Google Ads for auto repair shops: 376,400 searches a month from drivers trying not to pay you!

PPC for auto repair fails in a way no other trade's does: the wasted clicks are your own customers.

On nearly every industry page we publish, the traffic you're throwing away belongs to somebody who was never going to buy anything. A student. A job hunter. Somebody looking for a movie. Auto repair doesn't work like that. The 169,000 people a month searching what the check engine light means all own a car that's telling them something is wrong. So do the 103,800 typing a code, or a list of codes, straight into the box. So do the 66,000 standing in a lot at 7am asking how to jump start a car. Every one of them is a driver with a broken vehicle and a phone, doing whatever it takes not to pay a shop.

That's a harder call than blocking a film title, and it still comes out the same way: block them. Not because they're worthless. Because your budget is small and the distance between reading a code on a phone and handing you a key is enormous. Meanwhile 130,000 people a month type “auto repair near me”, and that click runs about $1.80. There's an exception, it's real, and it comes with a condition - the playbook spells out both instead of pretending the question is simple.

Now the good news, and shop owners consistently underrate it. The terms that carry your account sit at or under $2.00, which makes them the cheapest set of buyer terms on this site, taken as a whole. $40 a day buys around 22 of them. A plumber spending that same $40 on emergency work gets 4 clicks and a month of noise. You get enough traffic to learn something before the month's out, and that changes the first move: build the thing properly instead of tiptoeing.

AdPilot connects to the account and reads it end to end, the way a decent consultant would before touching anything. Spend that was never going to become a repair order. Hours the ads run into a dark counter. Ads pulling clicks that fill no bays. Each finding turns into a card, written the way you'd explain it to somebody leaning on the counter, and you approve it or you skip it. Nothing lands in the account until you do.

You decide every step of it:

• You approve every change. The AI proposes; you decide. Every budget, bid, keyword, or ad edit arrives as a plain-language card you approve or skip with one tap. Nothing is ever applied on its own.
• A snapshot before every change. AdPilot saves exactly how things looked before anything is touched, so you always have a clean “before.”
• One-click revert. Don’t like a change once it’s live? Undo it in a single click, straight back to that snapshot.
• A full audit log. Who, what, when, the old value and the new value, every change recorded. Nothing hidden.
• You keep the keys. You connect with one secure Google sign-in, and it only ever makes changes you approve. The account stays yours start to finish - you never transfer ownership, and you can disconnect anytime.

Where auto repair ad budgets actually leak

US monthly volumes, Ahrefs, August 2026. Every row is a word repair shops bid on, and none of them ends with a key on your counter.

Split them and the shape shows up. 180,000 a month are the check engine light itself, asked short and asked long. 103,800 are codes and code lists. 83,600 are people fixing it on the driveway. 9,000 want to know what a mechanic earns. That's 376,400 searches a month, and any row takes about a minute to verify. That's deliberate. You should be able to check the argument before you believe a word of it.

Now the uncomfortable part. On the lawyers page the junk is a TV show and blocking it takes two seconds. Here, the person who typed “p0171” has a car running lean, and plenty of them end up in somebody's shop inside the month. You're still going to block them. The playbook shows why the arithmetic says so - but block them knowing what you're blocking, and read the exception before you decide it applies to nobody.

US monthly search volume, Ahrefs, August 2026. 376,400 searches a month, and the awkward truth is that most of these people own a car that needs work.

Negative keywords for auto repair shops. Free, no email, no form.

Take it, whether or not you ever pay us a dollar. Ten minutes at the counter, and on a $40 day it buys back more clicks than any bid you could touch this month. Every entry is safe to paste as it stands with one exception, and it's marked: the group headed “Work you refer out” is written for a mechanical shop. If you're a body shop, a glass shop or a tire store, that group blocks your own front counter - go through it and delete the lines that describe your work before it goes anywhere near a live account. Then read the paragraph under the list. Several of these words have a twin that books a car.

Doing it themselves

how to, how do i, how do you, diy, do it yourself, myself, tutorial, step by step, youtube, video, repair manual, haynes, chilton, wiring diagram, fuse box diagram, serpentine belt diagram, torque specs, oil capacity, what oil does, socket size, how to jump start, jump start a car, jumper cable order

Reading the code instead of clearing it

p0420, p0171, p0300, p0455, obd2, obd2 codes, obd scanner, code reader, scan tool, elm327, check engine light mean, what does check engine light, check engine light codes, reset check engine light, clear check engine light

Parts and parts stores

auto parts, auto parts near me, parts store, autozone, o reilly, oreilly, advance auto, napa auto parts, rockauto, used parts, junkyard, salvage yard, pull a part, buy brake pads, brake pads for sale, oil filter for, spark plugs for sale

The career, not the customer

mechanic salary, mechanic salaries, auto mechanic salary, mechanic pay, mechanic jobs, mechanic job openings, technician jobs, now hiring, hiring mechanics, apprentice mechanic, become a mechanic, mechanic school, automotive trade school, ase practice test, ase study guide, mechanic resume

Work you refer out - delete any line you actually sell

body shop, paint job, dent repair, windshield replacement, window tint, car wash, detailing, upholstery, towing near me, roadside assistance, engine swap, dyno tune, rv repair, semi truck repair, motorcycle repair

Somebody else's paperwork

is it covered by warranty, does warranty cover, carshield, recall lookup, recall notice, is my car under recall, dealer covers it, goodwill repair, class action

Buying and selling a car, not fixing one

cars for sale, used cars for sale, sell my car, trade in value, kbb value, junk my car, cash for cars, dealership near me, car lease, car loan

Games, definitions and browsing

simulator, car mechanic simulator, forza, gta, meaning, definition, wikipedia, reddit, images, pictures, memes

The check engine light, your tech count, and five other things that beat fiddling with bids

• Block the check engine light. Then read the exception, because it's real. 180,000 searches a month are the light itself, and on $40 a day you can't afford them. The person typing it wants a free answer, and the whole first page of Google is built to hand them one before they ever pick up a phone. Block it and put the money on the 130,000 who typed “auto repair near me” at $1.80. Now the exception, and it comes with a condition. If your near-me campaign is already fully funded - not budget-limited, actually converting, calls coming in - and you have bays sitting empty, and you run a genuinely free scan you can afford to honor, then the symptom searches become the cheapest way to fill those bays. In that order, never before it. A shop that bids the light while its near-me terms run out of budget by 2pm has moved money from a booking to a lookup.
• Count your techs before you raise the budget. Four bays and two techs is six to ten repair orders a day, and the ceiling is the techs, not the bays - a fifth bay with nobody standing in it turns nothing. Work out your own number before you touch the budget: count what you actually turned out last week, brakes and diagnostics included, not what the shop could do on a day of oil changes. Now the part most marketing advice for shops skips. A campaign that fills your schedule on Tuesday and keeps spending through Friday is buying phone calls you have to turn away, and a driver you turn away calls the next shop and stays there. If the honest answer is “we're booked out ten days”, more clicks is the wrong lever. A higher average repair order is the right one, which means bidding brakes, diagnostics and AC rather than oil changes.
• Two dollars a click changes what you should do first. Every term that carries an auto repair account sits at or under $2.00, which no other trade on this site can say - the roofers page has one term at $0.70, but the ones with the volume run to $9.00. “Auto repair near me” is $1.80, “auto repair shop” is $1.60, and only the specialty work climbs - “european auto repair” tops out around $2.50. At $40 a day that's roughly 22 clicks, which clears the ten-clicks-a-day floor a new campaign needs before its numbers mean anything, twice over. A plumber pays $10.00 for an emergency click and gets 4. So the advice that fits every other trade page - narrow it, be patient, expect a thin month - doesn't fit you. Build it properly and you'll have a month worth reading in four weeks.
• The coupon crowd is real, and it isn't automatically yours. “Auto repair coupons near me” runs 27,000 searches a month at about $1.60. Those are customers, not tire kickers, and saying otherwise would be dishonest. They're price-led, though, so bidding them only adds up when two things are true at once: you have an offer you can honor without gutting the ticket, and you have capacity to fill. Booked solid through next week? Skip it. Slow February and a $29 oil-and-inspection special? Be all over it.
• Run AC on a season, not on a calendar year. “Auto ac repair near me” is 8,300 searches a month at about $2.00, and nearly all of that demand lands between the first 90-degree week and the end of August. Running it in February spends money on nobody. Give it its own campaign so it gets its own budget and its own on and off dates, and write the ad for July - warm vents, a hot cabin at pickup, cold air by Friday. Same logic in reverse for batteries and no-starts when the first hard freeze arrives.
• The car breaks at 7pm and your counter is dark. This is the scheduling problem specific to your trade. A plumber can answer at 2am and go. You can't open a bay at 9pm, but the alternator still dies on the drive home, and the driver is searching right then. Two honest options: the ads run into the evening and the landing page carries a booking form that says plainly “first call tomorrow at 7:30”, or the evening comes off and that money goes into the morning, when people book. What you must not do is run 8pm ads to a phone that rings out. That driver calls the shop that answered, and you paid for the introduction. The hour that catches everyone out is Sunday evening. You're closed, the phone's off, and it's the best booking window of the week - the driver who's been ignoring a noise since Wednesday finally sits down and sorts out Monday. Cut Sunday off the schedule and you've handed Monday morning to the shop with an online booking form. Add a call extension too, and schedule it separately from the ads: the ad can run at 9pm with the phone number hidden, so nobody rings out and blames you for it.
• Glass, body and European are three different businesses. “Auto glass repair” is 23,000 searches a month at $2.00 and “auto body repair” is 18,000 at $1.60, and in a keyword tool both look like your traffic. Neither one is, unless you do that work - a mechanical shop bidding glass buys chip repairs it has to hand to somebody else. European specialty runs the other way. “European auto repair” is 7,300 a month at about $2.50, the priciest click on the list, and if you're the shop in town with the software for it, that's the most profitable ad group you can build. Bid the work you do. Block the rest by name.

What one phone call is worth to the shop

Repair shops get the cheapest set of buyer terms on this site, which makes this arithmetic friendlier here than anywhere else. That price is measured, not assumed: “auto repair near me” is 130,000 searches a month at about $1.80 and “auto repair shops near me” 35,000 at the same price, from the same pull as the table above. The repair order value and the booking rate are a plausible independent shop and nothing more, so replace them with yours. Nothing leaves your browser.

The method, start to finish

AdPilot's agent works in the order below, and here it all is, so you can run the thing yourself with a laptop and a Google account. Two reasons for publishing it. Anybody who follows it ends up with an account that works, and a method you can audit beats one you're asked to take on trust.

A weekend to build it. Then an hour every other week, for as long as the ads run. Building it is the easy half. The hour is the one that vanishes the week the alignment rack is down and a transmission job ate Thursday - and the search terms pile up whether you read them or not.

Not a repair shop? Every trade gets its own version of this

1. 1

   Sort the drivers from the readers

   Twelve of the biggest auto searches in the US belong to people who'll never book anything, and most of them are your customer dodging you. Together that's 376,400 a month: 180,000 on the check engine light, 103,800 on codes and code lists, 83,600 fixing it on the driveway, and 9,000 pricing up the career rather than the repair. “Check engine light” on its own is 169,000. Cheap clicks, every one of them, and cheap is the trap: they'll clear a $40 day before a paying search gets near the page.

   The buying half is a far shorter list, and the prices on it are almost embarrassing. “Auto repair near me” is 130,000 a month at about $1.80. “Auto repair shop” is 61,000 at $1.60, “auto repair shops near me” 35,000 at $1.80, “auto repair shop near me” another 16,000 at $1.80. Seasonal and specialty: “auto ac repair near me” 8,300 at $2.00, “european auto repair” 7,300 at $2.50. Intent comes off the words themselves. “Near me”, your town, “shop”, “cost”, “estimate”, “same day”, “open now” - car keys in hand. “How to”, “what does”, “p0420”, “salary” - somebody who wants your knowledge for nothing.

   Then one last test before anything goes in. Read the term the way a stranger reads it, with nobody there to explain what your shop does, and ask what else it drags along. Bare “auto repair” is 104,000 a month at $1.50 and it looks like the bargain of the set, right up until you notice it carries repair manuals, shops for sale and the whole driveway crowd with it. Add the qualifying word or leave the term out. One-word keywords have no business in a small account, and no match type saves them: phrase carries every other meaning those words have, and broad only half-guesses your intent off the landing page.

   • Name the person behind the words: what are they holding when they type this, a phone or a socket wrench?
   • Buyer words: near me, open now, same day, your town, brake job, inspection, estimate, what will it cost.
   • Reader words: how to, what does, why is my, DIY, salary, torque spec, and a bare five-character code starting with P.
   • Two extra words usually halve the price and double the intent.
   • A big number tells you what a word costs you. It never tells you what it's worth.

   Ten real terms - four to buy, six to block - with the volumes and the click prices

   • auto repair near me - 130,000/mo, ~$1.80 - BUY. The backbone of the account.
   • auto repair shop - 61,000/mo, ~$1.60 - BUY, and cheaper again.
   • auto ac repair near me - 8,300/mo, ~$2.00 - BUY, June to August only.
   • european auto repair - 7,300/mo, ~$2.50 - BUY if you have the software for it. Priciest term on this list, and the only one that pays for itself twice.
   • check engine light - 169,000/mo - your customer, hiding. Block it, and read the exception first.
   • how to jump start a car - 66,000/mo - standing in a lot with the hood up. Block it.
   • p0420 - 47,000/mo - reading the code instead of paying you to clear it. Block it.
   • mechanic salary - 7,300/mo - somebody deciding whether to retrain. Block it.
   • how to change brake pads - 3,800/mo - already under the car. Block it.
   • obd2 codes - 1,800/mo - looking up the list. Block it.
2. 2

   Open the page the click lands on, on a phone, in a parking lot

   One page catches every ad, and that page is what turns $1.80 into a car on the lift or into nothing at all. Your homepage is the wrong page for almost every search. It answers “who are you”. A driver with grinding brakes is asking “can you take it Thursday, and roughly what am I in for”. Two real exceptions: somebody typing your shop's name, and a one-service shop whose homepage is the whole offer. If the honest reason is that the brake page doesn't exist yet, that's not an exception. That's a page you haven't built.

   So pull the brake page up on a phone and read it as somebody sitting in a parking lot who's been burned by a shop before. Does the first heading say what you fix and what town you're in? Can they tap the number without scrolling? Is there a price anchor anywhere - a diagnostic fee, brakes starting at, or a plain sentence saying you call with a number before anybody touches the car? That last one carries more weight in this trade than almost any other, because the thing the driver is afraid of isn't the repair. It's the surprise.

   Same visit, make a second list: the work you send elsewhere. Body work, glass, tint, transmissions, diesel, RVs, motorcycles. That list turns into cash in step five, where it becomes your negative keywords - jobs a shop refers out are the biggest hole in a small account, every time. And if the page you need isn't there, if you want European work and nothing on the site says so, build it first. Sending that traffic to the homepage buys you nothing except proof the page is missing.

   • The first heading names the work and the town: Brake Repair in Boise. Not “Welcome to our shop”.
   • A number you can tap without scrolling, on a phone. The footer doesn't count.
   • Hours in plain words, including whether Saturday is a real service day or just somebody on the counter.
   • A booking form that works when you're closed, because Sunday evening is when Monday gets planned.
   • A price anchor: diagnostic fee, brakes from, or “we call with a number before we start”.
   • Proof carrying a number: years open, how many ASE certifications, the warranty length, two reviews that name the car.
   • One thing to do. Call, or book. Not four.
3. 3

   Shape it first, then do the money math

   A campaign is one budget with one goal behind it, and nothing else justifies opening another. Not per keyword. Not per match type, not per zip code on your service map. All three do the same damage: they cut your conversions into heaps too small for Google to learn from. An ad group is a single theme - keywords near enough in meaning that one ad answers all of them without lying to anybody. “Auto repair near me”, “auto repair shop near me” and “mechanic near me” belong together. Brakes belong somewhere else, because they want their own headline and their own page. Don't build single-keyword ad groups either. Exact match has been matching close variants for years now, so the isolation those groups promised doesn't exist, and four groups sitting on three conversions apiece teach the bidding nothing while one group with twelve teaches it plenty.

   Now the arithmetic, and this is where auto repair parts company with every other trade on this site. $40 a day at $1.80 a click is about 22 clicks a day, roughly 660 in a month. The floor a new campaign has to clear is around ten clicks a day - below that, thirty days of data tells you nothing you can act on - and you clear it twice over on a budget a plumber would starve on. Say one call in twenty of those clicks: 33 calls in a month, past the thirty conversions the bidding wants before it stops guessing. One in ten and it's 66. Either way you've got a readable month, which is why the honest advice here is build it properly rather than tiptoe. And on the ceiling: Google can run up to twice the daily budget on one day, never more than 30.4 times it across the month, so $40 a day stops at $1,216 and not a cent past that.

   What kills a shop's first month is spreading that same $40 across brakes, AC, tires and oil changes to see which one works. Four campaigns at five clicks a day each, and thirty days later you've got four verdicts you can't trust. One fed campaign learns. Four starving ones don't.

   Start everything on phrase match. Later, once a particular search has actually rung the phone, put it back in as exact with an ad and a page of its own. Keep broad match off until you have conversion-based bidding, tracking that counts real repair orders, and a maintained negative list. Under roughly thirty conversions a month it'll spend the month proving it can't tell a driver from a reader. And pick a bid strategy the account can feed. Thirty-three calls a month sits right on the line Target CPA needs before it stops guessing, and it takes six weeks to get there, so start on Maximize Clicks with a CPC ceiling you set yourself - $2.50 covers everything on this page except the European work - or on manual CPC if you want to feel the levers under your hand. Move to conversion bidding when the account has genuinely produced thirty booked calls in thirty days, not when it looks like it might.

   The money math, done before a single click

   • $40/day at $1.80 a click (“auto repair near me”) = about 22 clicks a day, ~660 a month.
   • The floor for a readable month is ~10 clicks a day. $40 clears it twice over. The same $40 on emergency plumbing at $10.00 a click buys 4.
   • At one call per twenty clicks: ~33 calls in a month, past the ~30 conversions the bidding needs. At one in ten: ~66.
   • Ceiling check: Google can spend up to 2x the daily budget on one day, never more than 30.4x in a month. $40/day = $1,216 max.
   • Same $40 split four ways to “test” = 5 clicks a day each, and no verdict on any of them after 30 days.
   • Add the AC campaign when the first hot week lands, on its own $25/day, and stop it at the end of August. It's a season, not a line item.

4. None of this is difficult. It is a long run of small decisions, and each one quietly costs money when it goes the wrong way. The agent makes them for you and shows you every single one before it touches the account.

   Start free - no credit cardFree for 7 days. No card, no call, cancel whenever.
5. 4

   Twelve headlines, and then read the worst three

   Thirty characters a headline, ninety a description, and Google won't save the ad until you've given it three of the first and two of the second. Give it twelve and three. Count them by hand with the spaces in, because the box will happily take a headline that a phone then chops in half and never say a word about it. Twelve claims, not one claim written twelve ways - a near-twin burns a combination Google could have spent learning something. And whatever the driver typed comes back bolded wherever it lands, so spell it the way a driver spells it, not the way you write a repair order.

   The next bit is the one almost nobody bothers with. Google puts any three of your headlines side by side, in an order nobody shows you, and bills you full price for whichever trio it built. So before you hit save, pick out the three that say least about cars, read them as though they were the finished ad, and ask whether somebody who's never heard of you could tell what's being sold. “Honest, Reliable, Local” and “We Treat You Right” and “Family Owned Since 2011” is a finished ad for a dentist, a roofer or a dog groomer. Nothing in it mentions a car.

   Give each headline one job, and don't let two of them do the same one, even when both are about brakes. “Auto Repair Shop In Boise”, “Brake Pads From $189” and “Car Repair Quote First” all say car and none of them repeat, because one's the search term, one's a number and one answers the fear of the surprise bill. What ruins a set is five headlines pushing one trust claim through five different word orders. Vary the lengths deliberately too: when every headline runs 29 or 30 characters, Google frequently squeezes the ad down to two and a third of what you wrote never appears at all.

   Headline drafts, counted by hand, including two that failed

   • Auto Repair Shop In Boise (25) - the search term, spelled the way they type it.
   • Free Check Engine Scan (22) - the offer, stated once.
   • Brake Pads From $189 (20) - a real number, which filters harder than a bid does.
   • Car Repair Quote First (22) - the surprise-bill fear, answered.
   • 9,000 Repairs Since 2011 (24) - proof, with a number in it.
   • Boise Car Repair (16) - the short one, and it still names the work.
   • WEAK: Honest, Reliable, Local (23) - could be a roofer, a dentist, a dog groomer.
   • FIXED: ASE Certified Mechanics (23)
   • WEAK: We Treat You Right (18) - treat what? Nothing in it says car.
   • FIXED: Same Day Brake Service (22)
6. 5

   Block list first, budget second

   Negatives aren't upkeep on this page, they're part of the build. An auto repair campaign that serves one impression without them is buying code lookups, Haynes manuals and directions to the O'Reilly on the corner inside the hour, because that's what those words mean to everybody else. Campaign level, not ad group. And start with the work you refer out, because that's the biggest leak in a small shop's account and it's the only part of the list nobody else can write for you.

   Fault codes want their own treatment, and the detail matters. Block each code you actually see in the report - p0420, p0171, p0300, p0455 - as a phrase, never as a stem like “p0”, which catches nothing useful. Don't block them as exact either: negative exact stops that precise query and nothing else, so [p0420] kills the bare code and sails straight past “what does p0420 mean”, which is how most of them type it. Phrase catches both. And block the codes you see, not every code in the book - step 7 has the row that explains why.

   Then the thing that convinces shop owners negatives don't work: a negative doesn't stretch. No plurals, no synonyms, no close variants, ever. Block “mechanic salary” and “mechanic salaries” runs all month. Block “part” and you never touch “parts”. Put both in, then read the list back hunting for the word you only typed once. Use the narrowest form that stops the bleeding, too. A negative on broad kills any search holding all your words in any order, so “free” on broad takes “free estimate” and “free brake check” out with it, and both of those are drivers about to book.

   The day-one list for an independent mechanical shop

   • how to, how do i, diy, do it yourself, myself, tutorial, video, youtube
   • repair manual, haynes, chilton, wiring diagram, torque specs, socket size
   • p0420, p0171, p0300, p0455, obd2, obd scanner, code reader, elm327
   • check engine light mean, reset check engine light, clear check engine light
   • mechanic salary, mechanic salaries, mechanic jobs, now hiring, ase practice test, mechanic school
   • auto parts, parts store, autozone, o reilly, napa auto parts, rockauto, used parts, junkyard
   • cars for sale, sell my car, junk my car, kbb value, dealership near me
   • Everything you refer out: body shop, paint job, windshield replacement, tint, towing, rv repair, motorcycle repair
   • Careful: never bare “cost”, never bare “jobs”, never bare “used”, never bare “check engine light”, never “extended warranty”. Every one has a paying twin.

7. Everything past this point is the work that never finishes: reading search terms, adding negatives, checking what the budget did overnight. That is exactly the part the agent takes off your hands.

   Start free - no credit cardFree for 7 days. No card, no call, cancel whenever.
8. 6

   Count cars, not clicks

   Most shop accounts measure clicks, and clicks tell you nothing about whether a car came up the drive. Your work arrives on the phone, so start there. Google builds the call conversion itself, drops a forwarding number into the ad, and counts anything past a threshold you pick. It suggests 60 seconds. That's Google's guess about somebody else's business. A service call is “what's it doing, when can you get it in, we'll call you with a number”, and that's finished inside a minute more often than you'd think, so 45 catches the real ones without letting the wrong numbers in. Ten minutes of work and the phone starts naming the keyword that paid for it.

   Work arriving through a booking form needs a conversion firing on whatever page people see after they hit send, which is two snippets of Google's code pasted into the site. Check two things before you go near a web developer. Auto-tagging first: with it off, no booked repair can be traced back to the click that bought it, and campaigns that are working read as dead and get paused. Then go hunting for conversions Google already built out of your Analytics and quietly left disabled. Flipping those on takes a second, and they're sitting dormant in more accounts than you'd guess.

   Two numbers left, and only you have them. First, what a lead is worth: your average repair order times the share of callers who show up with the car. Without it, Google files a $49 oil change and a $1,900 timing belt as the same event and will cheerfully buy you a month of oil changes. Second, when somebody can answer. Ads run around the clock out of the box. If the report shows 19 clicks between 8pm and 7am and all of them hit voicemail, you paid for those drivers to call the shop down the road. Either the evening ads carry a booking form promising a callback at 7:30, or the evening comes off the schedule.

   • Auto-tagging on. It's a toggle, and with it off nothing ties a booked repair back to the click that bought it.
   • Check for conversions imported from Analytics and left disabled. Switch them on.
   • Build a call conversion with a length that matches how a service call really goes. 45 seconds, not the 60 Google suggests.
   • Value the lead honestly: average repair order times the share of callers who show up. Your numbers, not a benchmark you read somewhere.
   • Schedule around when cars get booked, not when the counter is staffed - and that means Sunday evening stays on. Sunday night is when the driver who limped home all week lines up a Monday drop-off, and a booking form takes that appointment while you're at home.
   • Tracking starts the day you turn it on and last month stays blank forever. You'll know it worked when a booked car shows up in the report, not when the setup is finished.
9. 7

   The every-other-week hour, and what it's actually for

   Days one and two, two questions and no others. Is money moving, and did the ad clear review? A flat spend counter on day two isn't patience. Something's broken - the ad got disapproved, a target is strangling it, or the radius somehow leaves out the block your shop sits on. Go find which. Nothing on the screen looks different between an account that can't serve and one that's warming up, so the keywords take the blame for both.

   Day seven, the search terms report, one job only: add negatives. Leave bids, budgets and targets where they are, don't harvest yet, don't conclude anything. Week one is noise, and a bid strategy still in its learning period takes fiddling personally. The real read is day fourteen to thirty. Cost per call, ad group by ad group, then back into the search terms - this time for waste and for gold. A search that produced a call gets promoted to exact match with an ad of its own. That promotion never stops being the job.

   Month two brings one decision per campaign, and there are only three: feed it, fix it, or kill it. Tracking working, a full month gone, three times your target cost per booked car and nothing on the board? That campaign isn't learning. Say it out loud and act. And be clear about what the expensive part is here. It isn't the setup. It's the two weeks you skip because a transmission job ran long and the parts truck came Thursday.

   Three weeks of search terms, row by row

   • “auto repair near me open now” - 14 clicks, 4 calls -> add as exact match.
   • “brake job near me” - 9 clicks, 3 calls -> add as exact match. This is why you never block the bare word “jobs”.
   • “how much to fix an alternator” - 7 clicks, 2 calls -> add as exact match. A buyer with “how” in the query, which is why you only blocked “how to”.
   • “p0128 thermostat replacement cost boise” - 6 clicks, 1 call -> leave it running. It carries a fault code and it's still a driver pricing a repair in your town. This is the row that argues against blocking every code you ever meet.
   • “check engine light reset” - 11 clicks, 0 calls -> negative phrase “light reset”. Leave “check engine light” itself alone.
   • “car mechanic simulator 2021” - 6 clicks, 0 calls -> negative phrase “simulator”. It's a video game.
   • “mechanic jobs near me” - 4 clicks, 0 calls -> negative phrases “mechanic jobs” and “mechanic job”. Plurals never carry over.
   • “transmission rebuild near me” - 5 clicks, 0 calls -> negative “transmission rebuild”, because you send those out.
   • Not one of those was stopped by a list already blocking “how to”, “salary”, “auto parts” and four fault codes as phrases. That's what the every-other-week read is for. A block list is never done, and the codes on it will cost you a booking now and again, which is exactly why you read the rows instead of trusting the list.
   • That's also only what Google's willing to show you. Anything below its privacy threshold never appears, so these rows will always come up short of your spend. Don't tell yourself the list is complete.

One finished ad, for the auto repair near me group

One ad group - general repair in Boise - for a shop that doesn't exist. The card draws the first three headlines; the other nine sit underneath as the pool Google picks from. That's the part people get wrong on their first ad: you never see the thing you wrote, you see one of the hundreds of trios it can build out of your twelve. One description here, though Google wants two before it'll save anything and will take four.

• Three get drawn. Twelve get written.. The card shows the first three headlines you wrote. The other nine sit in the pool below it, which is where Google actually goes shopping - it assembles a different trio for different searches and you never choose which. So every headline has to survive being read next to two strangers.
• Bold goes wherever the match lands.. Whatever the driver typed comes back in bold, in whichever slot that headline lands in, which is why “Auto Repair Shop In Boise” is written in exactly the words a driver uses. Google also asks for three headlines carrying no keyword at all, so the rest name the work some other way - brakes, AC, oil change, the certification.
• Sample the weakest trio, not a random one, including the 16-character headline.. Any three you happen to pick will pass. The three that decide the ad are the three carrying the least information - here that's “Independent, Not A Dealer”, “9,000 Repairs Since 2011” and “Car Repair Quote First”. Read them cold, as somebody who's never heard of you: a stranger still works out this is a car shop, that it isn't the dealership, and that they get a number before anybody touches anything. That's the test. If the weakest trio comes out as “Honest, Reliable, Local | We Treat You Right | Call Us Today”, the set isn't done, and the nine good ones don't rescue it.
• Twelve slots, twelve separate arguments.. The search term, the near-me variant, the free scan, the certification, the quote-first promise, a real price on brake pads, 9,000 repairs, oil and inspection, AC, same-day brakes, the not-a-dealer angle, and the short town version. What you won't find is one trust claim wearing four different hats. “Trusted”, “honest”, “family owned” and “you're in good hands” are a single claim in four sets of coveralls, and it gets a single slot.
• “Independent, Not A Dealer” does work no generic headline can.. It answers the comparison the driver is already running in their head - you or the dealership, and dealer service is the thing they're trying to escape. Every shop has one of these. Yours is usually the price against the dealer, or the fact that the same tech touches the car every time.
• A price in the copy filters harder than a bid does.. “Brake Pads From $189” is the cheapest negative keyword in the account. A driver who was never going to spend that reads the number, keeps scrolling, and costs you nothing. Put a real figure in, one you'll honor at the counter. If you quote instead of publishing prices, say that instead - which is the job “Car Repair Quote First” is doing two slots down.
• “Free Check Engine Scan” is a promise, so run it only if it's true.. It's also the one headline tying this ad to the biggest search category in your market. If you offer it and you have the bays, it earns its slot and it opens the door to the exception in the playbook. Charge for diagnostics instead? Take it out and put the fee in its place. A driver who was never going to pay $59 to have the code read reads the number and keeps scrolling, which is exactly what you want.
• Write the second description, because one of them is running blind.. Two is the minimum Google accepts, four is the most it takes, and it picks which one a given searcher gets. The one above carries the range of work, the certification and the quote. Write a second one around turnaround - most brake jobs back the same day - and a third only if it makes a claim the other two don't. Reword the first and all you've done is give Google a weaker twin to spend your money on.
• 16 to 25 characters, and the spread is deliberate.. Three headlines is the ceiling, and only when all three fit the space Google has. Run everything at 29 or 30 characters and the ad gets trimmed to two, so a third of what you wrote never sees daylight. Keep two or three genuinely short ones in the set, and make them still say car.
• Sitelinks are free frontage.. Skip them and your ad sits visibly smaller on the page than the shop advertising underneath you, for no saving at all - extensions are free. These four answer the second question a driver has, what a diagnostic runs and whether you do AC, without anybody rebuilding the landing page. Add a call extension in the same sitting, point it at the call conversion from step six, and schedule it to the hours somebody picks up.

The account a two-tech shop can actually run

Two campaigns doing work and one small one guarding your name. Four ad groups all in. Each is tight enough that one ad answers everything inside it and one page catches the click, and no tighter than that - at this spend, four groups already means each one sees a handful of calls a month, and a handful proves nothing either way. Everything starts on phrase match. Exact match is a promotion, and only a search that has rung your phone earns it.

Why it's shaped like this. General repair and brakes share a campaign because they share a budget and a goal, but they sit in two ad groups because a brake ad and a general repair ad aren't the same ad. AC is its own campaign purely so it can have its own on and off dates - a season needs a switch, and you can't switch off half of a shared budget. Brand sits alone because brand clicks are the cheapest and best-converting thing in any account, and folded into general repair they make a losing campaign read like a healthy one. Check somebody is actually bidding on your name first: if you own the free results and no one's showing above you, that $5 a day buys clicks that were already yours. Three things are missing on purpose. European specialty, because at about $2.50 a click it wants its own budget and its own page, and it only earns them if you really have the diagnostic software - then it becomes the best ad group in the account. Glass and body work, because a mechanical shop bidding those buys jobs it has to hand off, and 23,000 searches a month for glass repair is somebody else's money. And the coupon terms, because those turn on an offer you can honor and bays you need to fill, which is a decision about February, not a decision about keywords.

Google Ads management for auto repair shops, shown as cards you approve one at a time.

What you actually pay an agency for is the hands-on part. Deciding which of the check engine searches to pay for and which to block, building a campaign per job type, digging through search terms for money leaks, matching the ad hours to when cars get booked, writing ads that fill bays rather than collect clicks. AdPilot's agent does that work and shows its reasoning on every card.

Examples of what the cards look like - not results from your account

The car broke on Saturday. The report comes out on the 5th.

You keep the account, and nothing changes in it while the bays are full.

Most shop owners never learn how their own Google Ads account works. Nobody explains it, and an agency has little reason to start. AdPilot works the other way around. Every card tells you:

• What it changes - for example, pause a keyword that spent $187 on fault-code lookups and booked nothing.
• Why it helps - the argument behind it, not only the instruction.
• What it is based on - figures taken from your own account.

Approve one and the snapshot is taken before anything moves. Afterward a before-and-after view shows what changed - cost per booked car, clicks, spend - in language nobody needs a marketing course to follow. Two months of that and you can tell which levers fill bays and which ones only filled a report.

AdPilot is new - and we’d rather earn your trust than fake it.

We could plaster this page with five-star quotes and a “trusted by 10,000 businesses” banner. We won’t - because AdPilot is new, and inventing proof would be lying to you. Instead, here’s what actually protects you:

• You risk nothing to try it. 7 days free, no credit card. If it isn’t useful, you walk away having spent nothing.
• Your account is never at risk. Nothing is ever applied without your one-tap approval, a snapshot is saved before every edit, every change has one-click revert, and a full audit log records it. The worst case is a suggestion you don’t like - so you just don’t approve it.
• Guardrails you can’t trip by accident. AdPilot never deletes campaigns and never touches shared budgets. Some doors simply don’t exist.
• Built by an operator, not a VC-funded black box. Software made to give small businesses the Google Ads work an agency charges a retainer for - without agency prices or the loss of control.

Judge us on the work, the math, and the safeguards - on your own numbers, not our adjectives.