Local Business Google Ads Wasted Spend

A plumber in Sacramento spends $2,000 a month on Google Ads. His location targeting is set to a 25-mile radius. His keywords are five phrase-match terms, all with "plumber" in them. He has a decent landing page and call tracking. And roughly a fifth of his spend goes to people who were never going to call him.

They are not in Miami. They are in Sacramento. They are looking for a plumbing apprenticeship, or a plumber's hourly rate so they can argue with the one they already hired, or a YouTube video about replacing a wax ring. Location targeting caught all of them correctly. The problem was never geography.

That is the part the "stop wasting money on Google Ads" articles get wrong. They tell local advertisers to fix their radius, exclude other states, and turn off the Display Network. Those are real fixes and they take twenty minutes. Then the waste continues, because the waste in a local account lives in the intent of the query, not the location of the searcher.

Geography and intent are two different filters, and only one is automatic

Google gives you a reliable tool for geography and no tool at all for intent. Location targeting works — imperfectly, with its own quirks around "presence or interest", but it works. Nothing in the interface asks whether the person typing "dentist" wants a cleaning, a job, a school, or a picture for a slide deck.

So local advertisers spend their optimization time on the filter that already works. They tighten the radius from 30 miles to 20. They add city exclusions. They switch to "presence only". All fine. Meanwhile the search terms report has "dental assistant jobs sacramento" sitting at $86 in spend and zero conversions, and it has been there for four months.

The tell is simple: a term is a geography problem if a qualified buyer in your service area would never type it. A term is an intent problem if a person standing in your service area typed it and still will not buy. The second category is bigger, and it is the one nobody automates.

This matters more for local accounts than for national ones because of volume. A national ecommerce account running $80,000 a month has enough conversion data for Smart Bidding to learn which queries are dead and stop bidding on them. A local plumber with 14 conversions a month does not. The algorithm has nothing to learn from. It keeps buying the same bad clicks because it has no statistical reason to stop.

The five search-term patterns that drain local accounts

Across local service accounts we audit, the same five families of query show up again and again. Each one looks relevant enough to survive a quick glance at the report.

DIY and how-to

"how to unclog a kitchen sink", "fix running toilet yourself", "why is my ac blowing warm air". These people have a problem you solve and no intention of paying anyone to solve it. They are the most sympathetic group and the most expensive, because they search a lot and they click.

The nuance: some DIY searchers fail and call a pro two hours later. That is real. It is also rare enough that on a local budget you cannot fund it. If you have the data to prove that "how to" queries convert for you at an acceptable cost, keep them. Most accounts do not have the data, and the honest default is to block.

Negatives worth holding: how to, diy, yourself, myself, tutorial, step by step, youtube.

Employment

"plumber salary", "hvac technician jobs near me", "electrician apprenticeship", "how much do roofers make". Every trade account has these and they can be brutal, because "near me" makes them look geographically qualified. A job seeker in your city triggers every signal your campaign is built on except the one that matters.

We cover the full list in job seekers clicking Google Ads, but the core set is salary, jobs, hiring, apprenticeship, career, wage, pay, training, certification, license.

The exception: if you are actively recruiting and your careers page is the landing page, you may want some of these — in a separate campaign, with a separate budget, never mixed into the one that pays for lead generation.

Price research without purchase intent

"average cost to replace a water heater", "how much should a roof cost", "plumber hourly rate". Split these carefully. Someone searching "water heater replacement cost sacramento" is shopping and may well call you. Someone searching "average plumber hourly rate 2026" is checking whether the invoice they already have is fair.

The distinguishing signals are the word "average", the presence of a year, and the absence of any local modifier. "cost to replace water heater near me" behaves very differently from "average water heater replacement cost usa".

Wrong service, adjacent trade

A residential plumber picking up "commercial backflow testing". An HVAC company picking up "refrigerated truck repair". A residential roofer picking up "roof rack installation". Broad match produces these at a steady drip, and each one looks close enough to the business that it survives a skim of the report. If you are still on broad match in a small local account, the case for it in 2026 is worth reading before your next budget cycle.

Free, cheap, charity, grant

"free roof replacement program", "low income furnace replacement grant", "free legal advice". These convert at near zero for paid service businesses, and they carry high click volume because the words are emotionally loaded. Block free, grant, charity, low income, assistance program, government program unless one of them genuinely describes something you offer.

"Near me" is not the qualifier local advertisers think it is

"Near me" tells you the searcher wants something nearby. It tells you nothing about what they want to do when they find it. "plumbing jobs near me", "plumbing supply store near me", "plumbing school near me" and "emergency plumber near me" all contain the same two words and only one of them is worth $40 a click.

This is the single most common misread we see in local accounts. The advertiser scans the search terms report, sees the city name or "near me" in most rows, and concludes targeting is healthy. The geography signal masks the intent problem for the whole report at once.

A useful exercise: export your search terms, delete every instance of your city name, your state, "near me", and "in [neighborhood]" from the strings, then read what is left. "plumber near me" becomes "plumber". "plumbing apprenticeship sacramento" becomes "plumbing apprenticeship". With the geographic wrapper stripped, bad intent is obvious in a way it never is in the raw report. Half an hour of this on a $2,000/month account usually finds more waste than a week of bid adjustments.

The same trick works in reverse for suburb names. Local advertisers often add every surrounding town as a keyword and end up matching queries about those towns that have nothing to do with the service — "roseville city hall hours" against a keyword like "plumber roseville". Google's matching is looser than most people expect, and why Google shows irrelevant search terms explains the mechanics behind it.

Small budgets make every bad click cost more

A $2,000 monthly budget at a $12 average CPC buys about 166 clicks. If 20% of those clicks go to job seekers, DIY researchers and price checkers, you lost 33 clicks and $400. At a 6% conversion rate on the remaining 133 clicks, you got 8 leads instead of the 10 you would have got clean. Two leads a month, for a business where a lead is worth $300 in booked work, is $7,200 a year.

Run the same arithmetic on a national account spending $80,000 a month and 20% waste is a bigger number in dollars but a smaller problem in practice: the volume is high enough that Smart Bidding corrects for a lot of it, and the business has margin to absorb the rest. A local service business does not. That is the argument for treating negative keywords as a weekly habit rather than a quarterly clean-up, and how often you should check search terms goes into what cadence actually earns its time.

There is a second effect that is harder to see. Every bad click is also a bad conversion signal — or rather, the absence of one. Smart Bidding learns from what converts. Feed it 33 clicks a month of job seekers and DIY searchers, and the bid strategy is building its model partly on noise. On a small account where the whole model rests on 14 conversions, that noise is a meaningful share of the input.

Before you attribute a bad month to search terms, though, check the order of causes. Google Ads spending money with no conversions walks through what to rule out first — broken conversion tracking will make a perfectly clean account look like a disaster.

Performance Max is where local waste hides now

Google pushes local service businesses toward Performance Max hard, usually through the optimization score and through account reps who describe it as "Smart Shopping for leads". For a local advertiser, PMax has one structural problem: you can see far less of what you are paying for.

The search terms that PMax reports are grouped into categories, and a chunk of spend never appears as a query at all because it went to Display or YouTube placements that produce no search term. What the PMax search terms report actually shows covers the gap in detail. For a local account the practical consequence is that your "plumber salary" problem may be running inside PMax with no row in any report proving it.

You can still defend yourself. A Performance Max campaign accepts up to 10,000 campaign-level negative keywords. It does not accept shared negative keyword lists, which is the annoying part: if you run three PMax campaigns for one local business, your list of employment and DIY negatives has to be uploaded to each campaign separately and maintained in three places. Negative keywords in Performance Max and the 10,000 cap explains what fits inside that limit and what to do when the list grows.

Search themes make this worse if you are careless with them. A search theme of "plumbing services sacramento" invites Google to match broadly around plumbing in Sacramento, which includes plumbing careers in Sacramento. Search themes versus negative keywords in PMax is worth reading before you add any.

For a local service business with one location and one set of services, our honest view is that a well-negated Search campaign is easier to control than PMax and produces cleaner data. If you run both, where Performance Max wasted spend hides is the diagnostic to run first.

Building the negative list without strangling the account

The failure mode on the other side is real. A local account with 400 aggressive broad-match negatives can block more good traffic than bad. We see accounts where someone added free as a broad negative and lost every query containing "free estimate" — which is exactly the query a roofer wants.

Three rules keep this from happening.

Prefer phrase negatives to broad. Broad negatives behave unpredictably across match types and block far more than you expect. salary as a phrase negative kills "plumber salary" and "average plumber salary california" and leaves the rest of the account alone. The mechanics are genuinely counterintuitive and negative keyword match types don't work like positives is the short version.

Never block a term that has converted. Obvious, routinely violated, especially when someone builds a negative list from a generic "bad words for contractors" template found online. Before adding any negative, search your own last 90 days of search terms for it. If a query containing that word ever produced a lead, the negative needs to be narrower. PPCleaner enforces this as a hard rule — it never blocks a term that converted in the last 90 days — but if you are doing it by hand, the check is on you.

Watch for impression collapse. If impressions drop 30% the week after a negative upload, you over-blocked. Six signs your negative keywords are hurting performance lists the symptoms; the fix is usually two or three negatives, not the whole list.

One more thing specific to local accounts: keep your negatives in a shared list where you can. Search and Standard Shopping both accept shared negative keyword lists, so a single "employment" list and a single "DIY" list can cover the whole account and be edited in one place. PMax, as above, cannot use them. If you manage several local clients with the same trade — three plumbers in three cities — managing negative keywords across client accounts covers how to share the base list without copy-pasting mistakes between accounts.

What to do this week, in order

Thirty minutes, in this sequence, on any local account:

1. Pull 90 days of search terms, sort by cost descending. Not by clicks. Cost is the number that hurts.
2. Read the top 100 rows. Mentally delete the city and "near me" from each one. Mark every row where the stripped query has wrong intent.
3. Check conversions on each marked row before blocking. Any row with a conversion stays, no matter how odd it looks.
4. Add the survivors as phrase negatives to a shared list for Search and Shopping, and separately to each PMax campaign.
5. Write down the date. Come back in seven days. The report refills.

That last point is the whole problem with treating this as a project. It is not a project. New queries arrive every day, and a list you built in January is stale by March. How to find wasted ad spend in Google Ads covers the finding; the maintenance is the part that quietly decides whether a local account works.

If you want a broader sweep before you start, the Google Ads audit checklist you can run in an hour covers the settings-level checks — location targeting, network settings, conversion actions — that should be right before search terms are worth your attention. And if you are deciding whether this work is worth paying someone else to do, agency or do it yourself runs the numbers at local-budget scale.

Where this approach does not help

Search term hygiene fixes one category of waste. It does not fix a landing page that takes nine seconds to load, a phone number nobody answers after 5pm, or a conversion action that counts every form view. If your cost per lead is bad and your search terms are clean, the problem is downstream and no negative keyword will touch it.

It also cannot see terms Google does not report. The search terms report withholds queries below a volume threshold Google does not publish, and on a low-volume local account that withheld portion can be a meaningful share of spend. What the search terms report hides is the uncomfortable detail. You can block what you can see. The rest stays invisible, which is an argument for keeping match types tighter in small accounts than the advice you will get from Google.

And Display, Video and App campaigns report no search terms at all. If a local business is running a Display campaign for "brand awareness", negative keywords are not the lever. Placement exclusions are, and that is a different job.