What AI Overviews actually changed on the results page
Walk through what a homeowner now sees when they search "AC repair near me" on a phone. At the very top, in most markets, is a Local Services Ads row with the Google Guaranteed badge and a few screened shops. Below that, Search Ads. Below those, on more and more queries, an AI Overview: a generated paragraph that answers the question and links out to a handful of sources. Then the map pack. Then the blue links, pushed further down the page than they have ever been.
Notice what moved and what did not. The AI Overview pushed the organic blue links down. It did not push the ads down. LSA still sits on top. Search Ads still sit above the map. If you were relying on ranking a service page in the top three organic spots, the AI answer just stole a chunk of your real estate. If you were running paid, your position on the page is essentially where it was.
That is the part the "is PPC dead" headlines get backwards. The channel AI Overviews threaten most is free organic clicks, because the model answers the question before the homeowner scrolls to a blue link. Paid slots are a different shelf. Google is not going to bury the ads that pay its bills underneath a free AI answer.
So the real question for a contractor is not "do ads still work." It is "now that fewer people scroll past the AI answer to the free results, is paid a bigger or smaller share of the leads I can actually reach." For most home-service trades, the answer is bigger. The paid slots are where the ready-to-book intent still lands after the AI answer has done its summarizing. You just have to be in them, and you have to make the click count.
Why the click that survives is worth more now
Here is the shift that matters on your P&L. AI Overviews absorb a lot of the low-intent, just-researching traffic. The homeowner who wanted to know "how long does a water heater last" now gets that answer in the AI box and never clicks anything. That was never a great lead for you anyway, and you were often paying for that click and getting a bounce.
What is left clicking through is more likely to be someone who already has a problem and is ready to hire. The tire-kickers got their answer for free. The person who clicks your ad after reading the AI summary skipped the research phase and is closer to picking up the phone. In other words, the click got more expensive to earn and more valuable when you earn it. The mix of who reaches your ad quietly improved even as the raw count of clicks may have dropped.
That cuts two ways for a contractor. The upside: a well-built call-only or LSA campaign is now catching a warmer, more decision-ready caller. The downside: you cannot afford to waste those clicks the way you could when volume was cheap and plentiful. A weak landing page, a phone that rings out to voicemail, a CSR who fumbles the booking, and you paid a premium price for a premium-intent lead and dropped it on the floor. The penalty for sloppy execution went up right alongside the value of the lead.
This is exactly why the old habit of watching cost-per-click is the wrong meter now. Cost-per-click going up is not automatically bad if the clicks convert to booked jobs at a higher rate. A shop obsessing over a rising CPC while ignoring its booking rate is reading the wrong gauge on the dash. The number that matters is cost-per-booked-job: what you spent divided by the jobs that actually hit the calendar. Run that math and a more expensive, higher-intent click often looks like the best money you spent all month. It also changes what you optimize: not the cheapest click, but the campaign, page, and phone process that turns the pricier clicks into the most jobs.
Local Services Ads: the slot AI can't summarize away
If there is one paid product built for this moment, it is Local Services Ads. LSA sits at the very top, above both the AI Overview and the Search Ads, and it carries the Google Guaranteed badge, which is exactly the trust signal a nervous homeowner wants when a machine just told them who to call.
Two things make LSA hold up as AI reshapes the rest of the page. First, its placement is above the fold and above the AI answer, so it is the first thing a searcher sees, not the last. Second, LSA is pay-per-lead, not pay-per-click. You are charged when a qualified lead contacts you, not for a curious tap. In a world where a chunk of the tire-kickers now get answered by the AI box and never reach a paid slot, a pay-per-lead product that only bills you for actual contacts is well positioned.
There is real work to get there. Google Guaranteed screening is a gate: license verification, insurance proof, background checks on the business. That screening is a barrier, and a barrier is a moat once you are through it, because the shop down the street that never bothered is not in the row. Getting the badge live and the profile answering leads correctly is paid-channel setup work, and it is separate from your Google Business Profile in the map pack (that is a different lane).
- Placement: top of page, above the AI Overview and the Search Ads.
- Billing: per qualified lead, not per click.
- Trust: Google Guaranteed badge, backed by Google's screening.
- Barrier: license, insurance, and background screening, which keeps casual competitors out.
It is worth being clear about the trades where LSA earns its keep hardest. The strongest fit is service work a homeowner books under pressure: no AC in August, a leak overnight, a garage door stuck with the car inside, a panel that keeps tripping. That searcher wants a screened, badged shop at the very top of the page and is not scrolling to compare ten options. For big-ticket project work with a longer decision, Search Ads and remarketing often carry more of the load, because the homeowner is researching, not dialing the first badge they see.
For most established home-service contractors, the smart sequence is LSA first for the high-intent local calls, Search Ads second to fill the gaps LSA does not cover. Which one runs first for your trade and market is a real decision, and it is the kind of thing worth ten minutes on a call rather than a guess.
The wrong conclusions contractors draw from the AI shift
When the "PPC is dead" articles started, plenty of contractors drew the exact wrong lesson and either killed their paid budget or doubled it in a panic. Both moves cost money. Here are the misreads worth naming so you do not repeat them.
The first is cutting paid because organic traffic dropped. If your website's free clicks fell after AI Overviews rolled out, the instinct to spend less feels logical: fewer visitors, tighten up. It is backwards. The AI answer ate your free traffic, which makes the paid slots above it more important, not less. Pulling paid at that moment hands the ready-to-hire searcher straight to the competitor still bidding.
The second is chasing volume that is gone. The homeowner who used to click through to read three articles before deciding now gets a good-enough answer from the AI box. That researcher was never going to book today anyway. Trying to buy that top-of-funnel volume back with a wider, cheaper campaign just pours budget into curiosity clicks. The intent that is left is narrower and hotter; aim at that, not at the volume you lost.
The third is ignoring remarketing. In an AI-era funnel, a homeowner may read the AI summary, click one shop, get distracted, and leave without calling. That person is not lost. Paid remarketing puts your name back in front of them across the web over the next few days, while the job is still on their mind. When the top of the funnel is thinner, staying in front of the warm ones you already touched matters more.
The clean lesson underneath all three: the AI shift did not shrink the paid opportunity for a local contractor, it concentrated it. Fewer, warmer, pricier clicks, plus a real reason to catch the ones who bounced. Read it that way and you spend smarter instead of spending scared.
How to build paid so the AI-era click converts
The mechanics of a paid campaign that works in this environment are not exotic. They are the same fundamentals that always separated a profitable account from a dashboard that just spends money, only now the margin for sloppiness is thinner because every click costs more and carries more intent.
Build the campaign around dispatch reality, not vanity metrics. A contractor's ad exists to make the phone ring during business hours in the service area, with a caller who has the job you actually want. That means call-only and call-extension campaigns for the trades where people pick up the phone in a panic, ad scheduling that matches when your crew and CSR are actually available, and geo-targeting tight to the service area so you are not paying for clicks from towns you will not drive to.
Then defend the budget with negative keywords. Home-service accounts hemorrhage money on searches for DIY instructions, jobs, parts, and warranty lookups. Someone searching "how to fix a garbage disposal myself" or "plumber jobs hiring" is not a customer, and every click from them is a paid dollar set on fire. A disciplined negative-keyword list is the single most common thing missing from an account a contractor built alone.
Last, the landing page. The ad's only job is the click; the page's job is the booking. A page built for ads is not your homepage. It matches the ad's promise, loads in under two seconds, puts the phone number and a call button where a thumb can hit it, and states the service, area, and why you over the shop next door before the visitor scrolls. When the click is warmer and pricier, a page that converts at a higher rate is worth more than a lower bid.
Get those four right, call handling, scheduling and geo, negatives, and the landing page, and a paid account earns its keep even as the top of the results page fills up with AI.
Where paid stops and the durable stuff starts
Being honest about what paid does not do is part of running it well. Google Ads buys you the top of the page today and stops the minute the card declines. That is the product, and it is a good product when your calendar needs filling this week. It is not an asset you own.
The AI Overview reshaping the page is also a reminder of a second front you cannot buy your way onto. When a homeowner asks ChatGPT, Perplexity, or Gemini "who's a good roofer in my area," that is not an ad auction. There is no slot to bid on inside a conversational AI answer. Being named there is earned through the way your business is described and corroborated across the web, and it lives in a different lane than paid. We build that too, but it is not something you turn on with a budget, and any agency telling you that you can bid your way into ChatGPT is selling you something that does not exist.
The clean way to hold both in your head: paid is the faucet for demand searching right now, and it runs above the AI answer where the ready-to-hire clicks still land. The durable visibility work, organic and AI-answer, is the asset that keeps naming you after the spend stops. They are not rivals for the same budget. Paid keeps this month full while the durable channel is built underneath it.
For most established contractors the answer to "do Google Ads still work" is therefore yes, run them, but run them as the fast lever, not the whole plan. Keep a paid floor live so the phone rings this week. Build the earned visibility so a year from now you are renting fewer of your leads. A shop that only rents stays on the treadmill. A shop that refuses to rent starves while the durable work lands. Run both on purpose, and neither trap is yours.
One last thing worth saying plainly, because the headlines will not: the AI shift did not fire your paid channel, it promoted it. The slots you can buy now sit above the answer everyone is talking about, and the buyers who reach them are more ready than they were two years ago. That is a better paid landscape for a well-run local contractor, not a worse one. The shops that read it as a death notice and cut spend are the ones handing their market to whoever kept running the ads and made the clicks count.