How to Analyze Google Ads Campaigns: The Right Order

Thursday, nine in the morning. A marketing manager at a mid-sized electrical supply distributor opens the Google Ads dashboard and sees a 7.2% CTR, cost per click lower than last quarter, and an average Quality Score of 8. Everything green.

That same week, the sales team complains that almost every quote request from the website comes from a homeowner asking the price of two meters of wire.

Both are right. The campaign is healthy on the inside and useless on the outside. That gap is the core problem with analyzing Google Ads: the dashboard measures the health of the auction, not the health of the business. Most analyses start at the wrong metric, which leads managers to optimize a layer that was only a symptom of another one.

What does it actually mean to analyze a Google Ads campaign?

Analyzing a Google Ads campaign means checking, in this order: whether the conversion you measure is the one that matters, which search terms actually spent the money, how much of the auction you lose to budget versus rank, and only then ad copy and landing page. Each layer caps the next one.

The order is not arbitrary. If conversion tracking is broken, every number below it is broken too, including the CPA you used to kill a campaign last week. If half the budget went to irrelevant queries, testing headlines changes nothing. And if you lose 50% of impressions to rank, raising the budget barely moves volume.

Step 1: is the conversion you measure the one that pays the bills?

Before touching anything, open the conversions column and ask what is being counted. In service and B2B accounts it is common to find three or four actions summed into one number: form submissions, WhatsApp button clicks, contact page views and sometimes time on site.

Three of those four are not sales. They are intent. When Smart Bidding optimizes toward that sum, it learns to buy the cheapest click that produces the easiest event, which is almost always a tap from someone who just wanted a price.

The practical test takes ten minutes: take last month's conversion count and ask sales how many real opportunities they logged in the same window. If the gap is wider than 30%, stop the analysis and fix tracking first. It also pays to review how credit is split across channels, which we cover in conversion attribution.

Step 2: the search terms report (and what it hides)

A keyword is what you bought. A search term is what the user typed. With broad match and automated campaign types, the distance between those two has become the main budget leak in small and mid-sized accounts.

Sort the report by cost, not by impressions, and read the top 30 rows. Add up the spend on terms with zero conversions over the last 60 days. That figure, not your average CPC, tells you whether the campaign is calibrated.

One detail changes how you read it: the report is not complete. Google's own help documentation states that terms without enough query activity are omitted from the report in order to maintain data privacy standards (Google Ads Help).

In practice you usually see most of the spend, but never all of it. That is why reactive negative keyword work, term by term, has a ceiling. Past a certain point the gains come from structure: splitting campaigns by intent, tightening match types on the lines that convert, and maintaining negative lists by theme rather than by isolated word.

Step 3: impression share tells you where your ceiling is

Impression share is the percentage of impressions your ads received compared with the total number of impressions they were eligible to receive. Google publishes the formula plainly: impressions divided by total eligible impressions.

The number on its own is close to useless. What decides your next action is the split between share lost to budget and share lost to rank. One is a funding problem. The other is a competitiveness problem.

The scenario in the chart plays out constantly in smaller accounts: the manager asks for a budget increase, gets it, and three weeks later volume has barely moved. It was predictable. The bottleneck sat in rank, which responds to bids, ad relevance and landing page quality, not to funding.

The inverse reading matters just as much. High share lost to budget on a campaign that already converts inside target CPA is the closest thing to idle money in paid media.

Step 4: Quality Score does not do what most people think

This is where the most expensive misunderstanding lives. Google's documentation is explicit that Quality Score is not a factor in the ad auction, and that it exists as a diagnostic tool to show how your ad quality compares with other advertisers (Google Ads Help).

Chasing a 10 as a goal, then, is confusing the thermometer with the fever. It runs on a 1 to 10 scale at the keyword level and rests on three components: expected click-through rate, ad relevance and landing page experience.

The useful application is diagnostic. Filter for your highest-spending keywords with a score under 5, then look at which of the three components reads "below average". If it is landing page experience, no headline test will save you. If it is relevance, the cause is usually an ad group stuffed with mismatched intents.

Step 5: read the number that decides, not the one that comforts

Run the numbers on a typical niche e-commerce account. A monthly budget of $12,000, 2,400 clicks, an average CPC of $5.00. A 4% conversion rate gives 96 orders. With an average order value of $180, that is $17,280 in revenue.

ROAS of 1.44. Looks positive. Except the catalog runs a 45% contribution margin, which means $7,776 of contribution against $12,000 of media. The campaign burns $4,224 a month while looking healthy.

Break-even ROAS here is 1 divided by 0.45, or 2.22. Anything below that is a loss wearing the costume of a result. If that threshold is not part of your routine yet, start with how to calculate ROAS and the logic behind cost per lead for service businesses.

Symptom, likely cause, first move

What you see	Likely cause	First move
High CTR, low conversion	The ad promises what the page does not deliver	Compare the headline with the landing page hero
Conversions rise, revenue does not	Weak events summed into one column	Split primary and secondary conversions
CPC climbing with no bid change	New competition or lost relevance	Check impression share and Quality Score components
Volume flat after a budget increase	Share lost to rank, not to budget	Revisit target bids and page quality
Plenty of clicks, few qualified leads	Broad match without negative lists	Sort search terms by cost and negate by theme

Where this reading breaks down

It breaks in low-volume accounts. Below roughly 30 conversions a month, most weekly movement is noise, and treating every swing as a signal leads to changing things that were fine. Work with 30 or 60 day windows there and accept deciding more slowly.

It also breaks with long sales cycles. If deals take 45 days to close, the September report is judging a July investment. Month-over-month comparisons in those accounts produce inverted conclusions more often than people admit.

And it breaks for pure top-of-funnel campaigns. A new brand reaching an audience that has never heard of it will not return immediate ROAS, and demanding it is the fastest way to kill future acquisition.

Frequently asked questions

Where is the search terms report in Google Ads?

In the left menu under Campaigns, open the keywords section and choose "Search terms". You can also open it from inside a specific ad group. Sort by cost to see where the money went first.

What counts as a good impression share?

There is no universal number. In competitive verticals, 35% to 50% is healthy as long as CPA sits inside target. What matters is the split: share lost to budget points to funding, share lost to rank points to bids, relevance or landing page.

What is the difference between a keyword and a search term?

A keyword is what you add and bid on inside the campaign. A search term is the actual query someone typed that triggered your ad. With broad match, the two can end up very far apart.

How often should campaigns be analyzed?

Search terms and negatives, weekly. Structure, impression share and profitability, monthly. Adjusting bids daily on a mid-volume account usually disrupts algorithmic learning more than it helps.

Do Google Ads and Meta Ads get analyzed the same way?

The layered logic carries over, but the levers differ. On Google you diagnose search intent; on Meta, creative carries most of the performance. The contrast is laid out in how to analyze Meta Ads campaigns.

The real work happens outside the dashboard

People who analyze Google Ads well spend surprisingly little time inside Google Ads. They spend it cross-checking what the platform reports against what the CRM recorded and what finance confirmed. The dashboard is an honest source about the auction and an incomplete one about the business, and treating them as the same thing is what keeps bad campaigns alive for months.

That cross-check is usually the part that stalls, because the data exists but sits in three separate systems. Sherlok connects Google Ads, GA4, your CRM and spreadsheets in one place and answers plain-language questions like "which search terms produced closed deals in the last 60 days", with no manual exports and no spreadsheet gymnastics. Worth testing against your own numbers.