What Is a Good Close Rate for a Contractor or Service Business?

Why a close rate means something only by lead source and salesperson

A company-wide close rate mixes leads that arrived ready to buy with leads that arrived because someone sold the same request to several companies. A referral from a past customer, a tech-generated lead from a service visit, a web lead, a storm canvass and a bid invitation are different conversations before anyone names a price. When the mix shifts, the company-wide rate moves with no change in how anyone sells.

Illustrative example, with hypothetical numbers: two salespeople each present 40 estimates in a quarter. Rep A’s are 30 referrals, of which 15 sell, and 10 web leads, of which 2 sell: 17 sold, 42.5 percent. Rep B’s are 10 referrals, 6 sold, and 30 web leads, 8 sold: 14 sold, 35 percent. Rep A looks better, yet Rep B is better on both kinds of lead: 60 percent against 50 on referrals, about 27 against 20 on web leads. Whoever handed out the leads decided the ranking.

Small counts add noise of their own. A close rate is a proportion, and the NIST/SEMATECH e-Handbook of Statistical Methods sets out the Wilson method for putting a confidence interval around one. For a hypothetical salesperson who sold 10 of 30 estimates in a month, a 33 percent rate, the 95 percent interval runs from about 19 to 51 percent. A quarter at the same rate, 40 of 120, narrows it to about 26 to 42 percent. A ten-point swing between two months of 30 estimates each can be chance alone.

So compare salespeople only on the same lead source and job type, over at least a quarter. Before ranking anyone, check that every estimate reaches the record: a salesperson who enters an estimate only when the customer says yes prints a high rate on a short list. Match each person’s calendar appointments against the estimates in the system; an appointment with no estimate is either a no-bid with a written reason or a missing record.

What roofing, HVAC replacement and remodeling close rates hide

Each of these trades has its own way of putting the wrong things into the fraction, and each fix is a definition, not a sales script.

• Roofing: inspections and insurance work. An inspection booked from a door knock after a storm is a different event from a homeowner calling for a replacement price, so count inspections and presented estimates as separate steps. An insurance job can carry two signatures: a contingency agreement before the carrier decides the claim, and a build contract after the carrier approves the scope. Count the build contract as the sale, and keep retail and insurance work in separate rows. In Florida, § 489.147 bars a contractor from offering a gift, cash, a deductible waiver “or any other thing of value” in exchange for an inspection or a roof claim, and lets the homeowner cancel a roof contract within 10 days, or by the official start date if that comes first, when it was signed within 180 days of events covered by the Governor’s state-of-emergency declaration for the area. In those months a signed roof contract is not a sale until that window closes.
• HVAC replacement: two lead types and a setting. Replacement opportunities come from marketed calls and from technicians on service visits; read them apart. Check the sold threshold on every job type. ServiceTitan’s documentation calls it the dollar amount at which a job counts as sold or converted, suggests at least $1 over the dispatch fee, and says that at $0 every completed job counts as converted. A wrong threshold rewrites the close rate without a single sale changing. And a replacement sold at the kitchen table is a sale made at the home: the FTC’s Cooling-Off Rule, 16 CFR Part 429, lets the buyer cancel a sale of $25 or more made at the residence until midnight of the third business day, and its exclusion for a repair visit the buyer asked for does not reach additional goods or services sold during that visit, other than the parts the repair itself used.
• Remodeling: two closes and a cancellation window. A design-build remodeler can sell twice, a paid design or preconstruction agreement first and the construction contract after it. Report both rates; a strong first rate with a weak second one means design work is being sold that never becomes a build. Where the contract is signed in the homeowner’s house, the same three-day window applies, so count the sale once the window has closed and record cancellations as cancellations.
• Every trade: price shoppers. A phone ballpark is not a presented estimate. A full visit with a written price is, and if it is lost on price, the lost-reason field says so. Leaving price shoppers out of the count is a choice to write down, not a way to raise the rate.

This is what the rule texts say; the company’s attorney decides how they apply to its own contracts. Not legal advice.

How a booking rate differs from a close rate: a service call and a moving company

A booking rate and a close rate measure two hand-offs with two owners. A booking rate is inquiries that became a booked appointment or job, divided by the inquiries that could have; it belongs to whoever answers the phone. A close rate is estimates sold divided by estimates presented; it belongs to whoever presents the price. One blended number from first call to signed job cannot tell a phone problem from a selling problem.

On a service call the first number is the CSR booking rate: lead calls that became a booked job, where the software decides which calls count as leads. ServiceTitan’s CSR dashboard documentation, for example, counts an inbound call as a lead when it lasts at least 60 seconds and is not excused, and excuses a call tied to an existing job or given the call reason that marks it as not a service request. Once a technician is in the house, ServiceTitan’s documentation separates two more: a conversion rate for service, maintenance and install jobs whose invoice subtotal meets the sold threshold, and a close rate for sales jobs whose sold estimates meet it. The CSR books, the technician converts, the salesperson closes, and each number has its own owner.

A mover’s booking percentage hides the same split, so count it in two steps: inquiries to estimates, then estimates to booked moves. For an interstate household-goods move, 49 CFR 375.401 requires a written estimate based on a physical survey of the goods, which the individual shipper may waive only in writing, signed before loading, and 49 CFR 375.103 defines that survey as one done on site or virtually, through live or pre-recorded video that shows the goods clearly. So the estimate count is a count of written estimates, surveyed unless the shipper waived it, not of prices quoted on the phone. Mark a move booked when the date is held the way the company requires, such as a deposit or a signed order, record cancellations before move day, and give leads bought from a service that sends the same request to several movers their own row.

What moves a close rate besides price

Price decides some losses, and the lost-reason field shows how many if it is filled in every time. Before changing a price, split sold and unsold estimates on each field below. Each is a date or a yes-or-no already on the estimate record, or one that can be added this week and read after a quarter.

• Days from inquiry to appointment: same or next day, two to four days, five or more. If the sold share falls as the wait grows, the fix is in scheduling, not in the pitch.
• Everyone who decides in the room: yes or no. An estimate presented to one of two owners leaves the decision to a conversation the salesperson is not part of.
• Presented or only sent: walked through in person or on a call, against emailed with no conversation.
• What the estimate shows: options side by side, and exclusions and allowances in plain words, so the customer can line it up against another bid.
• The first follow-up: days from presenting to the next touch. An estimate with no second touch has not been lost yet; it has not been worked.
• The start date offered: for example, a price with a start eight weeks out competes with one that starts next week, so backlog can move the rate without anyone selling differently.
• Financing, where the company offers it: whether it was presented, and on which estimates.

Read the dollars beside the rate. Sold dollars per presented estimate, the close rate times the average sold job, says more about a salesperson than either number alone: one who closes fewer but larger jobs can bring in more per appointment than one who wins small ones. And a close rate that climbs while gross margin on sold jobs falls can mean the price is doing the selling.

Why published benchmarks do not settle it, and the count that does

An outside close rate can be compared with the company’s own only if both counted the same things the same way. Ask any published figure five questions: what counted as an estimate, every visit or only priced ones presented to a decision-maker; what unit, each option or each customer and job; when a job counted as sold, at signature, at deposit, after a cancellation window or at the start of work; which leads, sources, job types, season and region; and who was counted, a sample drawn to represent a trade or the customers, members or respondents of whoever published it. A figure that cannot answer them cannot be compared with anything.

Even inside one product the words move. ServiceTitan’s documentation defines close rate as closed sales opportunities divided by sales opportunities, while its legacy Technician Close Rate by Job Type report treats a service technician’s converted jobs as that technician’s close rate. The comparison that holds is the company against its own last four quarters, counted the same way, from one row per presented estimate. The salesperson fills in the row at the appointment, in the field-service software, the CRM or a spreadsheet:

• Customer, job, job type, lead source and salesperson.
• Inquiry date, appointment date and the date the price was presented.
• Presented value, by the one rule for options, and whether everyone who decides was in the room.
• Next follow-up date, and each touch made after presenting.
• Outcome: sold, with date, value and deposit; lost, with a reason; no decision, with a date to look again; or canceled after signing.

Every Friday the office manager or a sales manager, not the salesperson whose numbers they are, runs three lists: calendar appointments with no estimate row, estimates past their decision window with no outcome, and sales signed this week still inside a cancellation window. Each item gets a name and a date before the meeting ends; the quote line of the Friday Ledger counts the follow-ups that were due and how many happened. The owner reads the close rate once a month by lead source and by salesperson, by count and by dollars, and once a quarter against the last four quarters.

When not to pay anyone for this, and where the Review fits

A company where the owner presents every estimate, knows each open one by name and records every outcome does not need outside help here; the definitions above, a spreadsheet and a Friday check are the whole job. If the rate holds steady by source and the open-estimate list is short and dated, the time is better spent on how many inquiries arrive, or on how fast sold work gets billed.

Where estimates come from several salespeople and several lead sources, and nobody can say how many presented estimates ended with no outcome, the Revenue Review counts this under the Quote → Contract hand-off. From read-only exports of the field-service software or CRM, the calendar and the invoices, it rebuilds presented estimates by lead source and salesperson, lists the ones that ended with no decision recorded, and puts a monthly dollar figure on them where the records support one. It sets no sales targets and trains no salespeople, and the written report is the company’s to keep.

To start smaller, the Revenue Self-Audit asks two questions about this hand-off in two taps. The 20-minute call is free, and it is where we say whether the count above, done in-house, is enough, which may well be the honest answer.