How Many of Your Unbooked Calls Were Actually Bookable? Lead Quality Benchmarks for Home Services (2026)

SearchLight Digital · Data through September 30, 2026 · Updated October 8, 2026

In September 2026, SearchLight Lead Grading graded 232,635 inbound calls, forms and chats across 1,500+ home services contractors. 57,508 of them booked on the spot. That leaves 175,127 unbooked contacts, the number every "revenue at risk" dashboard in the industry is built on. 78% of them were never bookable: existing customers confirming or moving an appointment, billing questions, dead air, solicitors, wrong numbers, people outside the service area. Only 35,728, one in five unbooked contacts, were a homeowner with a real service request who left without an appointment. And of those bookable-didn't-book contacts, roughly half go on to book anyway once the office follows up.

So the honest "revenue at risk" number is about 11% of unbooked contacts, not 100%. The rest of this page shows where those 11% come from, by reason and by channel, and why the number a contractor should manage to is bookable book rate (61.7% network-wide in September), not raw book rate (24.7%).

Raw book rate vs bookable book rate: the denominator problem

Most booking-rate benchmarks in home services divide booked calls by all calls. That is raw book rate, and in September it was 24.7% across the network. It is a useful number for sizing call volume, and a terrible one for judging a CSR team, because 59% of what came in was unbookable before anyone picked up the phone.

Bookable book rate divides booked contacts by bookable contacts only: the homeowner wanted service the business offers, in an area it serves. In September that was 61.7%. The gap between the two numbers is the share of calls that were never a job, and it varies more by channel than the booking skill does.

Read across the Google Business Profile row: a 24% raw book rate looks like a problem, and a 64% bookable book rate says the CSRs are doing fine. The listing is simply where existing customers call, so 61% of what it produces is confirmations, reschedules and billing questions. Facebook is the opposite case: only 27% bookable, and the bookable ones book at 43%, the weakest of any channel, which is a lead-quality problem and a talk-track problem at once. Direct mail is the standout, 53% bookable and 82% of those booking, because the homeowner called with the postcard's offer in hand.

What the unbookable contacts actually were

SearchLight Lead Grading, 137,222 unbookable contacts across 1,500+ contractors, September 1 to 30, 2026.

The thing marketers worry about most, calls for a service the business doesn't offer or from outside its service area, is 4.4% of unbookable contacts. That is still worth fixing where it lands on a paid campaign, because it is the one category negative keywords and geotargeting can remove. But it is not where the volume is. Existing customers (administrative questions, confirming, rescheduling and cancelling appointments, billing, membership, callbacks) are 45.8% of everything unbookable, and dead air (the caller hung up at the menu, said nothing, or the transcript was empty or cut off) is another 36.9%.

Neither of those is a CSR failure, and neither is lost revenue. An existing customer calling the number on a Google Ad to confirm tomorrow's appointment is an operations finding: nothing after booking gave them a better number to call. Dead-air calls on a paid line are worth a look at the phone menu and hold times, and on an AI-answered line they are feedback on the agent, not on the campaign.

Why the bookable contacts didn't book

SearchLight Lead Grading, 35,728 bookable-didn't-book contacts across 1,500+ contractors, September 1 to 30, 2026.

Planned follow-up is the biggest bucket, 52.6%, and it is not lost revenue yet: the caller and the office agreed on a next step and no appointment was made on the call. Add form submissions (8.7%), which by definition cannot book at the moment of submission, and 61% of the bookable-didn't-book pile is pipeline waiting on the office's follow-up process rather than a lost caller.

The two reasons a business decides on the call itself are pricing concerns (15.4%, most often the diagnostic or trip fee) and scheduling availability (11.9%, the caller needed a slot sooner than offered). Together they are 27% of bookable-didn't-book contacts, or 9,745 contacts in September across the network, and they are the ones a talk track and a capacity decision can move. The remaining 10% is voicemail and callers who needed a human after reaching an AI agent, both of which are answer-path problems rather than sales problems.

How many of them booked later, and the honest "revenue at risk" math

A grade describes the contact at the moment it happened. A caller graded bookable-didn't-book on Tuesday who books on Thursday's callback keeps the Tuesday grade; the booking shows up on the customer. So the right question for revenue at risk is not how many bookable contacts failed to book on the call, but how many of those customers never booked at all.

For August 2026, the most recent month with enough time for follow-ups to land, 46.2% of bookable-didn't-book contacts came from customers who went on to book. 53.8% did not. Across all bookable contacts in August, 75.6% came from customers who booked at some point, on the call or after it. That is in line with the 73 to 74% SearchLight published in July on the May and June cohorts.

Put the pieces together for a September-sized month:

Eleven percent of unbooked contacts is the population that was a real job, was not booked on the call, and was not recovered. Any tool or report that prices every unbooked call at an average ticket is overstating the opportunity by roughly nine times. The useful version of the number is that 19,222 contacts, at the network's average ticket, is still a very large amount of money, and it is concentrated in two reasons the business controls.

What to do with this

Manage to bookable book rate, not raw book rate. A CSR team's number is booked divided by bookable. Network-wide that was 61.7% in September; the best channels run above 70%. If a report hands you a 25% booking rate with no bookability split, it is telling you about your call mix, not your team.

Sort unbookable by who can fix it. Service-not-offered and out-of-area contacts on a paid campaign go to the agency as negative keywords and geotargeting. Existing-customer calls on a paid line go to operations: what does a customer receive after booking, and does it carry a number to call back on. Dead air goes to whoever owns the phone menu, the hold path and the AI agent.

Work the two reasons the call decides. Pricing concerns and scheduling availability were 27% of bookable-didn't-book contacts, 9,745 in a single month. The first is a talk-track question, usually the diagnostic fee. The second is a capacity question, and the contacts tell you which business unit and which days are short.

Stop counting follow-ups as losses, and start counting the ones that never close. Half of the bookable-didn't-book pile is a planned next step, and about half of those book. The leak is the other half, and the only way to see it is to tie each graded contact back to the CRM and check whether a booking ever appeared.