Ask a service company owner what they know about their business, and the answer usually sounds like this: “We had a good month. Busier than March. Maybe.” Then you ask which client is actually profitable, what a new customer costs to win, or how much of your schedule is quietly leaking money, and the silence gets comfortable.
That silence is the real problem. Most service businesses are running on vibes that feel like data. The phone rings, the truck rolls, the money comes in, and the instinct is to assume the business is healthy. But health and busyness are not the same thing. You can be booked solid every day and still be losing money on half your jobs. You can be “slow” on the surface and quietly profitable underneath, just priced for a cost structure that has changed.
Analytics is how you find out which one you are. Not the glossy dashboard kind that impresses nobody, but the plain, weekly kind that changes what you do on Monday morning. And the good news: for a service company, you do not need a data science team to do this. You need a short list of the right metrics, honest numbers, and a tool that pulls them together without eating your Saturday.
The trap: tracking activity instead of results
Before the list of metrics, the mindset, because this is where most analytics efforts die.
Service businesses generate a lot of activity data: jobs completed, invoices sent, miles driven, calls answered, hours on site. All of it feels important, because all of it is work. But activity tells you what happened. It does not tell you whether what happened was good.
The distinction is the whole game. Here is a quick test you can run on your own number of “jobs completed.” Two companies each finish 100 jobs a month. Company A finishes 100 small jobs that take two hours each and net $40 profit after materials and labor. Company B finishes 100 jobs that take two hours each and net $90. On the activity metric, they look identical. On the metric that matters, Company A makes $4,000 a month and Company B makes $9,000, and one of them is probably quietly dying.
So the rule for picking metrics is simple: keep only numbers that change a decision. If a metric would not make you price differently, hire differently, stop doing something, or chase something, it is a decoration. Every metric in this post passes that test.
Metric 1: Gross profit per hour (or per tech, per crew)
This is the single most useful number in a service business, and it is the one owners skip most often.
The formula is not scary:
Gross profit per hour = (Total revenue – direct costs) / total direct labor hours
Direct costs are the costs tied to the job itself: materials, consumables, job-specific subcontractors, fuel for the runs. What you are not counting here is rent, your own salary, insurance, the office wifi. Those are overhead, and they come later. For now, this metric answers one question: for every hour a person is out working, how much money actually lands?
Why per hour and not per job? Because time is the scarce resource in a service company. You can win more jobs, but you cannot buy more hours in a week. The job that looks like a big deal but takes three days and needs two techs may be worth less per hour than the small one that takes an afternoon. Per-hour gross profit lets you compare jobs, clients, and techs against each other on the same ruler.
How to use it:
- Look at it by client. You will almost always find a client you love who is actually expensive to serve: endless changes, long drive time, materials markup that does not cover itself. That is your pricing conversation.
- Look at it by job type. The category you “sell more of” may be the one dragging your average down.
- Look at it by tech or crew. A big spread between your best and worst is a training and dispatching problem, not a morale problem.
A practical threshold: most well-run field service businesses sit somewhere between $100 and $150 of gross profit per direct labor hour. Under $75, the model is leaking. Over $200, you may be under-staffing and burning people out. The exact number depends on your trade, but the direction of the trend matters more than the absolute value.
Metric 2: Billable utilization rate
Utilization answers a different question than gross profit per hour. Gross profit per hour asks “how good is the work?” Utilization asks “how much of the time is actually being paid for?”
The formula:
Billable utilization = (Billed hours / total available working hours) x 100
In a field or professional service company, a typical range is 65% to 80% billable, depending on the trade. That means for every 40-hour week on the clock, 26 to 32 hours show up on invoices. The rest is travel, admin, meetings, dead time between jobs, and the stuff nobody schedules but everybody does.
Why this matters so much: utilization is the cheapest growth there is. You do not need a single new customer to raise it. You need better dispatch, shorter admin time, less dead travel. A company at 70% billable that gets to 78% with zero new sales has grown about 11% in revenue without touching a single phone number.
Metric 3: Customer retention and churn
In a service business, retention is not a loyalty program. It is the foundation of the entire financial model, because a returning customer is the cheapest revenue you will ever have.
Track it the honest way:
Churn rate (monthly) = (Customers who stopped in the month / customers at the start of the month) x 100
For recurring-service businesses (cleaning, HVAC maintenance, property management, agencies, salons, anything with a schedule), this number is the whole ballgame. A 5% monthly churn sounds gentle. Over a year, it means you lose roughly half your customer base and you are always filling the leak. That is why a business with 5% monthly churn feels like a treadmill, no matter how busy it is.
The companion metric is customer lifetime value, and the two have to be read together:
CLV = average monthly spend per customer x average months a customer stays
The reason you need both: a customer who spends $200 a month for 8 months (CLV $1,600) is a different animal from one who spends $80 a month for 24 months (CLV $1,920), even though the first one looks “bigger.” The second one is worth more and costs you nothing to keep.
How to use it: when churn ticks up, do not ask “why are we losing customers” in the abstract. Segment it. Is it a specific service line? A specific region? A specific year of tenure (most churn happens in the first 90 days, which is an onboarding problem, not a service problem)? The segment tells you where the fix actually lives.
Metric 4: Customer acquisition cost, and the ratio that makes it useful
CAC is the cost of winning one new customer:
CAC = (Total sales and marketing spend in a period) / (new customers won in that period)
For a $6,000-a-month marketing budget that wins 30 customers, that is a $200 CAC. Clean, simple, done.
But CAC alone is a number you can never act on, because it has no context. A $200 CAC is fantastic or terrible depending entirely on what that customer is worth to you over time. So the useful version is the ratio:
CLV to CAC ratio = customer lifetime value / customer acquisition cost
A common working target for small and mid-size service businesses is a ratio of 3 to 4. That means every dollar of acquisition spend returns three to four dollars of lifetime value. Under 2, the growth channel is eating itself. Over 10, you are probably leaving a lot of the market on the table because you are not spending enough to grow.
Two practical notes. First, be honest about what goes in the numerator: the ads are only part of it. Include sales labor, time spent on proposals, and any discounts or freebies that land the deal. Most companies’ “real” CAC is 30 to 50 percent higher than the ad spend they look at. Second, track payback: how long until a new customer’s profit repays the acquisition cost? Under 12 months is a healthy line for most SMB customer relationships. If it takes two years, your cash flow is quietly subsidizing your growth, and that is a conversation your bank will have for you eventually.
Metric 5: No-show rate and schedule fill rate
Service companies sell time that only exists in one specific place at one specific moment. When a customer no-shows, you lose the revenue of that job AND the opportunity to put someone else in that slot. In a busy business, a no-show is not a small miss. It is a compound one.
No-show rate = (no-showed appointments / scheduled appointments) x 100
A no-show rate under 5% is a well-oiled operation. Above 10%, you are running a leaky calendar and probably overstaffing to cover it, which is a silent tax on every job you do complete. The fixes are almost never mysterious: confirmation reminders (the automated kind, not the “we will call you on the day” kind), deposit policies for first-time customers, and a waitlist that actually gets used. All three are software features, which we will come back to.
Schedule fill rate is the companion metric:
Schedule fill rate = (booked time / sellable time) x 100
Read it against utilization. High fill rate and low utilization is an interesting combination: the calendar is full, but the jobs are short. That is a bundling and upsell problem. Low fill rate and high utilization means you are running hot and probably losing work you could take on. These two, side by side, tell you whether your capacity problem is demand-shaped or pricing-shaped.
Metric 6: Revenue per employee (and the trap hiding inside it)
This one sounds obvious and is misused constantly, so here is the honest version.
Revenue per employee = (Total revenue / number of employees)
As a trend line, it is genuinely useful: it tells you whether your workforce is producing more or less over time. As a target, it is a trap, because revenue and profit are different animals. A company can grow revenue per employee by selling more low-margin work, and the metric will cheer while the bottom line shrinks.
The healthier habit is to track revenue per employee AND gross profit per employee, together, always. The gap between the two tells you what the growth is made of. If revenue per employee climbs and profit per employee stays flat, you are growing your way toward thinner margins, and eventually the rent, the payroll, and the truck payments will meet that growth and absorb it.
Also worth a line: for field businesses, remember that “employee” should usually mean your billable field staff, not the whole headcount. The office does not drive revenue the same way the trucks do, and mixing the two muddies the number.
Metric 7: Customer satisfaction, measured in a way you will actually use
Net Promoter Score, satisfaction scores, review averages: all of them are fine as long as you treat them as lagging indicators, which is what they are. By the time a customer drops a star, the experience already happened. The value of CSAT is in the trend and in the text behind the number, not the score itself.
The leading signals are the ones worth putting on your dashboard:
- Complaint volume and time-to-resolution (how fast does a complaint get a human answer, and how long until the customer calls it closed?)
- Repeat service requests on the same job (a “second visit” rate; this is your quality defect line)
- Referral or review rate per customer cohort (not just counts, but the rate, because the counts track volume and volume tracks growth, not loyalty)
A service company that reads complaint volume weekly fixes problems before they become churn. One that reads a quarterly satisfaction survey fixes nothing before it is too late. The cadence of your measurement should match the speed at which your customers can leave you.
How to actually use the numbers: thresholds, trends, and one decision each
A metric with no threshold is a number. A threshold makes it a signal. So for each metric on your dashboard, write down the line where you act. Not in a policy document. One line, in the dashboard itself or on the page where you look at it:
- Gross profit per hour under $75 for two consecutive months: pricing review, this week, not “someday.”
- Billable utilization over 85% for a month: add capacity or re-dispatch. People are burning.
- Churn over 4% monthly: pull the last 10 lost customers and interview them before you change anything.
- CLV:CAC under 2 for a channel: cut that channel or fix the onboarding that is shortening lifetimes.
- No-show rate over 10%: turn on automated confirmations and deposits. Today.
And the trend rule: one bad month is noise; two is a trend; three is a pattern. Act on the pattern, not the blip, and you will stop over-reacting and stop under-reacting at the same time.
The final discipline is the best one: every metric on the board should have exactly one decision attached to it. Gross profit per hour by client decides pricing. Utilization decides dispatch and hiring. Churn decides onboarding. CAC decides where marketing money goes. When a number has no decision attached, it belongs in a filing cabinet, not on the dashboard.
Building the software around your metrics, not the other way around
This is where we get to the part we are biased about.
Most software is bought first and metrics are discovered later. You buy the scheduling system, you buy the invoicing tool, maybe a CRM, and then someone builds a spreadsheet that joins them on Tuesdays. The spreadsheet starts life accurate. By month four it is stale, patched, and half the team has quietly stopped trusting it, which is how the “we just feel like it was a good month” business is actually running.
The alternative is to start from the metrics. You sit down, you pick the six to eight numbers above that change your decisions, and you build the system so those numbers are produced by the workflow, not imported from it. Dispatch logs utilization automatically. Invoicing ties job costs to clients, so gross profit per hour computes itself. Retention is visible the moment a recurring customer skips. CAC is real because the marketing spend is already in the same system as the customers it won.
That is the Acherus way, and it is the reason a purpose-built tool beats a pile of best-of-breed apps for this job:
- The data is captured at the moment of work, not reconstructed at the end of the month. A tech clocking a job into a field-ready app writes the utilization and job-cost rows as a side effect of doing their job. There is no “data entry after hours” step, which is the step where analytics die.
- The metrics are computed by the system, so they cannot be quietly wrong. When the spreadsheet formula breaks, nobody knows for a month. When the dashboard runs on live data, a broken metric is obvious, and you fix the cause, not the symptom.
- The thresholds live in the tool. “Churn over 4% triggers a lost-customer review” is a rule the software can run, surface, and assign. In a spreadsheet, it is a hope.
- The whole thing scales with the business. A two-truck operation and a twelve-truck operation need the same eight numbers; they just need the system to stop requiring a full-time person to maintain it.
The shape is simple: decide what you measure, then build the software so measuring is what the software does, not a chore bolted onto the end of the day.
Frequently Asked Questions
How many metrics should a small service business track?
Six to eight, and not more. The limit is not the software; it is attention. A dashboard with twenty numbers gets glanced at; a dashboard with eight gets read, and the owner actually acts on the two or three that moved. Start with gross profit per hour, billable utilization, churn, CLV:CAC, no-show rate, and revenue per employee. Add customer satisfaction trend and one trade-specific metric. That is a full command center.
Which metric should I look at first if I am starting from zero?
Gross profit per hour by client. It is the one that most often surprises owners, it takes a day to compute from invoices and cost notes, and it immediately tells you where your pricing is wrong. Everything else in this list becomes more useful once you know which clients are actually earning their time.
How often should I review these numbers?
Weekly for the operational ones: no-show rate, utilization, complaint volume, schedule fill. Monthly for the financial ones: gross profit per hour, CAC, churn, revenue per employee. Quarterly for the strategic ones: CLV:CAC by channel and the retention trends by cohort. The rhythm matters more than the depth. Ten minutes on the right numbers every week beats a two-hour deep dive every quarter.
Do I need custom software, or will a dashboard off the shelf be enough?
An off-the-shelf dashboard is a great way to learn the metrics: you will quickly learn which ones you care about and which ones you ignore. But once the numbers start driving real decisions, the gap shows up in the data quality, because the dashboard is pulling from a system built for other purposes. If your scheduling, billing, and job costs live in three tools that do not talk to each other, no dashboard will ever be fully trustworthy. That is the point where a custom build, shaped around your own metrics and workflow, pays for itself in decisions you used to make on instinct.
About Acherus Inc.
Acherus is a software company that builds practical, well-engineered tools for small and growing businesses. We help owners and operators pick and implement the technology that actually fits how they work, so the software saves time instead of creating a new job.
© 2026 Acherus Inc. All rights reserved. This article is provided for general informational purposes and does not constitute professional advice. Prices and features are approximate and subject to change. Confirm details with each vendor before purchasing.