Reading a Google Ads report without going crazy.
Google Ads has 5 different campaign types, 100+ columns, and a “Recommendations” tab that’s mostly trying to get you to spend more. Here’s what actually matters, what each metric means in plain English, and how to spot when Google’s algorithm is wasting your money.
Campaign types matter.
Before you read any metric, you need to know which type of campaign you’re looking at. Google Ads is really 5 different platforms wearing one trench coat:
- Search — text ads on Google search results. Highest intent, most expensive clicks.
- Performance Max (PMax) — automated campaign that runs across Search, Display, YouTube, Gmail, Discover, and Maps. Black box, but powerful when fed good data.
- Display — image/banner ads on the Google Display Network (millions of partner sites). Cheap clicks, low intent.
- Shopping — product listing ads with images and prices. Ecommerce-only, driven by your product feed.
- Video — YouTube ads (in-stream, in-feed, shorts). Brand and awareness focused.
Why this matters: a “good” CPC on Search ($3-15) is a “terrible” CPC on Display ($0.30-1). A 1% CTR on Display is healthy; on Search it’s a sign of broken targeting. Always check campaign type before judging metrics.
As of late 2024, Google pushes Performance Max for almost everything. It hides most of the levers you’d normally pull on Search/Display campaigns — you can’t see exactly which keywords or placements are working. It’s powerful but opaque. Many advertisers run Search + PMax together so they can see what’s happening AND let Google’s algorithm do its thing.
The 9 metrics that matter.
Out of the 100+ columns Google offers, these are the ones we look at every time. Listed roughly in importance order — start with the top.
How many actual outcomes did your spend produce, and what did each one cost. This is your starting point. For ecommerce, also look at Conv. Value (revenue) and ROAS. For lead gen, this is essentially CPL. Like Meta, compare against your customer LTV — if cost-per-conversion is greater than what a customer is worth, you have a problem.
What you pay per click. Search CPC is dramatically higher than Display because intent is higher. Mortgage Search CPC: $5-25. Local services Search CPC: $3-12. Display CPC: $0.30-2. If your Search CPC suddenly spikes, check Quality Score and competition.
Of the people who clicked through, what % converted. This is the cleanest measure of whether your landing page works for the traffic this campaign sends. Healthy Search conversion rates: 3-8% for lead gen, 1-3% for ecommerce. If CTR is good but conversion rate is bad, the problem is the page, not the ad.
Of people who saw your ad, what % clicked. What “good” looks like depends entirely on campaign type. Search ads on branded terms: 10-30%. Search on competitive non-brand: 3-8%. Display: 0.3-1%. Shopping: 0.7-1.5%. Below baseline = ad copy/creative problem or weak relevance.
Google’s 1-10 rating of how relevant your keyword + ad + landing page combination is. Higher Quality Score = lower CPC for the same ad position. Aim for 7+. We’ll dig into how to actually improve this in the next section.
Out of all the times your keywords could have shown an ad, what % did. If Search IS is 30%, you’re missing 70% of available volume. Causes: low budget, low Quality Score, or low bids. Pair with the next two metrics to know which one.
If this is high (>20%), you’re hitting your daily budget early and missing impressions later in the day. Either raise the budget or lower bids to stretch coverage. Easy win when ROAS is healthy — more budget = more revenue at roughly the same efficiency.
Your bid + Quality Score isn’t winning the auction often enough. Three options: raise bids, improve Quality Score, or accept that this keyword auction isn’t profitable for you. If you keep raising bids without improving QS, you’re in a doom loop.
Same idea as Meta’s ROAS. For ecommerce this is the headline. For lead gen, only meaningful if you’re feeding offline conversion values back to Google (highly recommended). Healthy: 3-5×. Best-in-class: 8×+. If you’re below 2× consistently, your math probably doesn’t work.
Quality Score, demystified.
Quality Score is Google’s most-misunderstood metric. The bad news: it’s not directly fixable like a knob. The good news: improving it is straightforward, and it directly lowers your CPC.
How Quality Score breaks down
Each keyword gets a 1-10 score made up of three components:
1. Expected CTR — Google’s prediction of how often this keyword + ad combination will get clicked. Improve by writing better ad copy that matches the keyword’s search intent.
2. Ad Relevance — How closely your ad copy matches the keyword. Improve by including the keyword in your ad headline, description, and final URL.
3. Landing Page Experience — Page speed, content relevance to keyword, mobile-friendliness, and trust signals. Improve by sending each ad group to a page that matches that group’s intent.
The single biggest QS lever
If you only do one thing: break your campaigns into tighter ad groups. One keyword theme per ad group, one ad set per page that’s about that exact theme. Don’t run “marketing services” and “marketing automation” through the same ad group going to the same generic homepage. Tight matches = high relevance = high QS = low CPC.
A worked example.
Here’s a 30-day Search campaign report for a hypothetical home services business. What would you do with each campaign?
| Campaign | Spend | CPC | CTR | Conv. | Cost/Conv | Search IS | Status |
|---|---|---|---|---|---|---|---|
| A · Brand Search | $420 | $1.80 | 22% | 38 | $11 | 94% | Active |
| B · Service Keywords | $3,800 | $8.40 | 5.2% | 52 | $73 | 28% | Active |
| C · Competitor Bidding | $1,650 | $14.20 | 2.1% | 8 | $206 | 42% | Paused |
| D · PMax (all conv) | $2,200 | $3.60 | — | 31 | $71 | — | Learning |
How we’d read this
Campaign A (Brand Search) — Excellent. Branded search always performs well because anyone searching your name is high-intent. CPC is low ($1.80), CTR is high (22%), Cost/Conv is healthy ($11). Search IS at 94% means you’re capturing nearly all your branded traffic. Action: don’t touch this. It works.
Campaign B (Service Keywords) — Mixed. Cost/Conv of $73 is workable for home services. The big issue: Search IS at 28% means you’re missing 72% of available volume. Action: pull the IS Lost columns. If it’s “lost to budget,” raise budget. If it’s “lost to rank,” improve Quality Score by tightening ad groups and refining ad copy.
Campaign C (Competitor Bidding) — Broken. CPC of $14 is brutal because competitor terms always have low Quality Score (your landing page isn’t relevant to “competitor name”). 2.1% CTR is below baseline. Cost/Conv at $206 is unworkable. Action: stay paused. Competitor bidding rarely works for small/mid budgets — it’s a strategy for incumbents with deep pockets.
Campaign D (PMax) — Still learning. PMax needs 30-50 conversions to exit “learning mode” and stabilize. We have 31 in 30 days, so it’s just exiting now. Cost/Conv of $71 is similar to Campaign B — that’s a good sanity check. Action: let it run another 30 days, then compare. PMax often improves dramatically after 60+ days as it learns your conversion patterns.
Red flags.
Click “Insights” or “Search Terms” on a Search campaign. If you see queries unrelated to what you sell, you’re paying for clicks from people who’ll never buy. Add them as negative keywords. This is the single most-impactful weekly task in Google Ads.
Your CPC is paying a “Quality Score tax” — you’re paying more than you need to for the same ad position. Tighten ad groups, write keyword-rich ads, ensure landing pages match the search intent. Often improves CPC by 30-50% over 2-4 weeks.
You’re losing more than half your potential auctions to competitors. Either you’re outbid, your QS is too low, or this auction simply isn’t profitable for you. Investigate before raising bids — you might just be losing money faster.
CPC jumps 20%+ overnight without you changing anything. Usually it’s new competition entering the auction. Check Auction Insights to see who joined. You either accept higher CPC or differentiate (better ad copy, better page, better offer).
PMax hides which assets and audience signals are working. If you can’t see “Asset Group Performance” with at least Best/Good/Low ratings on each asset, you’re flying blind. Add 3-5 asset groups and feed PMax customer audience signals so it has something to work with.
Engaged-view conversions count someone watching 10 seconds of a YouTube ad. View-through counts someone seeing a Display ad and converting later. Both inflate conversion numbers without representing real intent. For honest reporting, exclude these from the main Conversions column or create a custom column with Click-only conversions.
Good signs.
Means your ad relevance and landing page are properly aligned. You’re winning auctions for less, your CPC is sustainable, and Google is actively rewarding you with better placement.
This is a great problem. You have profitable campaigns and you’re being capped by budget rather than market dynamics. Increase budget by 15-25% per week and watch what happens to Cost/Conv. If it stays stable, keep increasing.
Branded traffic should always be your cheapest. If brand search Cost/Conv is healthy and you’re getting decent volume, your other marketing is working — people are seeing you elsewhere and then googling you.
PMax learns from your conversion data. If month 1 Cost/Conv is $100, month 2 is $80, month 3 is $65 — the algorithm is finding better audiences and placements over time. Don’t kill it prematurely; the second 90 days are usually better than the first 30.
The “Recommendations” tab trap.
Google Ads has a “Recommendations” tab that scores your account and offers suggestions to improve “Optimization Score.” Most of these recommendations are not in your interest.
Common bad recommendations to ignore:
- “Apply ‘Maximize Conversions’ bidding” — gives Google control of your bids without setting a target Cost/Conv ceiling. Often 2-3× your CPC overnight.
- “Add broad match keywords” — broad match brings in irrelevant traffic. Sometimes useful in PMax-paired campaigns; usually a CPC drain in standalone Search.
- “Add automatically created assets” — Google writes ad headlines for you. Quality varies wildly. Disable unless you actively review what gets generated.
- “Increase budget” — only valuable if your Cost/Conv is below your target. Otherwise you’re paying Google more without proven ROI.
- “Apply final URL expansion” — sends people to whatever page Google thinks matches their query, including pages you didn’t choose. Loss of control.
Genuinely useful recommendations to follow:
- Adding negative keywords (when relevant)
- Improving ad strength on responsive search ads
- Fixing disapproved ads
- Adding sitelink and callout extensions
Rule of thumb: if applying a recommendation would change your spend, attribution, or targeting, evaluate it manually. If it just adds metadata (extensions, copy variants), it’s usually fine.
Takeaways.
- Always check campaign type first. Search, PMax, Display, Shopping, and Video play by different rules.
- Cost/Conversion is the headline. Compare it against your customer LTV every single time.
- Search Terms report is your highest-leverage weekly task. Add negative keywords. Always.
- Quality Score directly lowers your CPC. Tighten ad groups → improve QS → pay less per click.
- Impression Share Lost columns tell you why you’re missing volume. Budget vs rank requires different fixes.
- PMax needs 30-60 days minimum to stabilize. Don’t kill it after a bad week.
- Ignore most “Recommendations.” Google’s optimization score is for Google, not for you.
Google Ads is more deterministic than Meta but also more punishing. A bad keyword match can burn money fast. A neglected Search Terms report can waste 30% of spend. The good news: once you know what to look at, fixing things is mostly tactical work.
Want us to read your reports?
Our paid ads team manages $10M+ in spend across Meta and Google. We’ll audit your current campaigns and tell you exactly what we’d change. Free 30-minute strategy call, no obligation.
Book a free strategy call