Seasonal campaigns and when to turn them on

The other half of the rule is the harder one. Turn it off when the season ends, on a date you decided in advance, not when you notice the leads got expensive.

Seasonal spend has a shape. Most local advertisers get the shape wrong in both directions, arriving late and leaving late, which is the most expensive possible combination.

Why does turning on early beat turning on during the rush?

Two reasons, one about customers and one about the platform.

Customers research before they buy. The person who calls a roofer in the first heavy rain looked at roofers a few weeks earlier when the forecast changed. Housecall Pro's 2026 Home Services Report, combining a homeowner survey with its own first-quarter platform data, found that a sudden problem is what most often makes a homeowner pick up the phone at 58%, followed by a smaller issue getting worse at 51%, routine maintenance at 47% and planned projects at 38%. That is vendor research from a company that sells field service software, so label it as such. What it says about timing is that roughly half of the demand you are buying was already deteriorating before it became a call.

The same report found 72% of homeowners would pay more to resolve an emergency within 24 hours, and 68% get additional quotes when pricing comes in higher than expected. Urgency raises what people will pay and it does not stop them shopping. Showing up two weeks before the spike is how you become one of the names being shopped rather than the third quote.

The platform reason matters just as much. A campaign that has just been switched on has no recent performance history. Google's own documentation notes that some Smart Bidding strategies rely on a minimum volume of historical conversion data, and that automated bidding needs time to adjust. That means the early days of any campaign are the least efficient days. If those days coincide with your busiest fortnight, you burned your peak on a learning period.

Start early, spend modestly, and let the campaign be smart by the time the money matters.

What does arriving late actually cost?

Put a number on it, because "start early" is easy to nod at and easy to skip.

Define the season. Say your peak runs six weeks. That is 42 days.

Define the learning stretch. Call it the first 10 days of a cold campaign, which is conservative for an account with low conversion volume.

Now compare two starts. Turn on three weeks before the peak at a modest $20 a day and you spend $420 derived buying data at off-peak click prices. You arrive at day one of the season with a campaign that already has history.

Turn on at the start of the peak instead and those same 10 learning days land inside the 42, which is 24%derived of your season spent teaching the account at peak prices.

Price the difference. If your peak budget is $4,000 across those 42 days, 24%derived of it is $960 derived spent during the least efficient stretch of the campaign, at the moment clicks cost the most. The $420 you would have spent in advance is cheaper, and it is spent when a mistake is affordable.

Rerun it with your own season length and your own peak budget. The conclusion holds under most inputs, and where it does not is a useful thing to discover on paper rather than in June.

How do I know when my season actually starts?

From your own records, not from what the season is called.

Open your invoices from the last two or three years and count jobs by month. Your season starts earlier than you remember, and it usually has a leading edge of quotes that convert weeks later. That quote date is the date you care about, not the invoice date.

Then check search interest for your main term by month using Google's own tools. In the accounts I open, search demand leads the work by a few weeks fairly consistently, though I want to be clear that is an operator observation rather than a published finding. There is no dataset that publishes seasonal demand curves for local trades, and anyone selling you one is selling a guess. Take the earlier of your two dates and subtract another two weeks. That is your turn-on date, and it goes in the calendar now for next year.

Bay Area weather makes this less obvious than in places with real winters. Demand still moves, it just moves with rain, heat waves, and the fire and holiday calendars rather than with snow. Hawaii shifts on humidity, storm season, and the visitor calendar. The pattern is in your invoices either way.

None of this is readable without conversion data in the account, which is another reason tracking goes in before the first dollar of spend. A seasonal decision made from click data is a guess.

What should I actually change when the season turns?

Budget, schedule, ad copy, and what the page emphasizes.

Budget first, and schedule the change rather than making it. Google spends against an average daily budget, and its documented monthly ceiling is that daily figure multiplied by 30.4. So a seasonal increase is a daily-number decision, and raising it on the first hot day means the increase lands after the rush started. Set it in advance.

Then the hours. Google's ad scheduling controls which days and hours your ads run, and peak season is exactly when a mismatch costs the most. ServiceTitan's June 2022 analysis of its own home services platform found that booking rates vary by time of day with mornings highest, and that HVAC booking rates move with the season. That is vendor platform data rather than independent research, but it is observed booking behaviour. The hours your budget covers are not interchangeable.

Hours matter on the organic side too, and the interaction is easy to miss. Whitespark's 2026 Local Search Ranking Factors report, in which 47 local search experts scored 187 factors, has "Business is Open at Time of Search" entering at fifth among local pack factors. Darren Shaw, who runs the survey, describes businesses dropping off the local results in the hour before they close. Grade that honestly as expert opinion rather than test data. The practical version is that extending your advertised hours for a season only works if you are genuinely open, and Shaw's own warning is not to misstate hours for a business customers might show up at.

Ad copy should name the season plainly. People searching in a heat wave respond to language about this week, not to your general service description. In the shoulder season the winning message is about booking ahead at a calmer time.

The page has to keep up. A seasonal spike is exactly when a slow page costs you the most, because the traffic is impatient and often on cellular. The relationship between speed and conversion gets exaggerated by vendors, and separating the page speed myth from the measurement is worth doing before you spend on a rebuild. For a testable target rather than a scary number, use Google's Core Web Vitals thresholds: Largest Contentful Paint at or under 2.5 seconds, Interaction to Next Paint at or under 200 milliseconds, Cumulative Layout Shift at or under 0.1, all assessed at the 75th percentile of visits.

Trust content matters more in peak season too, because urgent buyers are calling strangers. Licensing and insurance presented as trust signals close the gap faster than any headline when somebody is choosing between three companies they have never heard of.

How do I handle the offseason without going dark?

Keep a small always-on presence and let the seasonal campaigns carry the swing.

Going completely dark costs you more than it saves. Your brand searches continue year round, and going quiet is exactly when a competitor buying your name gets a free run at your customers. Competitors bidding on your brand name is cheap to defend against and expensive to ignore during the months you stopped paying attention.

Run the swing with scheduled changes rather than manual ones. Google Ads supports automated rules that change budgets, bids and campaign status on a schedule you set in advance, which handles the on and off dates without depending on you remembering during your busiest week. Which scripts and automations are worth using covers the short list that earns its setup time for a small account.

Use the offseason for the work you cannot do when busy. Rewrite the landing page. Clean the negative keyword list. Gather reviews. Those all pay off during the next spike.

Where does the early-start rule break down?

In three cases, and each one is common enough to check before you build a calendar around the rule.

When your peak is a single event rather than a season. A holiday weekend or a one-day promotion does not have a leading edge worth buying. Ramping three weeks ahead of a two-day spike just funds three weeks of low-intent clicks.

When your season is your only demand. A business that does nothing for eight months has no history for the platform to carry across the gap anyway, and the pause resets much of the advantage an early start buys. Start early in relative terms, but expect to pay the learning cost each year rather than compounding it.

When you are already booked at peak. Advertising into a full calendar is buying leads you will disappoint. The correct move is to advertise into the shoulder months where the capacity is, or to raise prices at peak and advertise nothing.

What to do this week

Pull three years of invoices and count jobs by month. Find the first month where volume climbs, then subtract three weeks. Put that date in your calendar with a reminder.

Run the four-line arithmetic above with your own season length and peak budget so you know what arriving late costs you specifically.

Write next season's ad copy now, while you have time to think. Save it in the account as a paused ad group so it is ready to enable rather than ready to write.

Set an end date at the same time you set the start date, and build both as scheduled rules rather than reminders. A campaign with no scheduled end runs until someone notices the leads got expensive, which is usually a month of wasted spend later.

Questions about your seasonal timing?

Email me at eric@seod.com with your trade, your city, and the two or three months you consider your busy season. I will send back the turn-on and turn-off dates I would put in the calendar, what I would change in the ads at each one, and the learning-cost arithmetic run on your numbers.

If your business turns out to be flatter than you think, I will tell you that and save you the work. That answer takes me a few minutes.

Or keep reading more on paid search for local businesses.