Google’s interface is built to nudge you toward the easy button. Open a new account and you’re steered toward Smart campaigns or Performance Max with a friendly “let our AI do the work” pitch. For a lot of advertisers that’s fine. For a local home-service business spending its own money on leads, it’s usually the most expensive way to learn what doesn’t work.
This isn’t a “PMax is evil” rant. Performance Max is a genuinely strong product for the job it was built for — ecommerce and broad-funnel demand with rich product data. The problem is narrower and more honest: for most local service-area businesses, it’s the wrong tool for the job. Here’s what these campaigns actually do, why that hurts local lead-gen, the few cases where automation is fine, and what to run instead.
What Smart campaigns and Performance Max actually do
Both products trade control for automation. The difference is how much.
Smart campaigns are Google’s beginner product. You write a few lines of copy, set a budget, pick a goal, and Google handles bidding, placement, and most targeting. There are no real negative keywords with match types — only blunt “negative keyword themes,” and they don’t even catch close variants reliably, so junk you thought you blocked can still slip through. You can’t see a normal search-terms report. It’s designed so a busy owner can launch in ten minutes. The price of that simplicity is that you have almost no levers when it spends on the wrong things.
Performance Max is the full automated product. One campaign feeds Google’s machine learning your text, images, video, and a conversion goal, and the system serves ads across every channel Google owns: Search, the Display Network, YouTube, Gmail, Maps, and Discover. You give it “search themes” (now up to 50 per asset group) as hints, not commands. You don’t pick placements — Google decides, optimizes toward your goal, and historically reported back only in aggregate.
That last part has improved. Channel-level performance reporting rolled out across accounts through late 2025, a channel timeline view followed, and Search Partner Network placement visibility arrived in early 2026. So the old “it’s a total black box” line is now too strong — you can see the Search/Display/YouTube/Maps split. What you still can’t do is fully steer where it spends, and there’s still no true query-level search-terms report. The shared trait that matters: you are not steering, the algorithm is. That’s the whole pitch — and the whole problem.
Why that hurts a local home-service business
A service-area business has a specific shape: you want homeowners with real intent (“AC not cooling,” “water heater leaking,” “roof leak repair”) inside a tight geographic radius, right now. Automated campaigns are built to find conversions wherever they’re cheapest — which is rarely the same thing.
You lose negative-keyword control
This is the big one. In a normal Search campaign you block junk with surgical precision: “jobs,” “salary,” “DIY,” “how to,” “free,” “rental,” brand names you don’t service, the next town over you don’t cover. You add negatives at exact, phrase, and broad match and tune weekly off the search-terms report.
Smart campaigns only give you broad “negative keyword themes” with no match-type control. Performance Max lets you add up to 10,000 negatives plus brand exclusions — but here’s the caveat that matters most: those negatives only apply to Search and Shopping inventory. They do nothing to filter Display, YouTube, Gmail, or Discover, which can be 40–70% of PMax spend. So the very channels where your budget bleeds are the ones the negative list can’t touch. To control those you need placement, topic, and content exclusions, plus account-level placement exclusions — a different toolset entirely, and one the product doesn’t put in front of you.
Budget bleeds into Display and YouTube
In PMax, a meaningful share of spend lands on the Display Network, YouTube, Search Partner sites, and mobile-app placements — cheap impressions that look like “reach” but rarely produce a qualified home-service lead. Mobile apps and in-game inventory are a long-standing source of accidental clicks; the fix is real but manual — account-level placement exclusions and excluding mobile-app categories, not the negative-keyword list owners assume protects them.
Spam leads and opaque search terms
Less intent filtering plus less visibility equals more garbage in the pipeline: wrong service, out-of-area, price-shoppers, and outright spam form-fills. Because there’s still no real search-terms report, you’re often diagnosing lead quality blind — you feel the bad leads in your inbox before you can see the bad queries in the account.
It can cannibalize brand and your LSAs
Left unconstrained, PMax happily eats your cheapest, easiest conversions — people already searching your business name — and takes credit for them. Brand exclusions are the specific lever that stops this. It can also overlap with your Local Services Ads, so you pay to reach the same homeowner twice and your reporting overstates how well automation is “working.” A related 2026 best practice: use Customer Match exclusions so PMax stops re-paying to reach customers you already have, and optimizes toward new-customer acquisition instead.
If you want the deeper version of this argument applied to a real account, our Google Ads management work is built around exactly these control points.
A 2026 wrinkle: AI Max for Search
One more thing that complicates the usual “just run a clean Search campaign” advice. In 2026 Google began replacing Dynamic Search Ads with AI Max for Search — broad-match-style matching, DSA-style page targeting, and auto-generated assets layered onto regular Search campaigns. It’s the same automation creep, now reaching into the channel we usually call the safe one.
AI Max isn’t automatically bad, but for a local SAB it widens matching and can pull in the out-of-area, wrong-service queries you spent years negating. If you turn it on, run it with an aggressive negative list and watch the search terms closely; if you don’t need it, you can keep it off. The point is that “Search = controlled” is no longer a default you can assume — you have to set it.
Where automation is actually OK
Being honest cuts both ways. Automation isn’t the enemy — undisciplined automation is. There are real cases where it earns its place:
- Remarketing and retargeting. Re-engaging people who already visited your site is a defensible use of PMax’s reach across YouTube and Display.
- Scaling an account that already converts well. Once you have clean conversion data and a proven cost per booked job from Search, a tightly constrained PMax — brand exclusions, Customer Match exclusions, aggressive negatives, and placement/content exclusions — can extend reach.
- Smart Bidding inside a real Search campaign. This is the automation worth using early. Letting Google optimize bids with Target CPA or Maximize Conversions (Target ROAS rarely fits a pure lead-gen account) — while you keep keywords, match types, and negatives — is very different from handing over the whole campaign.
The common thread: automation works when it sits on top of strong conversion data and tight guardrails, not when it replaces strategy. No automated campaign is smarter than the signal you feed it, which is exactly why tracking comes first.
What to run instead for local lead-gen
For most home-service owners the better-spending stack is boring and effective — built in order, on a foundation of clean tracking:
1. Local Services Ads first (for most trades)
LSAs sit above everything else on Google, are pay-per-lead instead of pay-per-click, and carry the Google Verified badge — the trust mechanic that makes them convert for home services. (In October 2025 Google folded Google Guaranteed, Google Screened, and License Verified into that single “Google Verified” badge and dropped the old money-back guarantee, so don’t promise homeowners a refund that no longer exists.)
A few honest specifics for 2026:
- Cost per lead runs around $53 blended across home services in early-2026 benchmark data — roughly HVAC $51, plumbing $57, electrical $39 — but it swings hard by metro and competition, so treat those as orientation, not a quote.
- Credits, not disputes. Google retired the manual lead-dispute button in 2024 and replaced it with automated crediting. Crucially, it generally no longer credits “job type not serviced” or “out-of-area” leads — exactly the junk owners most want refunded. Budget as if most bad leads are not creditable.
- What drives LSA rank: review count and rating, response rate and speed, proximity, and your bid/budget. Reviews and responsiveness do most of the work.
- CPL is not the scoreboard. LSAs book at roughly 44% on average, so judge them on cost per booked job, not cost per lead. A $70 lead that closes at 40% beats a $40 lead that closes at 15% every time.
LSAs are also a different surface from your organic presence — the Local Pack, Google Maps, and your service-area Google Business Profile are earned, not paid, and owners routinely conflate the two. Both matter; they’re just different levers. For HVAC, plumbing, roofing, electrical, cleaning, and appliance repair, this is usually where the first paid dollar should go. We set these up as Local Services Ads campaigns.
2. A well-structured Search campaign next
Then capture the demand LSAs miss with a real Search campaign: tight ad groups by service and intent, exact and phrase match on your money terms, a living negative-keyword list, and ad copy matched to dedicated landing pages. Set location options to “presence”, not the default “presence or interest” — that one setting is a top cause of the out-of-area clicks owners complain about, because it serves people merely searching about your area from somewhere else. This is where control pays for itself: you decide exactly what you pay for, in exactly which radius.
Worth a mention: Microsoft Advertising (Bing) captures the same high-intent local demand at often lower competition, and marketplaces like Angi, Thumbtack, and Yelp are lead-reseller alternatives — useful to fill gaps, but you’re renting someone else’s audience and usually sharing each lead with competitors, which is why owned channels (LSAs + Search) come first.
3. Conversion tracking as the non-negotiable prerequisite
None of this — and especially none of the automation — works without clean measurement. The campaign can only optimize toward what it can see, and if it can’t tell a booked job from a spam form, it will buy you more spam. For home services, phone calls are the dominant conversion, so call tracking with dynamic number insertion and qualified-call conversions is the operational heart of it, not an afterthought. Then close the loop: Enhanced Conversions for Leads and offline conversion import push your booked-job and ticket-size data back into Google so automation optimizes toward revenue, not raw form-fills. That feedback loop is the real 2026 answer to “feed the algorithm clean signal.” Proper conversion tracking is the prerequisite for any automation, Smart Bidding included.
One compliance note, since this stack involves recording calls and buying leads in the US: call recording requires consent in two-party-consent states, and lead-buying rules shifted after the TCPA one-to-one consent rule was vacated in early 2025. None of it is a blocker — it just means your call tracking and lead intake should be set up with consent handled correctly from day one.
Frequently asked questions
Are Performance Max campaigns good for local service businesses? For most local service-area businesses, no — not as the main campaign. PMax spreads budget across Search, Display, YouTube, Gmail, Maps, and Discover. Channel-level reporting arrived in 2025–2026 so you can see the split, but you still can’t fully steer placements and there’s no true search-terms report. A local HVAC or plumbing company wants high-intent searchers in a tight radius, and a well-built Search campaign with tight match types, negatives, and presence-only geotargeting delivers that with far less waste. PMax can earn a role later — remarketing or scaling once you have clean conversion data and a known cost per booked job — but it’s the wrong first tool for local lead-gen.
What’s the difference between Smart campaigns and Performance Max? Smart campaigns are Google’s simplified beginner product: minimal setup, automated bidding and placements, and “negative keyword themes” instead of real negatives with match types (and they don’t reliably catch close variants). Performance Max is the full automated product: one campaign serving every Google channel from your assets and a goal. Both hand control to Google’s automation — Smart is the lighter version; PMax is the heavier one. Neither gives you the keyword-level control a Search campaign does.
Can I use negative keywords in Performance Max and Smart campaigns? Partially, with one big catch. PMax supports account- and campaign-level negatives (up to 10,000) plus brand exclusions — but they only apply to Search and Shopping inventory, not Display, YouTube, Gmail, or Discover, which can be 40–70% of spend. For those channels you need placement, topic, and content exclusions. Smart campaigns only offer broad “negative keyword themes” with no match types. A standard Search campaign lets you add unlimited negatives at exact, phrase, and broad match.
Should I run Local Services Ads or Performance Max first? For most trades, start with Local Services Ads, not PMax. LSAs are pay-per-lead, sit at the top of Google, and carry the Google Verified badge. Early-2026 cost per lead runs around $53 blended (HVAC ~$51, plumbing ~$57, electrical ~$39), but varies hard by metro. Judge them on cost per booked job — LSAs book around 44% on average — not on CPL alone. Layer in a tight Search campaign next; Performance Max comes later, if at all.
The honest bottom line
Smart campaigns and Performance Max aren’t scams and they aren’t useless — they’re optimized for a different business than yours. They reward broad reach and rich data; a local service-area business rewards control, intent, and geography. Pushed as the default for local lead-gen, they quietly trade your budget for the platform’s convenience — and in 2026 that automation creep now reaches into Search itself via AI Max, so the levers matter more, not less.
Run the stack that respects how homeowners actually buy: LSAs first, a tightly controlled Search campaign second, automation only on top of clean conversion data and a real cost per booked job. If your account is leaning on Smart or PMax and the leads feel thin or junky, that’s usually a control-and-tracking problem, not a budget problem. Fix my Google Ads — we’ll audit where the spend is really going and rebuild around the levers that protect local budget.