What close rate is — and the definition that keeps it honest
Close rate is the share of sales visits that end in a sold job. That sounds simple, and most shops measure it loosely enough that it doesn't mean much: booked calls in the denominator, "sold" whenever an estimate is signed, techs and the office and the owner's test login all mixed together.
The definition that actually tells you something is stricter: close rate = sales visits that sold ÷ sales visits that actually ran. A canceled visit never counts. A visit that hasn't happened yet never counts. A tune-up where no price was given isn't a sales visit. Only real technicians and comfort advisors are in the number — not "Unassigned," not the owner's account.
Defined that way, the number can't be flattered by booking a lot of calls and can't be hurt by a rainy week of cancellations. It only moves when a tech stands in a home and something happens. That's the property you want, because every decision below depends on trusting it.
Two more rules keep it honest. Rank by sold / ran, not the percentage alone — 2 of 2 is 100% and means nothing; 19 of 29 at 66% means something. And don't trust any cell with fewer than about eight ran visits. Below that, you're ranking noise.
Why close rate leaks in the first place
Contractors usually blame lead quality first. Sometimes that's right. More often, the leak is in field execution, and it has five familiar causes:
- Technicians get technical training and almost no customer training. They can diagnose a failed capacitor in four minutes and have no plan for the first two minutes at the door.
- Every homeowner gets the same presentation. The one who's been burned by the last company and the one who's read every review of every unit hear the same pitch, in the same order.
- Dispatch optimizes for geography instead of opportunity. The replacement estimate goes to whoever's nearest — and whoever's nearest might be your best repair tech and your worst replacement closer.
- Managers review results after the opportunity is gone. A monthly close-rate number arrives with no read on which visits were lost or why.
- Techs walk in blind. They know an address and a complaint. Who they are, how long they've lived there, whether there are two decision-makers, whether a hailstorm rolled through last month — they find out at the door, if at all.
Good techs figure all of that out in the first five minutes. Great techs figure it out in the first two. Everybody else guesses. Guessing is the leak.
Step 1: Measure it by technician, by job type, and by conversation
Your company close rate is an average, and the average is hiding the fact that each tech is good at some visits and weak at others. Break it three ways:
- By job type. Replacement estimates, repairs, maintenance — as named in your scheduling software. Sort by revenue. Your money is usually in three to five job types; don't analyze all thirty.
- By technician inside each job type. Open the biggest job type and rank the techs: rate, sold / ran, revenue. This is the single most useful table in the business, because it's the table dispatch decisions come from.
- By the kind of conversation. A no-cool call and a replacement estimate are different job types, but so are two replacement estimates where one homeowner has been burned before and the other has researched every unit. The techs who win those two conversations are often different people.
Use 90 days for anything you'll make a rule from, 30 days for coaching, and Lifetime for renewal conversations. And write down the Before — your close rate from your software's history on the same techs and job types — before you change anything, so the After means something.
Step 2: Prepare every visit before the knock
The cheapest close-rate improvement available to a home-service company is to stop sending technicians in blind. A one-page read on the visit — who is likely to open the door, how long they've lived there, what the house tells you, what happened in the neighborhood, what the weather did last month — changes the first two minutes, and the first two minutes decide most visits.
What a useful pre-call plan contains, in three passes:
- The three-second read: any flags (a priority customer, a recent storm, a veteran household) and the posture to walk up with.
- The ninety-second read: an opening question, three questions that confirm the read, what to sell on, what to say if price comes up, and a closing question that actually asks for the business.
- The five-minute read: the homeowner (does the booking name match the owner on record?), the home (year built, size, permits pulled), the system (a 2004 air-conditioning permit means a system past twenty), the neighborhood, and recent hail, wind or heat.
You can build this by hand — a CSR with the call notes, the county property-appraiser site, the permit portal and a weather report can assemble it in about thirty minutes per job. That's the catch. Nobody has thirty minutes per job, which is why the visits that get prepared are the two big ones a week and everything else goes in blind. What pre-call planning contains, in detail.
Step 3: Name the conversation (the six plays)
Information isn't a decision. A tech can know the system is twenty years old, the owner has been there seventeen, and a hailstorm passed nearby in April — and still knock and ask "so what seems to be the problem?" The read was there. The conversation never used it.
Naming the conversation fixes that. Almost every sales visit is one of six:
| Play | What it means | You know it's right when… |
|---|---|---|
| Value | They want to see the math work. Cost of owning, not sticker price. | They ask 'what's this going to cost me over time' or mention the bill. |
| Research | They've read up and want to buy right. Often the biggest tickets. | They quote a model number or say 'I've been looking into this.' |
| Comfort | The house doesn't feel right — a room, noise, humidity. | They describe a feeling, not a failure. |
| Emergency | Something is broken now. Closes well if you don't slow it down. | The first thing they say is about today. |
| Trust | They've been burned, or they're careful. Prove before you propose. | They ask about licenses, warranties, or 'what happens if.' |
| Family | More than one decision-maker. Nothing closes without the second person. | You hear 'let me talk to…' or 'we'll decide together.' |
Each play has an opening question, three questions that confirm it, what to sell on, a price answer, and a close. The tech's first job at the door is to confirm the play; the second is to run it; and if the house disagrees — the brief said Comfort and the first thing they say is "the last guy ripped us off" — switch. The six plays, with the words that work.
Step 4: Route by close rate, not proximity
Here is the pattern that shows up in nearly every shop that measures by tech and job type. Same three techs, very different visits:
| Job type (90 days) | Marcus | Priya | Dale | Avg ticket |
|---|---|---|---|---|
| Replacement estimates | 66% (19/29) | 52% (14/27) | 31% (11/35) | $7,900 |
| Service / repair calls | 44% (38/86) | 40% (33/82) | 58% (61/105) | $890 |
Dale is the best repair tech in the shop and the worst replacement closer, and for months he got a third of the replacement estimates because his truck was usually closest. Moving 20 of his replacement estimates a quarter to Marcus and Priya — at their close rates instead of his — is roughly seven more sales, about $55,000 a quarter, on visits the company was already running. Dale keeps the repairs, where he wins.
The rules of thumb: rank by close rate, break ties by revenue; availability beats rank by one spot, not three (if #1 isn't free, #2 is fine — sending #4 because #1 was busy is how the leak happens); re-rank every 90 days. The full method, with the Dispatch Fit sheet.
Step 5: Coach one play a month, not 'close rate'
A monthly number is not a coaching plan. A tech told to "close more" hears that they're failing and has no idea at what. Close rate by conversation type changes the meeting: the same tech is third on the team at Emergency and last at Trust — he loses Trust visits in the first two minutes, and there's a teammate who wins them. That's a plan with a name, a partner and a date.
The monthly one-on-one, in ten minutes:
- Start with a win from the data — their best job type or play this month, with the sold / ran numbers. People repeat what gets noticed.
- Name one play to work on. Not "close rate." Show them who on the team closes it best and what that person does in the first two minutes.
- Agree on the next three weeks: a ride-along, a specific opening question to try, or a job type to get more of.
- Write the number down and revisit it next month.
And never coach on a bad Tuesday. The fastest way to lose a good technician is to coach them on noise. Judge every number against the tech's own last 28 days, against techs in the same role, and against the company trend before you call it a pattern.
Step 6: Watch five numbers every week
| Number | What it tells you | Healthy looks like |
|---|---|---|
| Close rate, last 7 and 30 days | Are we converting the visits we run? | Rising over the 30-day line; above your Before number |
| Before / After lift | Is the change paying for itself? | A steady positive gap after the first month |
| Average ticket on sold jobs | Are we selling the right thing, not just more things? | Stable or rising with close rate |
| Preparation actually used | Did techs read or hear the plan before the visit? | Above 80% of sales visits |
| Weakest play × strongest tech gap | Where is the biggest coachable leak? | The gap shrinking month over month |
Fifteen minutes on a Friday. If the fifth number is shrinking, everything else follows.
How ServiceScout does this before every appointment
A documented process is the starting point. ServiceScout makes that process happen before every appointment: it reads the job from ServiceTitan, Housecall Pro, Jobber or JobNimbus, builds the Scout Brief in about ten seconds, chooses Today's Play about ninety seconds later, delivers it to the tech by text, Slack, Teams, phone call or on the job, reads the outcome back from your software, and updates close rate by technician, job type and play — automatically.
The appointment enters ServiceScout
A job is created, scheduled, rescheduled or assigned in ServiceTitan, Housecall Pro, Jobber or JobNimbus. Nothing changes for the CSR.
2:00 PMBick, Harold · Est — AC Replacement · #48213BookedThe brief is assembled
About ten seconds later: the homeowner (and whether the booking name matches the owner on record), the home and its permits, the system's likely age, the neighborhood, and recent hail, wind or heat.
HomeownerOwner on record 17 years · booking name matches the deedThe system2004 condenser change-out, nothing since — past twenty, old refrigerantWeatherHeat advisory three of the last seven days · no hail within 10 miToday's Play is chosen
About ninety seconds later the play and the win condition are on the brief, with the Before You Knock card: opening question, three questions, what to sell on, if price comes up, closing question.
Trust PlayWin on Peace of MindOpening question“Before I touch anything — what happened last time?”Biggest mistakeLeading with price. A round number with no reasons ends the visit.The dispatcher reviews the opportunity
On the Dispatch board the visit shows its play. Replacement estimate + Trust → the best replacement closer who is also strong on Trust, if free. In Assign & Send, this step runs itself.
NeedsBick · Est — AC Replacement · TrustAssign Marcus (#1 · 66%)The tech hears the 90-second audio
Scout Audio, in the assigned tech's name, plays on the drive — from the brief, or as a phone call for the tech who never opens a link. Re-made automatically if the job is reassigned.
Scout Audio1:30 · “Marcus, this is your brief for Harold Bick…”The tech opens the brief
Thirty seconds on the card in the driveway. Then the knock. Every send is stamped sent, opened and listened, so the office knows the visit was prepared for.
Sent 1:31 PMOpened 1:34 PMListenedClosing question“If anything's off in thirty days, I come back at no charge. Fair?”The outcome is recorded — automatically
Sold, not sold, estimate open: read from your scheduling software after the visit. Nothing is self-reported. Only visits that ran count.
OutcomeBick · Est — AC ReplacementSold · $8,400Management measures the result
Close rate updates by technician, by job type, by campaign and by play. The Morning Brief flags wins and concerns; the Sales Plays report shows who wins which conversation.
Play Share Close Best Ride-along Trust 53% 41% → 47% Brandon 52% Jourdan 15% Research 9% 62% Robert 71% Ronnie 40%
Product mockups built from the ServiceScout interface and brand system. Sample data from a demo account; names changed.
What it's worth, in real numbers
From a real HVAC account, names changed: the most common play closed at 41%, two techs closed it at 46–52% and two at 13–15%. Two decisions came out of one screen — route Trust to the strong closers when free, ride-alongs for the weak ones. The full case study.
Run your own numbers
Close-rate revenue calculator
Stays in your browserGross revenue on visits you are already running. Use ran visits, not booked calls.
Frequently asked questions
What is a good close rate for HVAC, plumbing or electrical sales visits?
It depends on the mix of job types. Replacement estimates in the 40–60% range and repairs above 50% are common in well-run shops, but the useful comparison is your own Before number on the same techs and job types — and the spread between your best and worst tech on the same play, which is usually far larger than the gap between your company and the industry.
Should I fix close rate before buying more leads?
Almost always. A point of close rate is free visits — sales on appointments you already paid to book and already sent a truck to. At 400 visits a month and a $7,000 ticket, a point is $28,000 a month; five points is $1.7 million a year. More leads cost money and add visits; a point of close rate costs preparation and routing.
How many visits do I need before a technician's close rate means anything?
About eight ran visits on a job type or play before the ranking is more than noise. Under that, treat the cell as faint. Use 90 days to make dispatch rules and 30 days to coach.
Does close rate include canceled or rescheduled visits?
No. Only visits that ran count. A canceled visit drops out; a rescheduled one counts when it runs. That's what keeps the number from moving with booking volume or the weather.
What's the fastest close-rate win?
Stop sending your worst replacement closer on replacement estimates because he's nearest. It needs one table — close rate by tech inside your biggest job type — and one rule on the dispatch board. Most shops find seven or more sales a quarter in it.
Isn't naming the conversation just a script?
No. The play is a starting read, not a script. The tech confirms it at the door with three questions and switches when the house disagrees. Scripts tell a tech what to say; a play tells them where to start and what to listen for.
The short version. To increase your home-service close rate, measure it honestly first — sold visits divided by visits that actually ran, by technician and by job type — then prepare every technician before the knock, name the conversation they're walking into, route your biggest tickets to the techs who close them, and coach one skill a month instead of “close rate.” Before you buy more leads, price a point of close rate on the visits you already run: at 400 sales visits a month and a $7,000 average ticket, one point is four more sales — $28,000 a month.