What an audit protocol is
An audit protocol is a fixed procedure for turning a story you tell about yourself into a record you cannot argue with.
Every one of them runs the same four steps. Pick a domain. Record what actually happened, without editing it while you record. Compare the record against the person you claim to be becoming. Change one thing, then run it again.
They work because self-report fails in a predictable direction. People overestimate the time they spend on what they value and underestimate the time they spend on what they do not. The confidence is not evidence. It is narration, and it becomes convincing enough that most people never test it against a single data point. An audit removes the narration and leaves the record.
Why there are five, and not one
Time, money, attention and trust leave different evidence, so each needs a different instrument. The protocols differ on three axes: what they measure, how long the recording window is, and how often you run them. Running the wrong one is how people conclude that auditing does not work.
The Tuesday audit measures a day. The 168-hour audit measures a week. The 72-hour rule measures a single decision while it is still being made. The raise audit measures a fork you reach only a few times in a career. And the 6 AM test measures something you are not administering at all, because your partner has been running it on you for years.
The five audit protocols compared by what they measure, their recording window, how often to run them, and the trigger for running one.| Protocol | Measures | Window | Cadence | Run it when |
|---|
| The Tuesday audit | Daily defaults | One ordinary Tuesday | Monthly | Nothing is wrong, and nothing has changed in years |
|---|
| The 168-hour audit | Time allocation | Seven consecutive days | Twice a year | You believe you have no time |
|---|
| The 72-hour rule | Emotional spending | Every non-essential purchase | Continuous | Your bank statement contradicts your stated goals |
|---|
| The raise audit | Financial trajectory | Before any of a raise is spent | At every income increase | Your income just went up |
|---|
| The 6 AM test | Relational reliability | Your unguarded default | Continuous | Your partner says they do not feel connected |
|---|
The five protocols
Protocol 1 of 5
What it measures
It measures the distance between the day you actually lived and the day the person you say you are becoming would have lived. At the end of a Tuesday you write what happened, hour by hour. Beside it you write the realistic version the other person would have run. The gap between the two columns is the number.
When to run it
Run it when you are fine. Not in a crisis, not on a Monday, not after a bad week. Tuesday is the test day precisely because it carries no fresh-week energy and no Friday relief. It shows you what your life looks like with no occasion, no motivation, and no audience.
The move it produces
Change one default per month. Not all of them. The alarm time, the commute podcast, the evening screen habit. One default, replaced with a deliberate choice, held for thirty days. Then audit again and pick the next one.
Read the full protocol
Protocol 2 of 5
What it measures
It measures where a full week actually goes. Log every thirty-minute block for seven days without labeling anything during the tracking week. At the end, color the blocks green for what moved something forward, yellow for necessary maintenance, and red for hours you cannot explain. Count the reds.
When to run it
Run it the moment you catch yourself saying you have no time. Sleep eight hours a night and 112 waking hours remain. Work forty-five including the commute and sixty-seven remain. The audit does not create time. It shows you the hours you already have and are currently spending on nothing.
The move it produces
Reclaim one hour a day from the largest red block, which is almost always the phone. Redirect it to the one thing you keep saying matters. Do that for four weeks, then run the audit again and compare the two spreadsheets.
Read the full protocol
Protocol 3 of 5
What it measures
It measures how much of your spending was ever actually a choice. Wait seventy-two hours before any non-essential purchase, and during the wait write down one thing: the emotional state you were in when the urge appeared. Most urges do not survive the delay. Over ninety days, eighty-one percent expired.
When to run it
Run it continuously, because it operates at the moment of the decision rather than after it. This is the only protocol on this page that changes an outcome while the outcome is still being decided. Everything else here is a post-mortem.
The move it produces
Stop reading the bank statement as a budget document and start reading it as a behavioral one. It records what you reached for and when. Look for the clusters: late at night, right after work, alone on a Sunday. The pattern under the totals is the finding.
Read the full protocol
Protocol 4 of 5
What it measures
It measures what an upgrade actually costs. Before spending any part of an increase, answer one question in writing: if I invest this entire increase and change nothing about my lifestyle, what does my position look like in three years? Calculate it. Then decide, with the number in front of you.
When to run it
Run it the week the raise lands, not the month after. Lifestyle inflation does not arrive as one reckless purchase. It arrives as twenty individually reasonable upgrades that each feel earned, and by the time you notice, the entire increase has already been absorbed.
The move it produces
Pair the audit with two rules. Invest at least fifty percent of any raise before upgrading anything. And wait six months before committing to any recurring cost above one hundred dollars a month, because you cannot judge an upgrade until the novelty has worn off.
Read the full protocol
Protocol 5 of 5
What it measures
It measures what you reach for before your prefrontal cortex has booted up. At six in the morning you are not managing your image, so the reach is your honest default. Trust is not built through declarations. It is a ledger of small turn-toward and turn-away moments, and it has been running for years.
When to run it
You do not run this one. It is already running, and someone else holds the record. This is the only protocol here where the auditor is another person, which is exactly why the result is harder to argue with than any spreadsheet.
The move it produces
Change the default rather than making a gesture. Put the phone on a charger in another room before sleep. Turn your body toward your partner when they speak, not just your ears. Keep the small promises, all of them, because the ledger only tracks kept or broken.
Read the full protocol
How to sequence them
Start with the 168-hour audit. It has the widest aperture and the result is the hardest to explain away, which matters because the first audit has to survive your own defensiveness. Give it seven days.
After that, the Tuesday audit becomes the maintenance version. It costs ten minutes a month and catches drift long before a full week of tracking would. Turn on the 72-hour rule and the 6 AM test at the same time, since both run continuously and neither requires a recording window. Keep the raise audit in reserve. It fires only when your income changes, and that is the only moment it can do anything.
The failure mode
Auditing has one predictable way of going wrong. Measuring feels like progress. Building the spreadsheet, choosing the colors, counting the reds: all of it produces most of the satisfaction of changing the behavior at almost none of the cost. People run three audits, learn a great deal about themselves, and change nothing.
The rule that prevents it is a hard cap. One changed default per cycle. Hold it for thirty days. Then audit again. An audit that does not end in a changed default is not an audit. It is a nicer version of the same story.
Questions about audit protocols
What is an audit protocol?
An audit protocol is a fixed procedure that replaces a story about your behavior with a record of it. You pick a domain, record what actually happened without editing it, compare the record against the person you say you are becoming, then change one default and run the whole thing again.
Which audit should you run first?
The 168-hour audit, because it is the broadest and the result is the hardest to argue with. It shows you where a full week actually goes. Once you know that, the Tuesday audit becomes the monthly maintenance version, and it costs ten minutes instead of seven days.
How often should you run an audit?
It depends on what the protocol measures. The 72-hour rule and the 6 AM test run continuously, because both operate at the moment of a decision. The Tuesday audit runs monthly. The 168-hour audit runs twice a year. The raise audit runs only when your income changes.
Why do most self-audits fail?
Because measuring feels like progress. Collecting the data produces much of the satisfaction of changing the behavior and costs almost none of the discomfort, so the spreadsheet becomes the project. The rule that prevents it is a hard cap: one changed default per cycle, held thirty days, then audit again.