The Ads Layer

Google Ads for
Specialty
Automotive Shops.

We don't touch your ad spend, and we don't make a dollar more when it goes up. If we're doing this right, you spend less on ads every year — not more.

01
The Faucet

Where Paid Search Actually Fits for a Specialty Shop.

Almost every agency in this industry is paid a percentage of what you spend. Which means the person advising you on your ad budget gets a raise every time they talk you into a bigger one. That is the whole problem with advertising for small shops, and it's worth understanding before you read another word about keywords.

Paid ads are one of three channels, and they're the only one that stops the moment you stop paying.

Your Google Business Profile wins the map pack — the fast-turn, high-frequency work where somebody picks the closest good-looking pin and calls it. Your organic content wins the research-phase buyer — the guy pricing a full build who reads for three weeks before he calls anybody. Both of those keep producing after they're built.

Ads are the faucet. You turn them on where you have no map presence and no organic ranking yet, because those two take months to mature and you have bays to fill this quarter.

Ads buy attention immediately, at a price, for exactly as long as you keep paying for it.

That's not a knock. Buying time is a legitimate thing to buy. The failure is treating the faucet like the plumbing — running ads for five years with no organic plan underneath, and calling it a marketing strategy. That's renting attention you should own by now.

So the entire question with paid search isn't should you run ads. It's what are the ads covering for, and what's the plan to stop needing them there. The faucet-and-plumbing framing runs through everything Osiris builds — see how it plays out for lead generation more broadly.

02
The Structural Point

We Never Touch Your Ad Spend.

This is the structural point, and everything else on this page follows from it.

Your ad budget goes from your card to Google. Directly. It never passes through us. We don't mark it up, we don't hold it, we don't invoice you for it, and we don't take a cut of it.

Our fee doesn't move when your spend moves. Doubling your budget doesn't pay us a dollar more. Cutting it in half doesn't cost us a dollar. Turning ads off entirely in a city where you've started ranking organically — which is the actual goal — costs us nothing.

Compare that to how this normally works. The standard agency arrangement is a percentage of ad spend, usually somewhere between ten and twenty percent. Under that model:

Nobody in that arrangement is a villain. The incentive is just pointed the wrong direction, and over enough months incentives win. We removed it by not being paid that way.

It also means you own the account. It's in your name, on your billing, with your data and your history in it. If this partnership ends, your ad account and everything it has learned stays with you. We're not holding it hostage — that's not how any of this works here.

03
Say the Quiet Part

The Goal Is to Spend Less on Ads Every Year.

A well-run paid search program for a specialty shop should shrink. Not because ads stopped working — because the ground they were covering got taken by something permanent. Here's the sequence, service by service and city by city.

Month One

No organic ranking for "long-arm suspension install" in your market and no map presence for it either. Ads on, aggressively. Every one of those clicks costs money and every one is worth it, because the alternative is that nobody knows you do it.

Month Eight

The service page has matured, the content around it is indexed, and you're on page one for that term. The ad is now competing with your own organic listing for the same searcher — and you're paying for a click you would have gotten free.

Month Nine

Spend on that keyword comes down. Not off, necessarily — down. What comes off it goes to a service or a suburb where you're still invisible.

Year Two

The faucet is running on a fraction of what it was, pointed at your weakest ground, and the strong ground is producing for free.

Any agency whose plan for year three involves you spending more on ads than you did in year one has, by definition, failed at the durable part of the job. Ads should be the thing you need least by the time the infrastructure is mature.

04
Generic Advice Fails Here

Why a Specialty Shop Is a Different Advertising Problem.

Most advice about Google Ads is written for businesses that are nothing like yours, and following it will cost you money.

01

Your Ticket Is Enormous and It Varies Wildly

A tire mount and a full frame-off restoration both come in through the same phone. Advertising software that treats every lead as equally valuable will happily buy you a hundred of the cheap ones and report a fantastic cost per lead while your bays sit empty of real work.

02

The Decision Takes Weeks, Not Minutes

Somebody pricing a serious build clicks an ad in April, calls in May, comes in for a consult in June, and pays a deposit in July. Every conventional advertising dashboard has stopped watching by then. Judged on a 30-day window, that campaign looks like a failure right up until the deposit clears.

03

Your Customers Call. They Don't Fill Out Forms

Jobs at four and five figures are conversations. Nobody commits to a build through a web form. Which means if your advertising measurement is built around form fills — and almost all of it is, by default — you are measuring the smallest and least valuable slice of what your ads actually produced.

04

Your Volume Is Low by Design

A shop doing high-ticket work might get a few dozen genuinely relevant clicks a month per service. That's not enough data for the automated bidding everyone talks about to learn anything on its own. It has to be structured deliberately and fed information manually until there's enough history to hand over.

05

Your Customers Will Drive

Somebody looking for a shop that actually knows their platform will drive two hours past four closer shops. A tight radius around your zip code makes sense for an oil change and throws away money for a specialty build.

Every one of those changes how the account gets built. None of them show up in generic PPC advice.

05
What We Don't Run

Search Only. No Display, No Performance Max, No "Brand Awareness."

When you run Google Ads, Google will strongly encourage you to let it spend your money across its entire network — display banners on random websites, YouTube pre-roll, Discover feeds, Gmail promotions, and its Performance Max product, which spends across all of them with almost no visibility into where it went.

We run search only. Search Partners off. Display off. No Performance Max. No video.

The reason is simple: search advertising catches somebody who already wants what you sell and is looking for it right now. Everything else interrupts somebody who wasn't thinking about you. For a high-ticket considered purchase from a local shop, interruption converts badly and reports beautifully — the worst possible combination, because the report will tell you it's working.

Display in particular is where small ad budgets go to die. It generates enormous impression counts and click numbers that look like progress, at a cost per actual job that nobody ever computes because the dashboard doesn't show it that way.

We also don't run display remarketing. Following somebody around the internet with a banner for six weeks doesn't sell a $12,000 build; it just annoys them. Remarketing back into search — where somebody who already visited your site sees you again when they search a second time — is genuinely useful for long decision cycles, and it's the version we use.

06
The Wrong Metric

Cost Per Click Is the Wrong Number.

Almost every ads conversation a shop owner has ever had was about the wrong metric.

Cost per click tells you what you paid for attention. It says nothing about whether the attention was worth having. A $3 click from somebody looking for a DIY parts list is worse than a $14 click from somebody pricing an install, every single time, and the metric can't tell them apart.

Cost per lead is one step better and still wrong for your business, because it treats a tire quote and a full build inquiry as the same event.

The only number that matters is what a closed job cost you to get.

That means the measurement has to survive the entire gap between the click and the money — which for you is often six to ten weeks, usually runs through a phone call, and frequently ends in a payment that has nothing digital attached to it at all.

Which is a genuinely hard measurement problem. Most shops running ads have simply never solved it, so they're operating on a feeling about whether it's working. The next two sections are how it actually gets solved.

07
The Attribution Problem

Attribution Starts With a Phone Number That Only Exists in Ads.

Here is a finding from our own client accounts that changed how we build every campaign.

We pulled and analyzed several thousand inbound calls across two client shops. Every single native attribution field was empty. Source, keyword, referrer, landing page — blank, on all of them. Not most. All of them.

If your business runs on the phone, and the platform can't tell you where a single phone call came from, then your advertising reporting is describing a slice of your business that doesn't include most of your revenue.

The fix is unglamorous and it's the only thing that works: a dedicated tracking phone number that appears in your paid ads and absolutely nowhere else. Not on the website. Not on the Business Profile. Not on the truck, the invoice, the shop door, or a business card. One number, one place.

Then the logic is airtight. That number rang, so that call came from an ad. There is no interpretation, no modeling, no "assisted conversion" hand-waving. It rang or it didn't.

Calls route through to your normal line, get recorded and logged in the CRM against the right customer, and nothing changes about how your front counter works. The only difference is that at the end of the month you can say a specific number of calls came from paid search, and you can follow each one forward to whether it became a job.

And we count conservatively on purpose. A call from that number counts. A booking completed through the flow the ad click landed on counts. Somebody who saw an ad in March, thought about it, and searched your shop by name in May does not count as an ad conversion — that's a branded search, and calling it paid performance is how agencies manufacture results out of demand you already had.

08
Value-Based Bidding

Teaching Google What a Job Is Actually Worth.

This is the piece almost nobody does for a shop, and it's where paid search stops being a coin flip.

Google's bidding gets better when you tell it what happened. Left alone, it optimizes toward whatever event you told it to count — usually "a form was submitted" or "a call connected." So it goes and finds you more of those. Cheap ones. Lots of them. And it will report a beautiful cost per conversion while quietly buying the least valuable inquiries in your market, because those are the ones it can get cheapest.

It has no idea that one of those calls turned into a $9,400 build and forty of them turned into nothing. Not unless you tell it.

So we tell it. After the job is done and the money is actually collected, the real revenue figure gets sent back to Google and matched to the original click that started it. Not the estimate. Not the quote. The collected amount.

Do that for a few months and the bidding has a real history to learn from — which searches, at which times, from which areas, in which phrasing, turn into money instead of tire-kickers. It starts spending your budget on the expensive clicks that pay and starving the cheap ones that don't.

Two things about this that matter to you:

09
Clicks You Already Own

We Don't Buy Clicks You Already Have.

Two places where shops routinely pay for traffic that was already theirs.

Advertising on top of your own organic ranking. If you're sitting at position one or two organically for a term, running an ad above it means buying a share of clicks that were coming to you free. There are narrow cases where it's defensible — a competitor bidding hard on that exact term, a service where the map pack is pushing organic below the fold. It should be a decision somebody made on purpose, not the default.

Bidding on your own shop name. Somebody typing your shop's name into Google is already yours. In most markets, paying for that click is buying a customer you had. The exception is real and worth checking: if a competitor is bidding on your name, their ad can sit above your own listing, and defending the term is cheap because Google charges very little for an ad that's obviously relevant. So the honest answer is check first, then decide — not "always" and not "never."

Both of these get reviewed as part of managing the account, and both are places where the standard advice conveniently increases spend.

10
The Bucket With Holes

Ads Without a CRM Behind Them Leak.

The faucet analogy finishes itself here. Paid search turns the water on. The CRM is the bucket. If the bucket has holes in it, turning the water up harder just wastes more water faster.

A shop that pays for a click and then misses the call, or answers it and never follows up on the quote, is paying full retail for attention and then dropping it. And ads make that failure more expensive than it was before, because now every lost lead had a receipt attached to it.

There's also the measurement dependency: none of §07 or §08 works without a CRM. Attribution needs somewhere to record the source. Value-based bidding needs to know a job closed and for how much. The reporting that tells you what a job actually cost to acquire is assembled from CRM data, not from Google's dashboard.

Which is why ads aren't sold on their own here. They run on top of a site that converts, a CRM that catches and follows up, and reporting that connects a click to a paid invoice.

11
No Flattering Version

What the First Ninety Days Actually Look Like.

No agency wants to write this section because it isn't flattering. Here it is anyway.

Weeks One Through Three — the Learning Phase

The numbers will look bad. New campaigns start with no history, so the platform is genuinely guessing. Cost per click runs high, some spend goes to searches that turn out to be wrong, and the early conversion count is thin. This is normal and it is not fixable by fiddling.

The temptation — and the mistake — is to restructure everything in week two. Every restructure resets the learning and starts the clock over. Shops that do this monthly stay permanently in the expensive part of the curve and conclude that ads don't work. Ads worked fine; they just never got past the introduction.

What we're watching during that window is mechanical, not performance: are conversions actually firing, is the tracking number routing correctly, are budgets pacing, is anything obviously wrong. Not "is the cost per lead good yet." It isn't yet, and it isn't supposed to be.

Weeks Four Through Eight

The search-term data becomes useful. Now we can see the actual phrases people typed, which is different from the keywords we bought. Some of it is exactly right. Some of it is DIY questions, job seekers, wrong vehicle categories, people three states away. That gets filtered out, deliberately and one at a time — carefully, because plenty of terms look like the wrong intent and aren't. A word that reads DIY-ish is sometimes a real service term you'd be crazy to block.

Around the Third Month

There's enough history to change how bidding works and, once the job-value feedback is running, to start weighting toward the work that actually pays.

Ongoing

Through all of it you get reported on calls, booked appointments, closed jobs, and revenue. Not impressions. Not click-through rate. Not "brand lift." If a campaign isn't producing jobs, that should be legible to you in about eight seconds, and you should be able to tell us to kill it.

12
Said Plainly

What We Don't Do.

Shorter list, said plainly, because most of it is what you've been sold before.

13
Straight Answers

Ads Questions From Shop Owners.

You do, directly, on your own card. Ad spend never passes through Osiris and never appears on an Osiris invoice.

No. Our fee doesn't move with your ad budget in either direction. That's deliberate — it's the conflict of interest that ruins most agency ad relationships, and we removed it structurally rather than promising to resist it.

No. Ads management is part of what an anchor client gets, but running them is a decision you make. A shop that already ranks well and is booked out three weeks may be better served with the faucet nearly closed.

We run search. Search catches somebody who is already looking for what you do — social interrupts somebody who wasn't. For a high-ticket considered purchase from a local shop, that's a much harder and more expensive path. Social is part of the infrastructure as a brand recognition layer, not as a lead engine, and we'd rather be honest about the difference than sell you a channel because it's expected.

Usually no, sometimes yes. If nobody's competing for your name, you're paying for customers you already had. If a competitor is bidding on it, defending it is cheap and worth it. We check rather than assume.

The first two to three weeks are a learning phase and the numbers will look rough. Useful search-term data shows up around week four to six. Real optimization on job value needs a few months of history. Anyone promising results in week one is either guessing or spending recklessly to manufacture early numbers.

Spend on that term comes down and moves to ground you haven't taken yet. That's the plan, not an accident — the faucet closes as the plumbing gets built.

Yes. We audit what's there first and report what we find before touching anything. Existing history has value and we don't throw it away by reflex.

A dedicated tracking number that appears in your ads and nowhere else. If it rang, it was an ad. There's no modeling or interpretation involved.

The rest of the system
Territory Check

One Shop Per Niche,
Per Market.

Ads aren't sold on their own here, because on their own they're a faucet pointed into a bucket with holes. They run on top of a site that converts, a CRM that catches every call, and reporting that follows a click all the way to a paid invoice.

Check whether your slot is open.

No pitch. No deck. A direct conversation — about your market, your ad spend, and whether the position is still open.