Here’s a pattern I run into constantly, and it costs more than almost anything else I look at. Demand for home services moves in waves through the year. Ad budgets, meanwhile, tend to be a single number that got set once and never revisited. The two don’t line up, and the mismatch is invisible in every standard report.
September is a good time to bring this up, because for a lot of trades the fall stretch is the strongest run of the year and it’s already starting.
Why flat budgets quietly cost money
A flat daily budget behaves very differently in a busy month than a slow one. In your slow season, demand is thin, so your budget comfortably covers everyone searching. You show up for nearly every relevant search and you might even have money left over.
In your peak, the same budget runs out partway through the day. From then until midnight, your ads stop, and the searches keep happening. You don’t get an alert about this. The report just shows a slightly higher cost per lead and the same spend as always, which reads as “steady” rather than “we were closed for half of every day during our best month.”
This isn’t an oversight on anyone’s part. Nothing in the platform prompts you to revisit a budget seasonally, and the default reporting view compares you to last month rather than to available demand. The tool isn’t built to surface it.
The one number that tells you it’s happening
In Google Ads there’s a metric called search impression share lost to budget. In your campaigns table, click Columns, then Competitive metrics, and add “Search lost IS (budget)” along with “Search impression share.”
The translation matters more than the name. Impression share lost to budget does not mean customers chose a competitor. It means your daily budget ran out and Google stopped entering you into auctions you were otherwise eligible for. Those are searches you were qualified to win and simply weren’t present for.
Read it like this. Under about 10% is normal noise. Between 10% and 25% is worth a look at your peak weeks. Above 25% during your busy season means the budget, not the market, is capping your lead volume, and adding spend there is usually the cheapest lead you’ll buy all year.
There’s a sibling metric, impression share lost to rank, which is a different problem entirely. That one is about ad relevance, bids and landing page quality, and no amount of budget fixes it. Worth knowing which one you’re looking at before you act.
Finding your actual demand curve
Don’t guess at your season, and don’t borrow someone else’s. You have three sources and they take about half an hour together.
- Your own booked jobs by month, for the last two or three years. Count jobs, not revenue, so a couple of big tickets don’t distort the shape. This is the most trustworthy signal you have because it’s yours.
- Google Ads conversions and impression share by month over the same period. Segment by month rather than looking at a single total.
- Google Trends for your main service term, restricted to your state, over five years. This shows demand rather than your capture of it, which is useful for spotting a peak you’ve been missing entirely.
Put the three side by side and the shape usually jumps out. Where booked jobs rise but your impressions don’t, you were budget-capped. Where searches rise and your booked jobs don’t, something after the click is losing them, and that’s a different conversation.
Worked example
Take a drainage and waterproofing company on $6,000 a month, flat all year, at a $120 cost per lead. That’s 50 leads a month, 600 a year, and at a 30% close rate and a $6,000 average job, roughly $1.08 million in booked work.
Now look at the seasonality. Impression share lost to budget sits near 4% from January through July, then jumps to 34% across September, October and November. Those three months are the ones where the budget, not demand, set the ceiling.
Shift the shape without changing the annual total. Move $1,500 a month out of the four slowest months and add $2,000 a month to the three peak months. Same $72,000 for the year. In the slow months you lose very little, because you were already covering nearly all available demand. In the peak months you buy roughly 16 extra leads a month at a similar cost per lead, so about 50 more leads, about 15 more booked jobs, and about $90,000 in additional work for no additional spend.
I’ve rounded these numbers to keep the arithmetic readable, and your close rate and job value will differ. The point is the mechanism, not the exact figure. Reshaping a budget you’re already spending is the rare change that doesn’t need a bigger check.
How to reshape it without breaking things
A few practical notes. Move budget in steps of 20% to 30% rather than doubling overnight, because big jumps push smart bidding back into a learning period right when you need it steady. Make the increase before the peak starts, not three weeks into it, since the campaign needs a little runway. And check your capacity first, because more leads than you can schedule turns into slow callbacks, which costs you more than the leads were worth.
Then decide what you’re measuring. In peak season, cost per lead usually rises a little, and that’s fine. Total booked jobs at an acceptable cost per booked job is the thing to watch. If you only look at cost per lead you’ll talk yourself out of a good decision.
What good looks like
- You know your two or three strongest months by booked job count, from your own records rather than a hunch.
- Search lost IS (budget) sits in single digits during your peak weeks.
- Budget gets reviewed on a calendar, at least quarterly, with a note of what changed and why.
- Your seasonal increase lands two to three weeks before demand climbs, not after.
One thing to do this week
Open Google Ads, add the two impression share columns, segment by month, and look at your last two peak seasons. If lost impression share to budget was above 25% in those months, you’ve found the cheapest growth available to you right now.
If you’d like help mapping your demand curve against what you’ve been spending, that’s part of the free marketing audit I run. And if you want the wider list of numbers worth watching month to month, these are the seven I’d start with.