The Energy Audit, Not the Time Audit

Why this matters

A time audit tells you where your hours went. It will not tell you why a 44-hour week left you flat on Thursday and a 56-hour week two months earlier did not. Owners run the time audit, see the hours, cut the wrong things, and end up with a shorter week that feels exactly as bad. The reason is that hours are not the unit that ran out. What ran out was capacity to make decisions, and different work spends that capacity at wildly different rates per hour. This card is the instrument for measuring the rate, and the rule for what to do with each shape of result.

The unit is charge per hour of a class, not hours per class

A time audit produces a list: 14 hours on dispatch, 18 on admin, 41 in the field. Useful for pricing your own time, useless for explaining Thursday.

An energy audit produces a second number against each of those classes: how much capacity an hour of it costs or returns. Some work is genuinely restorative. Most owners find that running calls, the thing they are supposedly supposed to get away from, reads positive, and that fifteen minutes of a customer conflict reads worse than three hours of anything else. Neither of those shows up in a time audit at all.

Note the difference from finding your peak window. Peak-window work asks when you are sharp and puts hard work there. This asks what drains you regardless of when you do it. Both are real, they are different measurements, and the second one is the one that explains a week you cannot account for.

The logging method

Ten working days. Two days is noise, five days catches one bad week and calls it a pattern, ten is enough that a single awful Tuesday does not drive the result.

For each block of work longer than about 20 minutes, write three things:

• The class. Six to eight classes maximum. Field work, quoting, dispatch and firefighting, admin and paperwork, customer conflict, coaching and training, vendor and supplier, personal and errands. Fewer classes than that and everything blurs; more and you have too few blocks per class to average.
• The duration in hours, to the quarter hour. One unit throughout. Do not switch to minutes for the short blocks.
• The charge, on a scale of minus 2 to plus 2, taken about 15 minutes after the block ends, not during it. Minus 2 is "I need a break before I can do anything that matters". Plus 2 is "I would happily do another one". Zero is neutral.

The 15-minute delay matters. Read taken during the block measures how hard the work feels, which is not the same thing. Read taken the next morning measures your sleep. Fifteen minutes after is when you can tell whether the block took something from you.

Computing it

For each class: multiply each block's charge by its hours, add those up across the class, then divide by the class's total hours. That gives you a weighted average charge per hour, which is the number you compare across classes. Weighting by hours matters because a minus 2 on a 15-minute block and a minus 2 on a 3-hour block are not the same event.

Then compute the total deficit: add up the charge-hours of every class that came out negative. Each negative class's share of that total deficit is what tells you where to act.

A worked audit

An owner logs 10 working days, 96 hours across six classes.

Hours add to 96. Charge-hours net to positive 14.0, or about +0.15 per hour across the 96 logged hours.

The trap is the net. Read the net alone and this looks like a healthy week. It was not. The owner was flat by Thursday both weeks. Nets hide the shape, because the positive classes and the negative classes are not interchangeable and do not cancel in practice. You do not get to spend the surplus from Tuesday's field work on Wednesday's dispatch chaos.

The total deficit is 19.6 plus 5.4 plus 5.4, which is 30.4 charge-hours. Now the shares:

• Dispatch and firefighting: 19.6 of 30.4, about 64% of the deficit, from 14 hours, which is about 15% of the 96 logged hours.
• Customer conflict: 5.4 of 30.4, about 18% of the deficit, from 3 hours, about 3% of the 96 hours.
• Admin: 5.4 of 30.4, about 18% of the deficit, from 18 hours, about 19% of the 96 hours.

Three classes, three completely different shapes, and each one needs a different action.

The three shapes and what each one calls for

The rule for structural removal: a class earns a structural fix when it accounts for 30% or more of your total deficit charge-hours AND runs at least 8 hours across the 10-day log (unit: per class, on a log of at least 80 total hours). Both gates. Below 8 hours in the class you have too few blocks for the weighted average to be worth acting on. When the rule fires, the step size is to cut that class's hours roughly in half in the next 10-day log, not to zero, because halving is what lets you tell whether the drain rides with the hours or with the class.

Run the example through the rule as written. Dispatch: 64% of deficit, which clears 30%, and 14 hours, which clears 8. Both gates met, so it gets the structural fix. Conflict: 18% of deficit, under 30%, and 3 hours, under 8. Neither gate met. Admin: 18% of deficit, under 30%, and 18 hours, over 8. One gate only, so the rule does not fire, and that is the instructive case.

High rate, low volume (conflict at -1.8 across 3 hours). Do not try to eliminate this. Customer conflict is part of owning a shop and the three hours were probably three real problems that needed you. What you control is placement. A minus 1.8 block placed immediately before something requiring judgment contaminates the next block, and placed last in the day it follows you home. The instrument is sequencing: never schedule a known conflict conversation in the block before a quote, a hiring decision, or a price call, and never make it the last thing you do. Put a low-demand class after it as a buffer.

Moderate rate, high volume (dispatch at -1.4 across 14 hours). This is the one to change structurally, because it produces 64% of the deficit and the volume means there is something to cut into. Halve it: 14 hours to 7 across the next 10 days, by routing the classes of interruption that dominate it.

Shallow rate, high volume (admin at -0.3 across 18 hours). This is the one owners attack first and it is almost always the wrong first move. Eighteen hours is the second-biggest block in the log, so a time audit flags it hard. But at minus 0.3 per hour it is producing 18% of the deficit while consuming 19% of the hours. Cutting admin buys you hours back, which is a real and separate win, and it will barely touch how you feel on Thursday. Fix it for the hours if you want the hours. Do not expect it to fix the energy, and do not be surprised when it does not.

The re-read that tells you what you actually learned

Run the same 10-day log after the change. What you are reading is not whether you feel better, it is what happened to the charge per hour of the class you cut.

Say dispatch comes back at 7 hours and minus 1.4 per hour, so 9.8 charge-hours against 19.6 before. The drain rode with the hours. Halving the volume halved the damage, and halving it again is worth trying.

Now say it comes back at 7 hours and minus 2.1 per hour, so 14.7 charge-hours against 19.6. You cut half the hours and got less than a quarter of the relief. That means the easy blocks were the ones that left and the remaining seven hours are the genuinely hard ones. Volume was never the problem. Cutting further will keep disappointing you, and the fix is different in kind: it is who handles the hard escalations, not how many escalations there are.

That distinction is the whole reason for halving rather than eliminating. Eliminate the class outright and you never find out which of those two worlds you were in.

What flips the reading

A single dominating event. If one job, one customer, or one personnel problem ran through the whole ten days, the log is measuring that event, not your week. Note it and re-run after it clears. This is the most common reason a first audit reads strangely.

Seasonal load. A class that is restorative at normal volume can invert under peak load. Field work at plus 0.6 across 41 hours can read negative at 60 hours in the same class. The charge per hour of a class is not a fixed property of the class, it is a property of the class at that volume, which is exactly why the audit is repeated rather than done once.

Sleep. If the ten days included consistently short sleep, every class reads about half a point worse and the relative shape still holds. The ranking survives; the absolute numbers do not. Do not act on absolutes from a badly slept fortnight.

Verifying you ran it honestly

• Did you log the short blocks? The most-drained classes hide in fifteen and twenty minute blocks that never make it onto a calendar. A log where every entry is an hour or more has missed the interruptions entirely, and interruptions are usually where the charge is worst.
• Are your classes stable across the ten days? If "admin" meant paperwork on day two and meant anything-not-otherwise-classified by day eight, your weighted average is meaningless. One referent per class, all ten days.
• Did the charge readings vary? A log where nearly everything reads minus 1 is a log taken in one mood, usually all at once on the last day from memory. Real logs have positives in them, including on bad weeks.

References

• See related: The Energy Audit: When You Do Your Best Work
• See related: Batching vs Switching: The Cost of Interruptions
• See related: Decision Fatigue: Why Late-Day Calls Go Wrong
• See related: How to Recognize Burnout Before It Lands
• Trade-standard practice for owner workload measurement in small field-service businesses