Google Ads for foundation repair contractors

By Ahmed Imran · Updated August 2026 · 9 min read

Most foundation repair accounts I am handed are not short of budget. They are short of the right traffic, and they are counting the wrong thing as a lead. One contractor I work with went from $93,078 to $312,390 in quoted work in six weeks, while I cut 61 percent of the impressions.

What a foundation repair lead costs in 2026

Foundation repair sits at the expensive end of home services, because the job is worth four to five figures and every contractor in the metro knows it. Here is where the public benchmarks land this year, and where the accounts I actually run land.

That third row is the one nobody publishes, and it is the only one that matters. A $150 cost per lead looks excellent right up until you find out that half of those leads were somebody tapping your phone number and hanging up. The account I rebuilt this year reports $552.01 per quotable lead. That is higher than every benchmark above, and it is the honest number, because a human at the client marked each one.

The three problems in almost every foundation repair account

I have yet to be handed a home services account where at least two of these were not true. None of them are exotic and none need a rebuild from scratch.

1. Performance Max is eating the search budget

Performance Max is not the problem. Performance Max with no guardrails, buying display banners and YouTube views for a contractor who serves four counties, is. In the account I mentioned, two Performance Max campaigns produced 2.6 million impressions in 39 days. One of them took 716 clicks in a single week and generated $5 of quoted work.

Meanwhile the search campaign that was actually producing quotes was losing 43 percent of its available searches to budget. The money was in the room. It was pointed at the wrong half.

2. A phone tap counts the same as a quote request

This is the quiet one. Google was recording a tap on the phone number, a click for driving directions, and a completed quote request as the same kind of conversion. Smart Bidding then optimised toward whichever was cheapest to produce, which was the tap. The algorithm was doing its job perfectly. It had just been told to chase the wrong thing.

3. Competitor brand names are not blocked

Every foundation repair market has four or five big names that spend heavily on brand. If you are not blocking them, you are paying to show up for people who already decided to call somebody else. I found fourteen competitor brand names taking roughly $770 a month across the search campaigns with not one lead to show for it. They had been blocked in one Performance Max campaign only, and set to exact match, so they never caught the real searches.

What I changed, in order

The order matters more than the individual changes. I do not move budget until the numbers mean something, because before that I would just be guessing faster.

• ›Fixed the definition of a lead first. Demoted phone taps and direction clicks out of the bidding signal, and made the call tracking platform the single source of truth, with the client's own sales team marking every lead quotable or not.
• ›Cut Performance Max in stages rather than all at once, so nothing tripped a learning reset. One campaign was reduced, then switched off entirely after three weeks of zero conversions.
• ›Moved that budget into the search campaign that was losing nearly half its available searches to a budget cap.
• ›Built two search campaigns for services the client sells and had never advertised properly. One produced a $10,000 quotable lead on its first day.
• ›Repointed the landing pages. An entire campaign was sending drainage searches to a basement waterproofing page, so somebody looking for a french drain landed on a page about wet basements and left.
• ›Built a 26 term competitor blocking list and applied it to all six search campaigns, plus negatives for the wrong trades, out of area searches, and people shopping for materials rather than a contractor.
• ›Set the ad schedule from the client's own lead data by hour, after checking impression share by hour first, so it was a decision about demand rather than a guess about it.

What happened in six weeks

Both windows below are 39 days. Before is the period immediately preceding the engagement, after is the first six weeks of it. Same account, same market, same season.

Impressions fell 61 percent and quoted work rose 236 percent. Those two facts in the same sentence are the whole story. The account was never short of exposure. It was short of the right exposure, and it was paying full price for the difference.

Fewer leads is usually the right trade

In the final fortnight the share of leads worth quoting went from 40 percent to 69 percent. In raw count that is 20 leads down to 16. A lot of contractors would read that as a step backwards.

It is not. Every junk lead costs somebody in your office fifteen minutes, and it costs your sales team the belief that the phone ringing means anything. Sixteen leads where eleven are real beats twenty where eight are.

How to structure a foundation repair account

• ›One campaign per service line, not one campaign for everything. Foundation repair, basement waterproofing, drainage, crawl space and concrete leveling are different searches with different intent and different job values.
• ›Every ad group points at the page that answers that specific search. If somebody searches for a french drain, they should not land on a general waterproofing page.
• ›Keep Local Services Ads reporting separate from search. It bills per lead on a completely different model, and folding the two together flatters your cost per lead in good weeks and punishes it in bad ones.
• ›Run a competitor blocking list at phrase or broad match, not exact. Exact match negatives almost never catch the real searches people type.
• ›Feed quote values back into the account, so bidding optimises toward the jobs worth having rather than the leads that are cheapest to generate.

What to do first, this week

Open your conversion actions and look at what is set to primary. If a phone tap, a directions click and a submitted quote form are all sitting there as equals, that is your first fix and it costs nothing. Then pull a search terms report for the last 90 days and read it. Not skim it. Read it. The waste is almost always visible in the first two screens.