How to read a Google Ads report without the guesswork.

Most Google Ads reports get one of two reactions from a shop owner. Either it goes straight into a folder nobody opens, or it gets skimmed for one number, usually the one that happened to go up, and treated as proof everything is fine. Both reactions miss the point of the report entirely. A Google Ads report is Google telling you, in Google's own preferred language, how your money moved last month. Reading it correctly means knowing which of those numbers actually describes your business, and which ones only describe how hard Google tried.

Why your report reads like a different language.

Open the standard Google Ads dashboard and the first screen throws a dozen column headers at you at once: impressions, clicks, CTR, avg. CPC, cost, conversions, cost/conv, conv. rate, impression share, Quality Score, and a handful of others depending on what your agency or account manager chose to surface. None of those labels tell you, on their own, whether the account made or lost money last month. They tell you how the auction behaved.

That isn't an accident. Google Ads reporting is built around auction mechanics, because that's the system Google actually runs. Impressions tell you how many times your ad was eligible to show. Clicks tell you how many people chose to click it. CTR tells you what share of eligible impressions turned into a click. Every one of those numbers lives entirely upstream of the only question a shop owner actually cares about: did any of this turn into a car in a bay or a booked appointment on the calendar.

The report isn't lying to you, exactly. It's answering questions you didn't ask, with more confidence than the questions you did ask ever get answered. Fixing that starts with sorting the columns into two piles: the ones that describe the auction, and the one or two that describe your business.

The metrics that flatter the dashboard, not your bank account.

Four numbers show up near the top of almost every report, and all four are worth a passing glance and nothing more if the goal is deciding what to do with next month's budget.

Impressions. A climbing impressions count usually just means your budget or your bids went up, or a seasonal search spike lifted volume for everyone in your category. It says nothing about whether a single one of those impressions reached someone who was actually going to call.

Click-through rate. A high CTR means your headline matched the search well enough to get clicked. It says nothing about what happened after the click. An ad promising a service you don't actually offer can post a great CTR and a terrible result, because the people who click are the wrong people the moment they land.

Average CPC in isolation. A falling average cost per click looks like a win, and sometimes is one, but a $1.80 CPC on searches that never call is worse than a $4.50 CPC on searches that book. Cost per click measures what you paid for attention, not what that attention was worth.

Quality Score, treated as a grade to chase. Quality Score genuinely affects what you pay per click, and a Quality Score that's consistently low on your core terms is worth fixing. But it's a means, not an end. Chasing a 9 or 10 on every keyword while ignoring whether those keywords produce booked work is optimizing the wrong side of the equation.

The one number that decides if it's working.

Cost per conversion is the closest thing Google's report gives you to the truth. Cost per booked job is the truth itself.

Buried a few columns past CTR and Quality Score sits cost per conversion, sometimes labeled cost/conv. This is the number worth building a monthly habit around, with one important catch: Google's definition of a conversion isn't automatically the same as your definition of a booked job.

Out of the box, a "conversion" in Google Ads can mean a form submission, a click on a phone number, or a call that merely started and lasted ten seconds before the caller hung up on a wrong number. None of those are a booked appointment. If your account is still counting every click-to-call tap as a conversion with no minimum call length and no link back to whether the call actually turned into work on the schedule, your cost per conversion number is measuring something close to cost per lead, not cost per booked job, and the two can tell very different stories.

The fix is a conversion action built around calls that actually last long enough to be a real conversation, generally 60 seconds or longer, paired with import of real call outcomes wherever possible. Once that's set up correctly, cost per conversion in the Google Ads report starts tracking something close to the real number: what you're paying Google for a lead with a genuine chance of becoming a booked job. Getting all the way to cost per actual booked job, which also accounts for leads that never close, is a deeper tracking setup we cover in how to track your real marketing ROI, worth reading once this report-level fix is in place.

Four more metrics worth a glance, not a grade.

Past cost per conversion, four more lines in the report are worth checking on a regular cadence, not because any one of them should drive a decision alone, but because each one flags a specific kind of problem before it gets expensive.

Impression share

Impression share tells you what percentage of the searches you were eligible to show on actually showed your ad. A shop sitting at 35% impression share, lost mostly to budget, is leaving a predictable amount of volume on the table every single day, and raising budget is usually the straightforward fix. A shop sitting at 35% impression share lost mostly to rank, meaning the ad, landing page, or bid isn't competitive enough to win the auction even with more budget thrown at it, has a different problem that more spend alone won't solve.

The search terms report

This is the single most underused tab in the entire account. It shows the actual words someone typed before your ad showed, not just the keyword you targeted. A campaign targeting "auto repair near me" can quietly be spending real money on searches for "auto repair jobs hiring" or a specific competitor's name, and nothing in the top-level metrics will flag it. Fifteen minutes scanning this report every month, adding the obvious irrelevant terms as negative keywords, is often worth more than any bid adjustment.

Conversion rate

Conversion rate, clicks divided into conversions, is most useful compared against itself over time and against your other campaigns, not against some industry benchmark pulled from a blog post. A campaign converting at 4% that used to convert at 9% just told you something changed, in the ad, the landing page, or the competition, worth investigating immediately. A campaign holding steady at 4% for a year that has always converted at 4% is probably just your account's natural rate for that kind of search.

Device performance

The device breakdown often hides the simplest fix in the whole account. A shop running a desktop-designed landing page can show a respectable overall conversion rate while mobile, which is where the overwhelming majority of "near me" searches happen for a local service business, converts at half the desktop rate because the phone number isn't a one-tap call button or the page takes several seconds to load on a cell signal. Checking device performance separately, instead of as one blended number, is often the fastest fix available in the whole report.

Reading a real week of numbers, line by line.

Here's what this looks like applied to one representative week for a local service campaign, the same kind of week that lands in an inbox and gets either ignored or misread.

Impressions: 9,400. Clicks: 290. CTR: 3.1%. Average CPC: $3.40. Cost: $986. Conversions: 19. Cost per conversion: $52.

Read at face value, that looks solid. A $52 cost per conversion against a typical repair order of $300 to $400 has plenty of room. But the conversion action behind that 19 is counting every call tap with no minimum length. Pulling the call log behind those 19 conversions shows 12 calls that lasted under 20 seconds, mostly hang-ups and wrong numbers, and 7 that ran long enough to be a real conversation. Of those 7, 5 turned into booked appointments.

The honest version of that week isn't $52 per conversion. It's $197 per real booked appointment, $986 divided by 5, which is a very different number to make a budget decision against, and still a profitable one against a $300-plus repair order, just not the number the dashboard handed over first.

The search terms report for that same week shows two queries worth flagging: "auto repair shop reviews" and "auto repair technician jobs," together responsible for $64 of that week's spend and zero of the 5 real appointments. Adding both as negative keywords doesn't fix the account by itself, but it's $64 that week, repeated across 52 weeks, that was never going to become a booked job no matter how long the campaign ran.

Impression share for the week sits at 61%, split roughly 70% lost to rank and 30% lost to budget. That points to a clear next step: raising budget alone would buy more of the same mix of good and irrelevant clicks. Improving ad relevance and the landing page experience first, then raising budget once rank improves, gets more of the real appointments instead of just more volume.

The red flags that mean something's actually broken.

Most months, the report just needs a quick read and a shrug. A handful of patterns are worth stopping and acting on the same day you see them.

• Conversions climbing while cost per conversion also climbs fast. This usually means volume is coming from a weaker, more expensive source, often a broad match expansion or a new placement Google auto-added, not from genuine demand growth.
• Impression share collapsing week over week with no budget change. A competitor raising bids or a Quality Score drop on core terms can push you out of auctions you used to win, and it rarely fixes itself without a direct response.
• Conversion rate holding steady while booked jobs quietly drop. This is exactly the pattern that showed up in the sample week above: the report's own conversion count can look healthy while the underlying call quality degrades, which is why a minimum call-length threshold on the conversion action matters so much.
• A single search term eating a large share of spend with zero conversions. One runaway keyword, often a broad match term nobody deliberately chose, can quietly become the most expensive line in the account.
• Search Partners or Display placements showing up on a campaign built for Search. These placements can pad impressions and clicks cheaply while rarely producing a real local service call, and they're worth checking and excluding if they were never a deliberate choice.

None of these require advanced analytics to catch. They require actually opening the report every month instead of skimming the top line and moving on.

What to ask, agency or not.

Whether a report lands from an in-house hire, an agency, or the Google Ads dashboard directly, the same three questions get past the noise every time.

What's the cost per real booked job, not cost per conversion? If the answer requires pulling call recordings or cross-checking the shop's own schedule, that's a sign the conversion action isn't built correctly yet, not a dead end. If the answer comes back instantly and matches the dashboard's cost per conversion number exactly, that's worth a second look too, since it usually means nobody has checked whether every counted conversion was a real one.

What showed up in the search terms report that wasn't supposed to be there? Every account accumulates some irrelevant traffic over time. An agency or account manager who can't name at least one thing they caught and excluded recently either isn't checking, or the account is cleaner than almost any account actually is.

What changed from last month, and why? A report should come with a sentence of explanation for anything that moved more than 15 to 20%, in either direction. "Conversions were up" without a reason attached is a headline, not an answer.

The five-minute monthly read that's actually worth doing.

None of this requires turning into a PPC analyst. It requires reading five numbers in order, every month, in under five minutes.

Start with cost per conversion, adjusted for a real minimum call length, not the raw dashboard number. Check it against your average repair order or job value with real margin room. Scan the search terms report for anything obviously irrelevant and exclude it. Check impression share and note whether lost share is coming from budget or rank. Pull one or two of this month's actual booked jobs and trace them back to confirm the conversion tracking matches reality.

That five-step read catches almost everything that matters, long before it turns into a quarter of wasted spend nobody noticed. The dashboard will keep handing you a dozen columns every month regardless. Which five you actually read is the only part of the report that was ever going to change how the account performs.

Frequently asked questions.

What's the one metric I should check first on my Google Ads report?

Cost per conversion, but only after confirming the conversion action requires a minimum call length, generally 60 seconds or longer, or an actual form submission rather than a landing page view. Without that fix, cost per conversion is closer to cost per lead than cost per booked job, and the two numbers can tell very different stories about the same account.

Why does my cost per conversion look good but my phone still isn't ringing more?

The conversion action is almost certainly counting activity that isn't a real booked job, most commonly short calls under a minute or clicks on a phone number that never connected. Pull the actual call log behind the conversion count for a week and compare it to the dashboard number. The gap between the two is usually the whole explanation.

What is impression share and should I worry about it?

Impression share is the percentage of searches you were eligible to show on that your ad actually showed for. It's worth checking monthly, not obsessing over weekly, and the useful detail is always the split between share lost to budget versus share lost to rank, since the fix for each one is completely different.

How often should I actually sit down and read my Google Ads report?

Monthly for a full read, with a quick glance at spend pacing weekly so nothing runs away between full reviews. Reading it daily usually just reacts to normal day-to-day noise in the auction, and reading it quarterly or less lets a broken conversion action or a runaway search term burn budget for months before anyone catches it.

Should I fire an agency over one bad month in the report?

Not on one month alone. A single slow week or month can come from seasonality, a platform-wide auction shift, or plain variance, and judging on one data point is how good channels get killed for the wrong reason. A pattern across two or three months with no clear explanation, or an agency that can't answer what changed and why, is the actual signal worth acting on.

Mitchell Wolfert Founder, Provision Digital. Running paid media, SEO, and funnels for local and national brands from Queen Creek, Arizona.

Related reading: how to track your real marketing ROI, how to track which ads drive phone calls, why your Google Ads aren't getting calls, and Google Ads for local service businesses.