’s exercises. The example row is hypothetical, with no invented time saving or revenue result.

Field What to record Hypothetical website example
Recurring task The action and why it exists Repeatedly opening a search report to look for a change.
Observed effort Actual time and interruptions during the baseline Record it; do not guess a number.
Intended outcome The reader or operator problem it addresses Notice a meaningful indexing or query change.
Proposed change One limited change to compare Review the ordinary report at a planned time.
Quality check What must continue to work Urgent operational alerts still reach the operator.
Decision Keep, revise, reverse, or leave inconclusive, with a reason Compare actual usefulness and missed issues.

A task should not disappear merely because its benefit is hard to measure. Corrections and maintenance can prevent harm before a complaint or failure makes the benefit visible. Consider that possibility when interpreting the table.

Download the blank work audit worksheet (TXT, English). It includes separate baseline and follow-up logs, workload and quality checks, and a decision record. Fill it in on your own device; there is no form here to submit your answers. Use your browser’s Print command to save or print this article, with its sources and limitations included.

Change one variable and keep a way back

Choose an ordinary routine under your control. Reducing repeated nonurgent checks is a modest candidate; suspending a security alert or ignoring a rights complaint is not. Define the change, its review point, and the condition that would make you reverse it.

Compare effort with what the routine was supposed to accomplish. Fewer interruptions with a missed important issue is a mixed result. A smaller change may be better than keeping the original proposal unchanged. A short observation period can reveal a problem; it cannot establish that a system will remain reliable forever.

Keep search demand separate from production volume

A website’s count of articles or translated URLs measures output. It does not tell you whether a particular reader needs the next article. Use actual search queries, editorial questions, or reader feedback when available, and distinguish them from guesses about a topic.

If data is too sparse, improve the evidence and clarity of an existing page and record the question a new page would answer. That is a concrete next step without pretending that demand has been verified. Changes to search outcomes can have several causes; do not credit a one-week work audit with every later movement in traffic.

Review the trade-off, then return to the book

The useful result is a reasoned decision about one routine, including the work it protects and the work it consumes. Any time recovered should be counted separately from income and from the value of the activity you choose next.

Our DEAL reading guide explains the broader framework. Our automation and delegation review asks which responsibilities remain when a routine is handed over.

Sources and limits

The discussion refers to chapters 5 and 9–10 of a Chinese revised-edition text of Timothy Ferriss’s book, with translation credited to He Mei. The audit, developer examples, and checks are our own proposals. They are not Ferriss’s words, reported reader successes, or a promise of financial results. Edition metadata and every historical commercial example have not been independently authenticated.

Use the publisher’s book page to find the original work and legitimate editions. These independent comments are meant to accompany reading the book, not replace it.