The Seasonality Tax: What Field Service Companies Pay for Guessing at Demand

Colleen has run a mechanical services company in Brandon for sixteen years. She can tell you exactly when her year gets hard.

The second week of June, when every commercial client remembers simultaneously that their cooling equipment exists. And the second week of November, when the same thing happens with heating.

Sixteen years of the same two spikes, in the same two weeks, with a predictability that borders on comedic. And every year she handles them the same way: overtime, deferred maintenance work, apologetic phone calls, and a stretch of about five weeks where her crew is exhausted and her dispatcher stops answering the phone by mid-afternoon because there is no point.

Then in late July and late February, she has technicians washing trucks.

When her accountant finally quantified it, the two peaks and the two troughs were costing her about $290,000 a year between them. Not from bad service or bad pricing. From treating a completely predictable pattern as an annual surprise.

The number that bothered her most was not the overtime. It was the work she turned away in June. Two hundred and six service requests declined or deferred beyond the client's acceptable window, in a business with a 61% close rate and an average job value of $840.

Seasonality Is Not the Problem

Every field service business is seasonal to some degree. HVAC has two peaks. Landscaping has one long one. Cleaning has an autumn contract cycle and a summer lull. Plumbing has a freeze-thaw spike. Pest control tracks the weather almost exactly.

The seasonality itself is not a business problem. It is a fact of the market, it is shared by every competitor, and it is not going away.

The problem is that seasonality is treated as weather rather than as data. Owners know the busy season is coming in the same way they know winter is coming: as a general sense of impending difficulty, not as a forecast with numbers attached.

And there is a specific reason for that gap. Most field service operators have five to fifteen years of demand history sitting in their own systems and have never once looked at it as a time series. The forecast they need is already in the building.

The Cost Shows Up on Both Ends

At the Peak: Overtime, Refusal, and Quality Drift

Peak-season costs are visible and painful, so most operators can name them. Overtime premium on labour. Emergency subcontracting at unfavourable rates. Expedited material purchases at retail.

Less visible, and usually larger: the work you decline. Colleen's 206 declined requests at $840 average value and a 61% close rate represent roughly $105,000 in revenue that walked to a competitor, and a meaningful portion of those clients never came back to ask again.

Least visible, and most corrosive: quality drift under load. When a crew is running at 110% of sustainable capacity for five weeks, documentation gets thinner, checklists get shortcut, first-time fix rates fall, and callbacks rise, which consumes capacity you did not have in the first place.

Peak season callbacks are the cruelest cost in field service, because they are caused by the peak and they make the peak worse.

At the Trough: Paying Full Price for Nothing

Trough costs are less painful and therefore less managed. Full payroll against reduced billable hours. Fixed vehicle and facility costs against lower revenue. And the quiet one: skilled technicians with nothing engaging to do, which is a retention risk on its own, for the reasons we covered in why good technicians leave.

Nobody quits because of one slow week. People do reconsider their situation during long stretches of feeling underused.

The Composite Picture

For an operator doing $2.6M annually with fourteen technicians and a two-peak seasonal pattern:

Roughly 11% of revenue, in a sector where net margins commonly sit in the 6 to 9% range. The seasonality tax is frequently larger than the profit.

Not all of it is recoverable. Some peak overtime is genuinely the cheapest available option, and some trough idleness is unavoidable. But operators who forecast and plan against their own history typically recover 40 to 60% of the total, which on these numbers is $116,000 to $174,000 a year.

Forecasting With What You Already Have

You do not need demand planning software. You need three years of your own job history and a spreadsheet.

Plot completed job volume by week for the last three years. Not by month, because monthly buckets smooth away the spike you care about. By week. Colleen's June peak is a two-week event that a monthly view renders as a mild increase.

Separate the segments. Emergency versus scheduled, commercial versus residential, service type by service type. Different segments peak at different times, and the blended curve hides the structure. This is the same segmentation discipline that makes operational dashboards useful instead of comforting.

Overlay capacity. Available technician hours by week against demanded hours by week. The gaps are your peaks and troughs, quantified, with dates.

That exercise takes a day and it produces something most field service owners have never had: a numbered forecast of when they will be short and by how much.

Three years is enough to see the pattern. Five is better. And the pattern is usually remarkably stable, because it is driven by weather and contract cycles rather than anything about your business.

The Levers, In Order of Cost

Once you have the forecast, the responses are ordinary business decisions rather than emergencies.

Move flexible work out of the peak. This is the cheapest and largest lever and almost nobody uses it fully. A meaningful share of most field service books is recurring maintenance with flexible timing. Deliberately scheduling that work into your known trough weeks flattens both ends of the curve at once, using capacity you already pay for, at no incremental cost.

Colleen's June crisis was roughly 30% preventable by this lever alone, because she had been letting quarterly maintenance visits land in June by administrative accident.

Pre-position materials before the peak. Buying at planned volumes weeks ahead beats emergency retail purchasing every time. This requires knowing the peak is coming, which is the entire point of the forecast.

Build a known seasonal labour bench. Not scrambling for temporary help in week two of the spike, but a small roster of returning seasonal or part-time people who already know your systems, standards, and sites. A seasonal technician on their third year with you is a productive technician. One hired in a panic is not.

Line up subcontractor capacity in advance, at negotiated rates. Contracting for peak capacity in April is a different conversation, at a different price, than contracting for it in the second week of June. The quality considerations from subcontractor management apply, and they are much easier to satisfy when you are not desperate.

Use price to shape demand at the margin. Off-peak incentives on flexible work, and honest premium pricing for genuinely urgent peak service. This is standard practice in most capacity-constrained industries and unusually rare in field service, where flat pricing is the norm regardless of when the work happens.

Fill the trough with work that only fits there. Deep maintenance, equipment overhauls, training, certification, facility work, and the internal projects that never get done. Trough weeks are the only time these are affordable, and treating them as planned work rather than filler changes how the crew experiences the slow season.

The NowKleen Version

NowKleen.ca has a milder seasonal curve than an HVAC operator, but a real one: a heavy autumn contract-start cycle and a soft mid-summer.

They plotted three years of weekly job volume by segment, found the pattern was stable within about 8% year over year, and made four changes. Flexible recurring work was deliberately redistributed toward known soft weeks. Consumables were pre-positioned ahead of the autumn cycle. A returning seasonal bench of four people was established with standing arrangements. And deep-clean and equipment work was formally scheduled into July rather than fitted in.

The callback improvement was the one they had not expected and valued most. Reducing peak overload did not just save overtime. It restored the quality of work performed during the peak, which removed a category of cost that had always been treated as unavoidable.

The returning seasonal staff figure compounds. Their fourth-year seasonal people now need no onboarding and work to full standard from day one.

Start Here

Move one: plot three years of completed jobs by week, not by month. One spreadsheet, data you already own, one day of work. Monthly views hide two-week spikes, and two-week spikes are what break field service operations. You will be able to name your worst weeks by date before you finish.

Move two: find the flexible work sitting in your peak weeks and move it. Recurring maintenance with timing latitude is the cheapest capacity you will ever free up, and in most operations a surprising amount of it is sitting in the busiest weeks purely because of when the contract happened to start.

Move three: pick your two known trough weeks and put real work in them now. Deep maintenance, training, certifications, equipment overhauls, the internal project everyone agrees matters. Scheduled in advance, as actual work. Not as something to do if things are quiet.

Colleen still has two peaks. The second week of June is still the hardest week of her year, and it always will be.

She turned away eleven requests last June instead of two hundred and six. Her crew worked 94 hours of overtime instead of 312. And in late July her technicians did the equipment overhauls that used to get deferred into the following spring, which is part of why the June peak was easier.

Sixteen years of the same pattern. The pattern was never the problem.

*Basis: SynchronApp scheduling, recurring booking, and utilisation data, NowKleen.ca implementation results, and general field service operating benchmarks. Figures in the composite cost model are illustrative and scaled to a $2.6M, fourteen-technician operation. Content was rephrased for compliance with licensing restrictions.*