Service Agreement Profitability

The 100-point spread that nobody talks about

Across 259 eligible contractors with at least $10K annual SA revenue, company-level service-agreement gross margin ranges from -30% (bottom decile, losing 30 cents on every SA dollar) to +70% (top decile). The median sits at 37.9%. This is a cross-trade company distribution, not matched companies in the same market or the margin on every agreement.

That 100-point cross-company spread makes a line-by-line check in your own P&L useful, even when service agreements are not the largest revenue line. It does not measure what any one company can improve.

Four practical levers are worth testing: pricing discipline, scope control, customer mix and pull-through. Cost-to-serve visibility supports the first three. The benchmark does not measure how much of the spread each one explains.

Lever 1: pricing discipline

A common risk is fees based on an old price plus a small bump rather than current fulfillment costs. Check labor hours, parts, vehicle and dispatch costs, and when the fee was last rebuilt.

A practical repricing review prices direct cost-to-serve to a gross-margin target, then checks allocated overhead separately to establish net contribution. Reprice annually when wages, fuel or parts costs move, and decline agreements that cannot cover their required costs and contribution.

Test the attrition risk before repricing. A 20% increase holds revenue if attrition stays below about 16.7%; contribution still depends on retained scope and delivery costs. Test a scenario of a 20% price increase and 10% attrition: 90% of the original agreements x 120% of the old fee = 108% of original revenue. Attrition is an assumption, not a measured outcome; profit depends on retained work, costs and scope.

Lever 2: scope control (where SAs quietly bleed)

An SA covers a defined scope, typically scheduled maintenance plus minor adjustments. The bleeding happens when techs treat the SA as 'whatever the customer asks for.' Add-on tasks that should be billed separately get rolled in. Diagnostic time on issues outside the SA gets eaten. Free callbacks compound.

Start by testing process controls before repricing. Every SA visit needs a defined checklist (what's covered) and a quote-on-site mechanism for anything outside it. Techs need to be trained that 'I'm covered for the maintenance, but this repair is separate' is the customer expectation, not an upsell trick.

Use a 10-15-point improvement within 90 days as a planning test, not a measured customer promise. Compare the same agreements and supported direct costs before and after scope controls.

Lever 3: customer mix and the high-maintenance trap

Not all SA customers are equal. The retained 10-15% of customers / 50%+ of callouts reference is an unverified historical observation, not a measured typical distribution of callouts, complaints or scope creep. Test the concentration in your own book. Check whether high-callout customers also received discounts or regularly request work outside scope; this pattern is a practical hypothesis, not a measured outcome.

One option to test is repricing high-usage customers at renewal rather than terminating them. Quote renewal against actual cost-to-serve and the required margin, and explain the usage basis. In a planning scenario where 30% renew at a cost-covering price, verify their actual delivery cost and contribution; renewal and profitability are not guaranteed.

The signal to watch: SA customers whose unscheduled callout rate is more than 2x the book average. Those are the first customers to check for actual cost-to-serve and contribution; high usage can hide below-target margin in the average.

Lever 4: pull-through, a separate revenue opportunity

Service agreements aren't supposed to be the profit center. The agreement fee gets you in the door. The pull-through revenue, replacements, repairs, accessory upgrades surfaced during PM visits, adds billed revenue; its contribution depends on the labor, materials and other costs of that work.

Across 386 contractors, the median annual pull-through ratio is 8.7% (every $1 of SA revenue generates $0.09 of additional work). Top quartile: 29.6% (about 30%). At an illustrative $500K annual SA-revenue base, the median-to-P75 difference is $104.5K/year (about $105K) of billed add-on revenue, before delivery costs. This cross-company comparison is not a forecast for your company.

Retained trade references are mechanical 46%, HVAC 30%, plumbing 5.5% and electrical 18%; their per-trade eligible samples are unverified. They do not establish which process caused the difference. Test on-site recommendations and a 48-hour follow-up cadence against your own completed and billed work.

A 90-day test plan

• Pull SA-level P&L for the last 12 months. Allocate every PM visit's labor and parts cost to the agreement. Compute gross margin per SA. The bottom 20% is your reprice list.
• Build a checklist for every SA visit type. What's covered, what's not, what triggers an on-site quote. Train techs once and audit weekly.
• Set a 48-hour rule for follow-up on any tech-flagged opportunity. Every visit that surfaces a potential repair or replacement gets called within 2 business days. Compare technician-flagged opportunities with quoted and won work to check where follow-up is lost.
• Review the highest-callout 20% as a planning queue. Test renewal pricing against their actual cost-to-serve and required margin. Cost plus 30% is a 23.1% margin (30 / 130), not 30% margin. Communicate the usage basis and track actual renewals; departure rates are not measured here.
• Track SA gross margin, annual pull-through ratio and eligible-agreement renewal rate monthly. Review flagged follow-ups and visit-checklist exceptions weekly; these cadences are practical tests, not measured explanations of the decile spread.