Skip to content

When the crew is full in July, move the work into October instead of buying more July

Turning off weekends doesn't lower the bill, and the seasonality setting is built for a three-day sale. The lever is which months you buy work in, measured against what your crew can run.

By Lonta AIPublished Updated 12 min readHow we write

TL;DR

  • The HVAC slow season and the busy season are one problem seen from two ends. Your techs' hours are fixed and demand isn't, so extra July spend buys calls the board can't take while October trucks sit. Plan the ad year from crew capacity by month, not from last year's spend.
  • The budget and the target decide how much you buy. The ad schedule doesn't: Google paces toward 30.4 times the daily budget "regardless of how many days the campaign is scheduled to be active".
  • Since August 17, 2026, a Target CPA campaign limited by budget performs closer to its stated target, even where it used to beat it. Cut a July budget and leave a loose target, and the calls you still buy trend toward the loose price.
  • Seasonality adjustments aren't built for a season. Google calls them "ideal for short events of 1–7 days".
  • For the shoulder months, sell what doesn't need the weather, and check in Keyword Planner whether anyone in your area searches for it.

Is the HVAC slow season a budget problem or a capacity problem?

Most pages on this question list promotions and skip the constraint that decides the rest: how many calls your techs can run in a month. If payroll doesn't drop in January and the trucks don't multiply in July, the ad budget is the one cost you can move between months.

Start from two columns: calls run by month from your field software, and slots by month: techs × calls per tech per day × working days.

Illustration, not client data. Five techs running four calls a day over 22 working days give 440 slots a month.

MonthCalls runSlotsOpen slots
July440, more pushed to next week4400
October260440180
January300440140

An extra July call adds a job only if it goes to overtime, replaces a worse job, or the customer agrees to wait. October has 180 slots you are paying techs to stand by for. Every month where calls run fall short of slots is part of your slow season, whatever the calendar says.

What does one more July call buy when the board is full?

Compare months on cost per runnable call: cost per booked call from ads, divided by the share of those calls your crew ran within two days of the call.

Illustration, not client data. A booked call costs $150 in July, and with the board full one extra booking in three gets run within two days. The rest wait, cancel or push out a job you'd have run anyway. That's $450 per runnable call. In October the same call might cost more, say $200, and every one fills an open slot. October is the cheaper month to buy work in, even at the higher sticker price.

Your runnable share comes from your CRM: July jobs booked from ads that were run within two days of the call, divided by all July jobs booked from ads. That is an average across the month. The last bookings of a full week fare worse than the average, so treat it as the best case for an extra call. Cost per booked call by month comes from the same place, as cost per sold job does.

A capped season in an account we run

The Comfort Authority fixes air conditioning and plumbing in Tampa from one shop. Tampa doesn't have a shoulder season the way Dallas does: cooling runs most of the year, and the crew is at capacity from May to September.

The account didn't buy more calls in that stretch. It bought closer ones: the ad area went from forty-seven cities to what a van reaches on drive time (how that's set up), and calls from ads went down on purpose.

The Comfort Authority · Tampa, FLRead the case
1.6×
Jobs sold from paid search
Window
90 days against the 90 before
Source
Client CRM

The Comfort Authority, Tampa Bay. Radius rebuilt on drive time, search terms cleaned out, spend moved to answered hours. Jobs sold from paid search: 1.6×. Window: 90 days against the 90 before. Source: Client CRM.

Limit: the two windows are not weather-matched. Two Tampa quarters are never identical, and some of this movement belongs to the season rather than to us.

The account's ceiling was the calendar, not the budget: past a point, more July calls were not worth buying.

Which Google Ads settings actually turn July down?

The budget. For most campaigns, a day can spend up to twice the average daily budget and a month up to 30.4 times it. Change it mid-month and Google says spend "for the rest of the month won't exceed your new average daily budget multiplied by the remaining number of calendar days in the month".

Not the ad schedule. Campaigns run "All day" by default. You can cut days and hours, but Google's spending limits page says that with a schedule "the system will pace to reach this full monthly limit regardless of how many days the campaign is scheduled to be active". Our reading: switching Saturdays off to rest the crew leaves the same monthly target to be spent on the other days. Use the schedule to decide when calls arrive, in the hours your CSRs answer (what an unanswered one costs). Bid adjustments by hour won't do it either: Smart Bidding strategies "don't use your manual bid adjustments (for example, +20% on Saturdays)", though "they do respect your Ad Schedule settings."

The target. On August 17, 2026, Google changed how Target CPA and Target ROAS behave when a campaign is limited by budget. Such campaigns now "perform more consistently toward your bid target, including when you make budget adjustments". Google's example: a $10 target that had been achieving $5 "now delivers more closely to a $10 actual CPA". Our reading: cut the July budget and leave the spring target, and the fewer calls won't be cheaper ones. Set the target to what a July call is worth with a full board. Google's FAQ says "you can also adjust your bid targets to manage campaign spend", and its bidding guide says a target change "won't trigger a 'learning' status". Maximize conversions with no target set is different: Google says it optimizes "to spend your full budget without a target", so there the budget is the only dial.

Pausing. Google's bid strategy status page lists "Campaigns are paused" as a reason a strategy goes Inactive, and a strategy "recently created or reactivated" as a reason for Learning. Google's learning-period page says "It can take a few conversion cycles (1-2 typically) for the bid strategy to calibrate to the new objective". Our reading: a cooling campaign paused all winter and switched on in the first hot week starts the season in Learning. A reduced budget through the quiet months avoids that.

The seasonality setting isn't for your season

The name invites an owner to set one for June through August. Google's page says a seasonality adjustment tells Smart Bidding to expect a change in conversion rate during a future event, is "ideal for short events of 1–7 days", and "may not work as well" beyond 14 days. It adds: use it "only if you expect major changes to conversion rates, because Smart Bidding already manages seasonal events." For Search campaigns it works only with Target CPA or Target ROAS.

It changes the conversion rate bidding expects, not the number of calls you can take. Google's bidding guide says Smart Bidding takes seasonality into account and "in most cases, it will automatically handle seasonal increases in traffic without requiring any input." None of the budget, bidding or schedule pages we read has a place to tell Google how many calls your crew can run.

What to sell in the HVAC shoulder season

Two questions: does the product need the weather, and does anybody search for it? Our shortlist for the first:

  • Maintenance visits and plans, sold first to people already in your CRM rather than bought back through search (customer lists).
  • Replacing a system that still runs, with a fixed install date in the open weeks.
  • Duct, insulation and whole-home work, as in the case below.
  • The second trade: water heaters, drains and panels don't follow the cooling curve (two trades, one budget).

For the second question, open Keyword Planner, set the location to your service area, and use the date range, which Google says lets you "view search volume trends over specific time periods". Google notes the figures are rounded and shown for exact matches only. If October searches for a product are close to zero, search can't fill the slots with it. The work has to be proposed: to your customer list, by mail, or through ads that reach people who weren't looking.

A winter in an account we run

Way Cool sells plumbing, air conditioning and home performance in Phoenix. From late November there is close to nothing in cooling for about four months, and the crews stay on payroll. The account had been cutting the winter budget while the whole-home audit sat on the website and every campaign was built for someone whose air conditioning had failed.

The winter budget moved: search narrowed to audit and insulation work, and paid social went to neighborhoods with the right housing stock. The first offer, built on energy savings, ran for two weeks and drew enquiries at a price that would have ended the test. The one that worked led with the bedroom that never gets cold.

Way Cool Plumbing & Air · Phoenix, AZRead the case
2.3 to 2.5×
Audits booked in the off-season
Window
December–February against the year before
Source
Client CRM

Way Cool Plumbing & Air, Phoenix. Winter budget moved from cooling searches to the home audit. Audits booked in the off-season: 2.3 to 2.5×. Window: December–February against the year before. Source: Client CRM.

Limit: we doubled off-season spend. A doubling from a small base is an easier decision than it sounds, and this does not demonstrate the same return at a serious winter budget.

Way Cool Plumbing & Air · Phoenix, AZRead the case
−37%
Cost per booked audit
Window
December–February against the year before
Source
Call tracking

Same account. Cost per booked audit: −37%. Window: December–February against the year before. Source: Call tracking.

Limit: one winter against one winter. Two data points is not a trend, and Phoenix winters vary more than people outside Arizona expect.

A shoulder campaign with an end date

The campaign total budget is one amount over a start and end date, and Google says "you will never be charged more than your total budget". There is no daily cap. For Search campaigns the window is 3 to 90 days. It is chosen when the campaign is created, and "you can't change the budget type after a campaign is created". Google says these campaigns "won't require a new learning period if you're using account goal settings", and the August 2026 bidding change doesn't apply to them. That suits a shoulder push with a known end, such as heating tune-ups and replacement estimates from mid-September to mid-November.

What this doesn't cover

  • The October move itself. The Comfort Authority figure comes from a drive-time and search-term rebuild in a capped summer. Moving July demand into October is not a result we can show.
  • Local Services Ads. After the move into Google Ads, Google divides your historical average weekly budget by 7 to set a daily average budget, and the ad schedule moves over (what else changes). Throttling LSA in a capped month needs its own piece.
  • Staffing and the paid social setup. Whether to hire for July, and how Way Cool's social campaigns were built, are outside this piece.

Do this in your account

  1. Build the two columns: calls run and slots, by month, for the last twelve months. Mark capped and open months.
  2. Lay ad spend next to them (Campaigns, segmented by month). If spend peaks in capped months and dips in open ones, your ad calendar runs opposite to your crew calendar.
  3. Sort ad groups into work that must run this week (no cool, no heat, leaks) and work that can wait (tune-ups, estimates on a running system, duct work). An account that keeps HVAC repair and replacement apart has most of this split already. Before a capped month, turn the second group down and plan it into the shoulder months.
  4. Check Target CPA campaigns for "Limited by budget" and set the target for the month you're in. If you cut schedule days, cut the budget by the same share.
  5. Open Tools → Budgets and bidding → Adjustments → Seasonal and review any adjustment longer than 14 days.
  6. Run each shoulder product through Keyword Planner with your service area and October as the date range. Near-zero searches means your customer list, not search.

FAQ

When is the HVAC slow season? Your field software answers it better than a calendar: it is every month where calls run fall short of the slots your techs can cover. In the Phoenix account above, cooling drops from late November for about four months. In Tampa, the crew is full from May to September.

Should I cut my Google Ads budget in the slow season? Cut what has no demand, then move money to what does. The Way Cool account raised its winter spend and put it behind a product that didn't need the weather.

Should I pause Google Ads for the winter? Google lists paused campaigns as a reason a bid strategy goes Inactive, and a reactivated strategy as a reason for Learning, which Google says typically takes one to two conversion cycles. If you need the campaign at full strength the week the heat arrives, we'd keep a reduced budget running instead.

Does turning off ads on weekends save money? Not by itself. With an ad schedule, Google still paces toward the full monthly limit, so lower the budget to spend less.

Can I use a seasonality adjustment for the summer? Not for the whole summer. Google says they are ideal for events of 1–7 days, may not work as well beyond 14 days, and that Smart Bidding already manages seasonal events.

Sources

All read in full on September 24, 2026. Every quotation above is from these pages.

Read next

Budget & season

How much should an HVAC company spend on marketing? Start from trucks and tickets, not a percent of revenue

A percent of last year's revenue spends the most in the month the board is already full. How to size the budget from open truck slots, check it against what a sold job is worth, and enter it in Google Ads without the weekday-schedule trap.

Dec 10, 2024 · 13 min read
Budget & season

A maintenance plan sold only at the kitchen table stops growing. Sell it to your own list first.

Your techs can offer the plan only in houses where something broke. The other signups can come from your customer list and from the tune-up and price searches people already make.

Sep 9, 2025 · 12 min read
Google Ads

Two shops, one Google Ads account: stop letting the account decide which truck gets the call

Two shops on one website don't bid against each other in the auction. Put them in one campaign, though, and they share one budget, one season, one schedule and one phone number, and Ad Rank settles the towns between them.

Jun 16, 2026 · 14 min read

Request your free strategy

One page, five details, no card and no call.

Would rather talk it through? Ask for a callback on the form and pick a window. The person who would run your account calls you, not a salesperson.

This sets the testing pace. There is no minimum.

A good callback window or the first problem you want us to look at

No card and no call. Read-only access only.