Dale Whitmore sees space in next month's calendar and says, "We've got a slow month coming. Turn the ads up." A monthly total feels sensible when payroll and the marketing invoice both arrive by the month.
The month cannot take an appointment.
Dale is an illustrative five-technician owner in Springfield, Missouri. His recurring mistake was raising June ad spend from $3,000 to $8,000 while long attic jobs hid the real workload. This hypothetical planning exercise asks a narrower question: which particular weeks could take the work he wants to buy?
Which weeks disappear inside the monthly average?
Build a rolling weekly plan before assigning a seasonal budget. Start with your actual staff, skills, work mix, leave, and existing promises. Show capacity by work type where durations differ. A short service visit is not interchangeable with an installation simply because each produces one invoice.
For a compact illustration, assume these are comparable job units with the same mix and average time requirements. They are modeled planning quantities, not industry seasonal benchmarks:
| Planning week | Usable job capacity | Existing commitments | Remaining openings |
|---|---|---|---|
| 1 | 45 | 30 | 15 |
| 2 | 40 | 32 | 8 |
| 3 | 35 | 31 | 4 |
| 4 | 45 | 25 | 20 |
| Total | 165 | 118 | 47 |
The lower capacities in weeks two and three represent assumed staffing and availability changes. They are already usable quantities after the appropriate headroom, not gross hours waiting for another buffer. Dale's usual approximately 45-job target already includes scheduling reserve. Do not reduce it again by an automatic twenty percent.
For your real plan, use true weekly capacity and replace the comparable-job shortcut with qualified person-hours or crew blocks whenever the mix changes.
Dale had circled forty-seven. His dispatcher circles four. Both are correct, but only one describes the third week's ability to accept a caller.
What already owns the shoulder-season opening?
Before labeling a week slow, list work already owed. Include confirmed repairs, accepted installations, agreement visits due under actual contract terms, callbacks, carryovers, training, leave, and any other obligation not already deducted in your usable baseline.
Keep two boundaries clear. An absence deducted from capacity does not also belong in committed job hours. A confirmed maintenance visit included in commitments is not a fresh lead requirement. Record whether each deduction sits on the supply side or the obligation side, then count it once.
Maintenance timing should follow equipment requirements, local conditions, customer needs, and the agreement you actually sold. A shoulder-season plan is not permission to move every customer's visit into the same convenient month. Some appointments have limited flexibility.
Use the maintenance agreement capacity model for that obligation ledger. This seasonal worksheet coordinates all work types around it. Shared technician planning matters when an installation draws the same people who would otherwise service those agreements.
Check readiness too. An accepted install awaiting equipment may not consume this week's field block yet, but its likely arrival belongs in the forward plan. Show tentative obligations separately so uncertain deliveries do not disappear until they become emergencies.
Dale's dispatcher brings the agreement list to the review. The blank part of the calendar now has names waiting behind it. He needs a weekly demand case before calling that time surplus.
How should you model a hotter or quieter month?
Hold the earlier capacities and commitments constant for this exercise. Assume new demand arrives in the same comparable job units and needs service in its arrival week. Do not assume every homeowner can wait for week four.
In a base case, ten additional jobs seek each week: forty across the month. Total demand becomes 118 existing commitments plus forty new jobs, or 158, below the month's 165 usable capacity.
Yet week two has room for eight of its ten new jobs and week three has room for four. The model serves thirty-two new jobs and cannot place eight in their required weeks. Fifteen of the forty-seven openings remain unused elsewhere. The totals reconcile: forty-seven equals thirty-two served plus fifteen unused.
A quiet case of six new jobs per week produces twenty-four requests. Twenty-two fit, with two unable to fit in week three. A pressure case of eighteen each week produces seventy-two requests. Forty-five fit and twenty-seven do not, leaving just two openings unused in week four.
These are conditional scenarios, not weather predictions or permission to accept excess bookings. If customer-agreed timing is flexible, model that separately. If the new mix is heavier, rebuild hours rather than applying these job counts.
Dale can now see how a month can look underfilled and still disappoint eight callers. The problem is when the openings exist, not just how many appear in the total.
Which commitments can move without breaking trust?
Sort work into fixed, conditionally movable, and genuinely flexible. Fixed work includes commitments you cannot responsibly or contractually shift. Conditionally movable work needs a conversation and agreement. Flexible work still needs a deliverable date, parts, and qualified staff.
Ask the customer about a proposed change rather than assuming the company owns their calendar. Record the agreed date, who authorized the move, and whether any promise or price changes. Keep required service intervals and local conditions in view. An urgent repair cannot become a shoulder-season maintenance visit through a spreadsheet label.
Suppose four of the base-case requests that would not fit earlier can appropriately move to week four, with customer agreement and no conflict. The same model would then serve thirty-six new jobs and leave eleven openings unused. Four requests would still lack a suitable appointment. That is a separate revised case, not the original result.
Do not count the moved jobs again as fresh demand in week four. Move the existing records and retain their original inquiry dates. That lets you distinguish demand timing from completion timing.
Dale asks which customers actually accepted the change. The answer is a list of records, not an assumption that “people can wait.” The remaining shortage stays visible.
When should the campaign change?
Give paid demand a weekly assignment by work type and territory. If week three has four compatible openings, a campaign designed to fill twenty generic jobs does not match the plan. Week four's twenty openings deserve their own lead time and intake coverage.
For an explicit demand-planning illustration, suppose half of suitable inquiries book and eighty percent of those bookings complete. Four bookable openings divided by a fifty-percent booking rate gives an expectation of eight suitable inquiries. Four resulting bookings would produce 3.2 expected completions at the assumed eighty-percent completion rate.
These assumed rates forecast demand over comparable inquiries. Actual bookings must fit the four confirmed openings. The 3.2 is an expected value, not a fractional appointment or permission to book a fifth job. Do not assume canceled time can be refilled. As bookings arrive, reduce the remaining requirement.
Use how many leads your company needs to work backward from the gap. Use when to run HVAC ads to set a bounded test, with spend, response ownership, and a review point tied to observed bookings.
Dale's instruction changes from “fix the month” to the specific work and week the team can handle. His dispatcher can challenge that instruction before the spend creates a promise.
What belongs in Friday's seasonal review?
Compare the prior plan with actual staffing, received inquiries, bookings, cancellations, completions, time spent, carryovers, and customer-agreed moves. Keep the original forecast so a revision does not erase the evidence of what changed.
Refresh the near weeks in detail and the later weeks as ranges. Weather-sensitive demand is uncertain. So are parts arrivals and employee availability. Give each assumption an owner and a next check date. A seasonal label is less useful than knowing what would change the next decision.
Check whether unused capacity came from too little suitable demand, poor response, unready work, or the wrong skills in the wrong week. These call for different actions. Do not interpret every shortfall as a cue to increase every campaign.
Dale keeps forty-seven on the pad, then draws four weekly boxes underneath it. The four in week three gets its own circle. Before another seasonal promotion leaves his desk, the people who must deliver it can see exactly where its appointments would go.
