$53 Per Lead. 43.9 Percent Book Rate. $233 Per Paying Customer. The 2026 LSA Booked-Job Math.

The SearchLight LSA dataset for February 2026 dropped this week. $6.72 million in tracked spend. 126,650 home services leads. Every contractor running Local Services Ads needs to stare at these four numbers until they're permanent.

$53 cost per lead. 43.9 percent book rate. $233 cost per paying customer. $1,826 average ticket. 7.84x closed ROAS.

That is the math of home services demand capture in 2026. The shops still measuring channel performance by cost per lead are losing money next to the shops measuring by cost per paying customer. The gap is widening fast.

Why $53 Is The Anchor Number

SearchLight tracked the full February 2026 LSA flow across HVAC, plumbing, electrical, roofing, and drain/sewer. The blended home services CPL came in at $53. Inside that average the spread is real but tight:

• Electrical LSA: $39 per lead — the cheapest category in the dataset
• HVAC LSA: $51 per lead
• Plumbing LSA: $57 per lead
• Drain and sewer: $59 per lead
• Roofing LSA: $55 to $95 in non-storm months, $150-plus in storm windows

Compare that to the non-LSA Google Ads channel and the case for LSA stops being subtle. Plumbing non-branded search runs $183 per lead. Performance Max sits at $82. LSA is 49 percent cheaper than blended Google Ads and 64 percent cheaper than non-branded search inside the same SearchLight dataset.

That part is settled. The next part is where shops lose money.

The 43.9 Percent Book Rate Is Where The Money Lives

$53 per lead is gross. The dataset says 43.9 percent of those leads actually book a job. That math gives you a cost per paying customer of $233 against an average ticket of $1,826. Closed ROAS lands at 7.84x.

Two shops paying the same $53 per lead can end up with wildly different cost per customer. The difference is operational discipline, not channel choice:

• Answer the phone in five minutes during business hours — the LSA dashboard tracks call answer rate. Misses kill your match score and your book rate inside the same week.
• After-hours coverage answers within 30 minutes — voicemail is a missed lead in the LSA model. An answering service that actually books an appointment is the difference between a $53 lead becoming a $1,826 ticket and a $53 lead becoming a write-off.
• Dispute every bad lead inside 7 days — wrong service area, wrong service, spam call, no-show. Approved disputes refund the charge. A $5,000-per-month LSA budget typically has 15 to 25 percent disputable leads. That is $750 to $1,250 a month walking out the door for shops not running a weekly dispute review.
• Review the lead recording on every lost call — half the lost-job calls in the dataset traced to a CSR who quoted wrong, gave the wrong availability, or simply did not ask for the appointment. The lead arrived. The shop fumbled.

That is the difference between a $53 channel that prints money and a $53 channel that loses money. The number on the invoice is the same. The shop is what's different.

The Maximum CPL Formula Every Trade Owner Needs Memorized

BaaDigi's 2026 contractor breakdown gives the only CPL math that actually means anything:

Maximum Cost Per Lead = (Average Job Value × Profit Margin × Close Rate) ÷ Target ROI

A real example. Average HVAC job $7,500. 30 percent gross margin. 25 percent close rate from lead to booked job. Target ROI 3x. Your maximum CPL is $188. Anyone paying less than $188 per HVAC lead at those conversion numbers is making money. Anyone paying more is burning it.

Run the same formula on your own numbers. The math tells you immediately which channels to scale and which to kill. Most shops have never written this formula on a whiteboard.

What Changes In The Second Half Of 2026

LSA at $53 stays the best paid channel for the rest of the year. Three forces are about to bend the curve:

• Agentic booking is going to compress the lead-to-booked-job gap — when the iPhone Siri Camera mode ships in September and the AI Mode local pack starts routing bookings through agentic SLAs, the shops with sub-1-second scheduling endpoints get pre-qualified leads that book inside the answer. The 43.9 percent book rate moves toward 60 to 70 percent for shops on Housecall Pro, ServiceTitan, or Jobber with the endpoint live.
• The 32-percent AI local pack cut starves the laggards — Joy Hawkins measured that the new AI local pack surfaces only 32 percent of legacy 3-pack businesses. Shops cut from the AI pack will lean harder on LSA. CPL across the board likely climbs 15 to 25 percent through Q4 as displaced demand compresses into the LSA inventory.
• The dispute window matters more, not less — Google quietly removed three lead categories from automatic credit eligibility in May. Weekly dispute review goes from a nice-to-have to a survival habit.

The Move This Week

Three tasks before Friday:

• Write the Maximum CPL formula on a whiteboard. Plug in your real average job value, gross margin, and close rate. Compare to your actual blended CPL by channel. Kill anything above your max.
• Pull last 30 days of LSA leads. Rate every one. Dispute every wrong-service-area, wrong-service, spam, and unqualified lead inside the 7-day window. Expect 15 to 25 percent recovery.
• Listen to the recording on the last ten leads that did not book. Identify the failure pattern. Fix the one CSR habit that shows up three or more times.

The shops who do those three tasks before next Monday move their cost per paying customer down 20 to 30 percent without changing a dollar of ad spend. The shops who don't keep paying full freight on a channel they only half-understand.