Where an Owner's Hours Actually Go

Why this matters

Two owners hand you the same headline: about 60 hours a week, mostly on the tools. One of them is a week away from losing a large customer and does not know it. The other is fine and will still be fine in a year. The difference is not in the total, it is in the shape, and the shape is what almost nobody reads.

This card is about reading a distribution of owner hours the way you would read a set of pressures or readings on a system: not one number at a time, but the pattern they make together, and what that pattern predicts is going to break first. It assumes you have measured. If you have not, a sibling article covers how to run the study without producing numbers you cannot trust.

The seven buckets, and what each one actually contains

Category boundaries are where most self-audits go soft, so be strict about these.

Field. Hands on the work. Not driving to it, not talking about it, not writing it up afterward.

Drive. Windshield time between sites, including the stop at the supply house. Owners fold this into Field and it hides one of the more compressible parts of the week.

Routing. Answering somebody else's question so that person can proceed. Approvals, "where do we keep," "is it okay if," "which one did they want." The defining feature is that the work in question is not yours and you are the unblocking step. This is the bucket the whole card turns on.

Selling. Estimating, quoting, walking a job with a prospect, following up on outstanding quotes.

Admin. Invoicing, payroll, ordering, filing, reconciling, anything that leaves a record and could in principle be done by someone who has never held a tool.

Customer. Conversations not attached to a job in progress. Complaint calls, relationship maintenance, the long-standing account that likes to talk to the owner.

Owner work. Deciding where the business goes. Numbers review, hiring, pricing, capacity planning, deciding what the shop will stop doing. If it does not change the shape of the business, it is not in this bucket.

The bucket everyone underestimates

Routing is understated in every self-estimate, by a lot, for a mechanical reason: it arrives in pieces below the resolution of memory. A ninety-second question does not register as work. Twenty of them do not register as thirty minutes. And Routing is the bucket most likely to be lost in a time study's unaccounted hours, because you cannot log a two-minute interruption while it is happening.

So treat any measured Routing figure as a floor rather than an estimate. If the study says 15% of logged hours, plan as though the real number is meaningfully higher, and check it against the interruption count rather than against your sense of the week.

Routing also has a property no other bucket has: it is the only one that grows as the shop grows, without any decision being made. Add a tech and you add questions. Add an office person and, in the first months, you add questions. Every other bucket you can at least see coming.

The bucket everyone overestimates

Owner work, and specifically the part of it people count. Reading a report is not Owner work if nothing changes as a result. Thinking about the business while driving is not measurable and should not be logged. A conversation about whether to hire, held four times without a decision, is one hour of Owner work and three of worry.

Count only time that produced a decision, a number you now know, or a written plan. Owners who apply that test typically find their Owner work bucket falls to somewhere between a third and a half of what they first logged, and the honest smaller number is far more useful because it is the one that predicts whether the business will look different in a year.

Reading the distribution, not the buckets

Four patterns cover most small shops. Match on shape, then read what it predicts.

The technician shop. Field plus Drive over about half of logged hours, Owner work under 5%. Common and not automatically wrong, especially in the first years. What it predicts: the shop's ceiling is the owner's own capacity, and it will be discovered suddenly rather than gradually. The break is usually a large opportunity that arrives during a busy stretch and gets a slow, thin quote, or a quality slip on a job the owner got pulled off halfway.

The switchboard shop. Routing above roughly 20% of logged hours, and interruptions running above about 15 a day on a sustained basis. Field can be anything. What it predicts: output falls out of proportion to the owner's absence. Not by a bit, disproportionately - a shop that runs at 100% with the owner present can run at half that in the week the owner is out, because the unblocking step is missing.

The sales shop. Selling plus Customer above roughly 30% of logged hours with delivery genuinely delegated. Usually a shop that has hired well on the field side. What it predicts: sold work outruns delivered work. The break shows up on a lag, typically six to ten weeks after a strong selling month, as a schedule that no longer has slack in it and a rise in callbacks from work done in a rush.

The back-office shop. Admin above roughly 25% of logged hours. What it predicts: the slowest and most survivable failure of the four, which is why it persists for years. The break is quiet - receivables discipline drifts because the invoicing is being done at nine at night, and nobody notices until a cash squeeze that has been building for two quarters.

There is a fifth condition that is not a pattern in the buckets at all, and it is the one that catches people who look balanced. Alongside your bucket totals, record the longest uninterrupted block you got in the whole study period. A distribution can look reasonable in every bucket and still be unworkable if nothing in it lasts longer than 25 minutes, because some of the work in the Selling and Owner work buckets cannot be done at all in fragments.

Two patterns at once, and one worked read

Patterns combine, and the combination is usually more informative than either alone.

A three-tech shop, owner plus a part-time office helper. Two-week study, 96 logged hours:

Longest uninterrupted block in the whole two weeks: 55 minutes, once.

The read. Switchboard and back-office at the same time. Routing at 26 hours and Admin at 21 hours is 47 of 96 logged hours, about 49%, so close to half of measured working time is going to unblocking other people and to paperwork. Field, at 22.9% of logged hours, is not the story here even though it is what the owner would name first.

The 55-minute figure changes what you can do about it. An owner in this position often reaches for the obvious move, which is to carve out planning time. It will not hold. A shop generating this much Routing does not have a 90-minute gap in it to protect, and a protected block that gets broken twice stops being protected.

What it predicts, and how to test the prediction cheaply. This shape predicts that a two-day owner absence costs more than two days of output. Test it deliberately rather than waiting for a flu: schedule two days out, tell nobody it is a test, and afterward count three things - jobs that slipped, decisions that waited, and questions that queued up for your return. A shop in this pattern typically comes back to a queue rather than to a resolved week, and the size of that queue is the honest measure of how much of the business is stored in one person.

What it implies for the fix, and why the obvious order is wrong. The instinct is to hire office help to take Admin, because Admin is 21.9% of logged hours and looks like the clean removable block. But Routing at 27.1% is larger, it grows with the shop, and it is fed by the absence of written rules rather than by the absence of hands. Hiring into a shop with this much Routing usually raises Routing in the first months, because a new person generates questions. Write the rules that retire the top few repeating questions first, then hire, and the same hire lands on a shop that can absorb them.

That order flips under one condition: if the Admin hours are concentrated in a hard deadline you personally cannot move, such as payroll, hire first and accept the temporary Routing increase, because a missed payroll is not a productivity problem.

The second axis that changes every conclusion

Bucket totals alone will mislead you, because two shops with identical distributions can be in completely different positions depending on who else could hold each bucket. Carry a second column.

The row that surprises people is Routing. It is the largest compressible bucket in most small shops and it is the one that cannot be handed to a person, because what you are handing over is not tasks but answers. Until the answers exist somewhere outside your head, every hire routes through you and the bucket holds its share of the week no matter how many people you add.

References

• See related: How to Run a Time Study on Yourself
• See related: Owner Time Audit: Where the Week Goes
• See related: The Owner Bottleneck: When Everything Needs My Approval
• See related: The Owner's Calendar: Protecting Time to Think
• Trade-standard practice for owner workload analysis in small field-service businesses