1,100 Maintenance Agreements. Where Are the HVAC Hours?

Tony Delgado remembers the dinner after selling 400 maintenance agreements in one fall push. "Agreements are pure upside. They keep us busy when things slow down." Predictable renewals feel like protection when you have twelve technicians to keep working.

The sale does not create the hours to deliver it.

Tony is an illustrative Mesa owner, not a client case study. By the following July, tune-ups crowded the same calendar he needed for repairs and replacements. He now has about 1,100 active agreements.

Two Aprils ago, he faced the same mismatch from another direction: a six-week install backlog, four replacement prospects lost to a faster competitor, and two service technicians sent home early on Thursday. His marketing report had one line item.

This time he wants the agreement count translated into something his dispatcher can schedule.

What did 1,100 agreements actually promise?

Count obligations before counting available appointments. Export each active agreement with the covered address, systems, included visits, last completion, next due window, promised response benefits, and relevant customer constraints.

An agreement is not necessarily one system or one visit. A home with two systems may take longer. A commercial plan may have a different service frequency. Use the actual scope, equipment needs, and terms, including agreements already sold.

For this worked model, assume each of Tony's 1,100 agreements covers one system and promises two maintenance visits per year. 1,100 × 2 = 2,200 visits annually. Assume the roster remains unchanged for the modeled year. Renewals, cancellations, and partial-year starts need their own due dates.

ENERGY STAR recommends yearly equipment tune-ups. That does not establish a universal two-visit schedule. The two visits here are an assumed promise whose timing must fit the system, climate, and agreement.

The capacity planning checklist asks whether maintenance commitments crowd out other work. This is the ledger that lets you answer.

Tony had been reading 1,100 as a recurring revenue base. The same list now shows 2,200 visits his team needs to deliver. His sales count was not wrong. It had never been the whole instruction.

How long does one completed visit really take?

Measure elapsed labor, not just time beside the equipment. In this example, a single-technician visit needs:

• 90 minutes on site for the agreed work.
• 20 minutes of travel allocated to the visit.
• 10 minutes for job notes and completion paperwork.

That is 120 minutes, or two technician-hours per completed visit. The annual requirement is 2,200 × 2 = 4,400 technician-hours.

These are planning assumptions, not industry benchmarks. Check recent work by system type, geography, and scope. An average that hides long multi-system visits will underestimate the weeks where those visits cluster. If two technicians attend, count both people's time.

ACCA's residential maintenance standard describes minimum inspection tasks as a basis for a maintenance program. Define the work your agreement includes, then measure its duration. Do not shorten necessary work merely to make the calendar balance.

Travel and job notes are already inside this two-hour unit. Do not subtract them from available hours and charge them to each visit a second time.

Tony's dispatcher can work with two hours. "A tune-up" had been a smaller box on the screen, even when driving across the East Valley made the visit occupy most of a morning.

Which of Tony's twelve technicians can deliver those hours?

Start with one labor pool. Tony has twelve technicians and three install crews. For this allocation, assume each install crew uses two of those twelve people. That leaves six assigned to service, not twelve plus six installers.

At forty paid hours each, the service group has 240 gross technician-hours a week. Assume 24 hours are committed to leave allowance, training, meetings, and other fixed duties, excluding the visit travel and paperwork already counted. That leaves 216 hours to allocate.

For a flexible-demand week, protect 72 hours for repair work and 24 for unexpected work or callbacks. 216 − 72 − 24 = 120 hours available for maintenance, a ceiling of sixty two-hour visits.

For a peak-demand week, protect 144 repair hours and the same 24-hour reserve. 216 − 144 − 24 = 48 maintenance hours, a ceiling of twenty-four visits. Repair blocks include their own related travel and documentation.

The six installers retain their separate 240 gross hours. Their work, travel, leave, and contingency must fit inside that allocation. Borrowing an installer or service technician requires changing both boards. Use your team's true weekly capacity, not headcount alone, to replace these assumed budgets.

Tony's 40/60 service-install revenue split does not allocate those hours for him. Neither do three CSRs and a dispatcher create another technician. Someone finally has to decide which work each paid hour can accept.

Where do 2,200 visits fit across the year?

Test a proposed calendar against the due-visit ledger. For illustration, divide the year into two service cycles, each containing sixteen flexible-demand weeks and ten peak-demand weeks. These are planning windows, not a recommendation that every climate has the same shoulder season.

Schedule fifty-five completed maintenance visits per flexible week and twenty-two per peak week:

Each cycle delivers 1,100 visits. Each flexible week consumes 110 of its 120 maintenance hours, leaving ten for recovery. Each peak week consumes 44 of 48, leaving four. Those gaps total 400 hours annually.

The table passes arithmetic. It passes scheduling only if the individual visits fit: one appropriate visit per customer in each modeled cycle, within the actual promise. If more visits must happen during peak, rebuild the allocation. A neat annual total cannot override a due date.

Tony can now see a possible year, including 440 peak-period visits. Moving every tune-up out of summer was never a workable instruction. Finding which visits can move, with the customer and the agreement supporting it, is.

What happens when a visit cannot move or does not happen?

Book within allowed windows early. Ask about access, customer availability, equipment use, and timing needs. Group nearby work where practical. Preserve any response commitments and benefits already promised. Change future offers deliberately rather than silently changing existing service.

Keep two records: visits still owed and hours already consumed. A no-show may use travel time without completing the promised visit. Record that time, the contact attempts, and the new appointment. Do not count the visit complete or assume the original slot cost nothing.

Use the ten flexible-week recovery hours or four peak-week hours for actual overruns and rescheduling. Those allowances are assumptions to test. If misses consume them repeatedly, revise the workload, confirmation process, or staffing plan. Do not quietly reduce the repair reserve to make the maintenance report look current.

Review overdue and approaching-due visits every week. Give each a named owner and a next action. Carry unfinished visits forward visibly, preserving the original due date. Customer contact, any agreed change, and the reason for delay belong in the record.

Tony wants his dispatcher to have a permitted response to a missed appointment. "We'll squeeze it in" recreates the July he is trying to avoid. A recovery slot and a visible queue give the promise somewhere to go.

What would another 400 agreements consume?

The 400 sold in Tony's earlier fall push are already included in his current 1,100. Do not add them twice. Under this model, that historical cohort alone represents 800 annual visits and 1,600 hours within the existing workload.

Another 400 agreements on top of today's base would create 1,500 agreements, 3,000 visits, and 6,000 hours annually once all are active for a full year. The increase is 800 visits and 1,600 hours. Stage a partial-year sales forecast by its actual service windows.

The current maintenance ceilings provide 32 × 120 + 20 × 48 = 4,800 annual hours. Existing visits use 4,400. Even spending every one of the 400 recovery hours would leave a 1,200-hour shortfall against the expanded 6,000-hour obligation, with no recovery space remaining.

You need a supported change: more deployable skilled hours, genuinely shorter measured visit time without cutting scope, a different mix of new offers, or a smaller enrollment target. Do not assume 400 agreements justify borrowing the same install hours already sold elsewhere.

This is the practical continuation of service versus install capacity planning. Demand for one department cannot consume the other department's hours invisibly.

Tony can still want another strong fall. He now has a question to ask before the dinner: where will the extra 1,600 hours come from?

What does the agreement price actually pay for?

Use the same promised workload when testing price. Suppose your allocated technician-hour cost is $70, including technician pay, employment costs, vehicle, and allocated overhead. In this illustration, it excludes the office administration and supplies listed separately below.

• Four technician-hours per agreement annually × $70 = $280.
• Supplies at $10 per visit × two visits = $20.
• Agreement administration = $24 annually.
• Expected included-benefit and rework cost allowance = $16 annually.

Modeled annual cost: $340 per agreement. At an illustrative $360 annual price, $20 remains for unmodeled costs and profit. That is a thin difference to investigate, not a recommended price or margin benchmark.

Use your actual allocation method and benefits. Avoid charging the same overhead twice. Include promised discounts, priority response, payment costs, and other obligations where relevant. Do not make an unpriced replacement sale the assumption that rescues the agreement's economics.

Tony had treated recurring receipts as evidence the program worked. A delivery-cost column gives him another test. He can sell a relationship confidently and still check whether the price funds the visits it promises.

What should Tony put beside the next sales target?

Keep one operating record: active agreements, visits due by week, measured hours per visit, protected repair and install capacity, recovery hours used, overdue visits, and delivery cost. Assign who updates each line.

Use those obligations before calculating how many new leads your company needs or deciding when to increase advertising. Existing customers already own part of the calendar you are about to market.

Every agreement needs a place in the year before its renewal becomes a reason to celebrate.

In Tony's illustrative shop, the next step is to sit beside his dispatcher with the 1,100 names. The report still shows his roughly $11,000 monthly marketing budget. Now the plan beneath it starts with 2,200 visits and 4,400 hours.

He puts the proposed sales target beside the additional 1,600 delivery hours. Before selling another 400, he can see exactly which hours are still missing. This time, the calendar reaches the sales meeting before the target does.