You do not need a wall of dashboards. Five numbers, checked every Friday, tell you almost everything about where your shop is headed — and which fire to put out first.
In this guide
Most owners check their numbers in one of two ways: never, or obsessively during a cash crunch. Both are flying blind — the second just adds panic. The alternative is not a wall of dashboards. It is five numbers, checked at the same time every week, until you know their normal ranges the way you know the sound of a healthy blower motor.
1. Booked revenue for the next two weeks
Not last month's revenue — that already happened and you cannot manage it. The number that predicts your next thirty days is what is currently on the schedule. Sum the value of booked jobs (use estimate values, or average tickets for service calls) for the coming two weeks.
Why it matters: this is your early-warning radar. Revenue problems show up here two to four weeks before they show up in the bank account, while you still have time to act — push maintenance-plan outreach, follow up on open estimates, run a small promotion.
What healthy looks like: enough to cover two weeks of fully loaded costs, with a cushion. Learn your seasonal baseline; the signal is the deviation, not the absolute number.
2. Booking rate on inbound calls
Of the people who called with a service need this week, what percentage ended up on the schedule? You need two numbers: booking-intent calls and booked appointments.
Why it matters: this is the leak detector for your most expensive asset — demand. Marketing spend, referrals, and reputation all funnel into the phone, and every unbooked call quietly refunds that investment to a competitor. Shops are routinely stunned when they first measure this honestly, especially once missed and after-hours calls are counted (which is exactly the gap a 24/7 AI receptionist closes).
What healthy looks like: 75 to 85 percent of booking-intent calls answered live should book. If you cannot even produce this number because calls are not logged, that is finding number one.
3. Average ticket
Total service revenue divided by number of completed jobs, tracked weekly and compared to your trailing few months.
Why it matters: average ticket is the compound result of your pricing discipline, your techs' option-presenting habits, and your mix of work. A slow drift downward — $385, $362, $340 — is invisible day to day and devastating annually. A three-truck shop running 40 jobs a week that lets average ticket slip $40 loses over $80,000 a year without a single bad month to point at.
What moves it: presenting good/better/best options, maintenance-plan discounts driving repair approvals, price-book updates actually reaching the field, and techs mentioning the membership at the door.
4. Callback rate
Of jobs completed in the period, what percentage required a return visit for the same issue?
Why it matters: callbacks are triple poison — unpaid truck rolls, shaken customer trust, and the earliest visible symptom of a training or parts-quality problem. One tech's rising callback rate is a coaching conversation; a shop-wide rise is a process problem (rushed diagnostics, a bad parts batch, overbooked days forcing shortcuts).
What healthy looks like: under 3 percent. Between 3 and 5, watch it. Above 5, stop and dig — the cause is almost always specific and findable if your job records note who did what.
5. Receivables over 14 days old
The total dollar amount of invoices sent more than two weeks ago and still unpaid.
Why it matters: profit on paper with cash stuck in receivables is how profitable shops miss payroll. This number also grades your invoicing process — shops that invoice at the door with payment links keep it near zero; shops that bill "when the office gets to it" watch it snowball.
What healthy looks like: less than one week's revenue, excluding commercial accounts on agreed terms. Anything older than 30 days should have a name attached and a next action.
The Friday ritual
Same time weekly — Friday afternoon or Monday morning — fifteen minutes, the five numbers written next to last week's. That is the whole system. Two rules make it stick:
Write them down by hand somewhere persistent. The physical act builds the intuition that makes anomalies jump out.
Every out-of-range number gets one action, not a study. Booked revenue soft? Twenty estimate follow-ups go out. Callbacks up? Pull the job records and find the pattern.
If your operations run through one system, the collection step disappears: Service Command HQ's reports pull revenue, job, and technician performance from the same records that run your day, so the numbers are a glance, not a spreadsheet project. But with or without software, the ritual is the point. Five numbers, every Friday, forever. Within three months you will catch problems weeks earlier than you used to, and that head start is the cheapest management consultant you will ever hire.