The Handful of Numbers a Small Shop Owner Should Actually Watch

Why this matters

Give a new owner a reporting tool and the instinct is to watch everything it can show. Six months later the reports stop opening because there is no room left in a busy week to study forty numbers, and the one metric that actually mattered got lost in the pile. Owners who stay disciplined about this do the opposite: they name a short, fixed set of numbers before they ever look at a report, one number per real business question, and everything else waits until a specific problem calls for it. The skill here is not which metrics to pick, it is how to pick them at all.

Start from the questions, not the metrics list

Do not open a reporting tool and scroll through every available field. Start with the handful of questions that actually run a service business, and find one number that answers each. A shop rarely needs more than these five questions covered:

• Are we busy enough? One number: work booked or scheduled for the coming stretch.
• Is the work itself earning? One number: what a typical job nets after its direct costs.
• Is money getting collected? One number: how long it takes from finished job to cash in hand.
• Is quality holding? One number: how often you have to go back and redo something.
• Are customers coming back? One number: the share of business that is repeat or referral.

Five questions, five numbers. Resist the pull to add a second number under any question "just in case." A question with two numbers under it usually means you have not decided which one actually answers it.

One number, one owner

Every number on the list needs a person who is accountable for it moving in the right direction, even in a shop of one. If nobody owns a number, nobody notices when it drifts, and a metric with no owner degrades into wallpaper within a month. In a bigger shop, assign the number to the role that can actually influence it: the dispatcher owns the booked-work number, the lead tech owns the redo rate, the office owns the collections number. Ownership turns a metric from a fact you glance at into a responsibility someone carries.

The test for adding a sixth

The list stays short by design, but real problems do come up that a fixed five cannot answer. Add a number only when you can answer yes to both of these:

• Did one of the core five already tell you something is wrong? A sixth number exists to explain a symptom the core list surfaced, not to look for problems the core list has not found yet.
• Will a specific person change what they do based on this number, starting the next time they see it? If the honest answer is "we would just note it," it does not earn a permanent spot.

A shop with a stubborn redo-rate problem might add a per-tech breakdown temporarily, work the problem, and then retire the extra number once it is solved. The core five stay constant; the extras are seasonal.

Guard against the number that flatters instead of informs

Some numbers feel good to watch precisely because they never demand anything of you. Total jobs done all-time, or a big pipeline of unqualified inquiries, climbs steadily and asks nothing in return. Test any candidate metric by imagining it dropping sharply: would that drop actually change what you do tomorrow? If a falling number would not trigger any different action, it was never really a working number, just a nice one to look at.

Match the number to who is going to read it

A number chosen well still fails if it is shown to the wrong audience. The owner's five belong on the owner's short list. A technician does not need the collections number; they need their own redo rate and their own booked hours. A dispatcher does not need the repeat-customer share; they need today's coverage and tomorrow's gaps. Keep one core list per role, each answering that role's real question, rather than one long combined list everyone half-reads.

Revisit the list on a fixed schedule, not constantly

The five questions rarely change, but the exact number under each can be worth revisiting. Do this deliberately, on a set cadence such as a quarterly check, not every time a number has a bad week. Ask: is this still the sharpest single number for this question, or has the business changed enough that a different measure answers it better? Swapping a metric mid-crisis just because it looks bad is how a shop ends up quietly hiding a real problem instead of fixing it.

The mental model to keep

A short list you actually read beats a long one you stopped opening months ago. Pick the number of business questions you can honestly hold in your head at once, usually four to six, put exactly one number against each, name who owns it, and let everything else wait for a specific reason to exist. The discipline is in the restraint, not the analysis.

References

• U.S. Small Business Administration (SBA), small business performance measurement
• Trade-standard practice for field-service operational review
• See related: The Dashboard Numbers That Matter; The Data You Should Actually Track; A Metric That Doesn't Change a Decision Isn't Worth Tracking