Google Ads: structure, exclude, measure
Google Ads buys intent: someone is already searching for what you sell. That is what makes it the paid channel with the fastest return, and also the easiest to waste. Most of the wasted spend comes from three places: irrelevant search queries, campaign structure and the landing page.
- Quality Score can make the cost per click vary twofold at an equal bid.
- The actual search terms, not the keywords you picked, reveal where the budget goes.
- A campaign that sends traffic to the homepage pays more and converts less. The effect compounds.
- Automated bidding needs valued conversions, otherwise it optimizes toward nothing.
On this page
Quality Score
Quality Score is Google's rating of how relevant an ad is, measured from 1 to 10. It combines three components: the expected click-through rate, the ad's relevance to the query, and the experience offered by the landing page. Together with the bid, it determines the position obtained and the cost actually paid per click.
Google Ads buys intent, not an audience
This is the fundamental difference from social advertising, and it drives everything else. On social media, you interrupt someone who was not looking for anything. On Google, you appear in front of someone who expressed a need two seconds ago.
Three practical consequences follow.
| Consequence | Why |
|---|---|
| Return comes faster | No need to create demand, only to capture it |
| Volume is capped | You cannot exceed the number of people searching |
| The message matters less than the match | An average ad on the right query beats a brilliant ad on the wrong one |
The volume ceiling is what surprises people most. Doubling a budget on a narrow market does not double the results: past a certain point, you buy less and less relevant queries, and the cost per acquisition climbs.
What to check before increasing spend
The wasted spend on this channel almost always comes from irrelevant queries that no one is watching. The search terms report shows exactly what your budget paid for, query by query, and it takes ten minutes to read.
- When was the search terms report last reviewed, and what came out of it?
- What share of the budget went to queries that have nothing to do with us?
- How many leads from this channel turned into sales last quarter?
- Do we know our cost per sale by campaign, or only the cost per click?
- At what cost per sale would we stop bidding on a keyword?
A good answer cites queries excluded recently and a cost per sale. A hollow answer talks about Quality Score, click-through rate, or proposes broadening the targeting to get more volume.
Knowing what share of your budget is lost to useless queries can be verified in a single meeting. A 90-minute consultation settles it on your actual account, with a written summary your team or your agency can execute.
When Google Ads is not the right first channel
The channel is excellent, it is not universal. Four situations where the money sits better elsewhere, at least for a while.
Nobody searches for your category
A genuinely new product has no keywords yet. Demand gets created elsewhere, on social platforms or through content, before paid search has anything to capture.
The site does not convert yet
Buying traffic to a page that does not convert means paying to measure a problem you already know about. Fix the page first, run the campaign after.
Nothing connects an enquiry to a sale
Without a system linking the enquiry received to the sale closed, you will steer on forms rather than on revenue. The channel will look like it is performing while your profitability stays unknown.
The budget does not reach the learning floor
Spend spread across too many campaigns never produces enough conversions for the system to learn. One market run seriously beats four markets barely touched.
Structuring a campaign without overcomplicating things
Complex structures with thirty ad groups date from a time when you had to guide the algorithm by hand. Today they fragment the learning data without adding anything.
One campaign per business objective
Not per product, not per region, unless the budgets genuinely have to be kept separate. Each additional campaign divides the conversions available for learning.
Ad groups by intent, not by keyword
"Emergency repair" and "repair price" are two different intents that call for two different ads. "Fast repair" and "express repair" are the same intent.
One landing page per group
The page headline echoes the ad's promise. It is the component of Quality Score most often neglected, and it acts directly on the cost paid.
A negative keyword list kept up to date
It is built from the search terms report, weekly at first, then monthly. It is the least spectacular and most profitable work on the channel.
Starting an account, from the settings to the first 90 days
Four things get settled before the account is opened. None of the four happens inside Google Ads, which is exactly why they get skipped, and why they get paid for during the first quarter.
Confirm the demand exists
Google Ads captures intent, it does not create it. Confirm that search volume exists on your terms, in your territory, before opening anything.
Wire up and test conversion tracking
A conversion that never reports back is a sale the system cannot see, so one it will not go looking for. Make a test call, fill in the form yourself, and confirm the event shows up in the account.
Prepare a page that repeats the promise
The landing page headline picks up the wording of the ad. That is what lowers the cost per click and raises the conversion rate at the same time.
Set the maximum cost per enquiry
Start from your margin per sale and from the share of enquiries that become sales. That number is your limit: it is decided before the first dollar is spent, never during.
The budget itself is not set in dollars per month. It is set by the number of conversions the campaign can produce each week, because that volume is what lets automated bidding learn. Multiply the average cost per click in your industry, the number of clicks it typically takes to get one conversion, and the number of conversions you want each week: that is your weekly floor. Below that floor, the campaign proves nothing, it just spends slowly.
| Period | What you do | What you watch |
|---|---|---|
| Weeks 1 to 2 | One campaign, one intent, search network with no partners | That ads are serving and conversions are reporting back |
| Weeks 3 to 6 | Weekly reading of search terms, negatives added every week | The share of budget spent on irrelevant queries |
| Weeks 7 to 10 | Move to automated bidding once conversions are stable | Cost per acquisition, not cost per click |
| Weeks 11 to 13 | The decision: scale, hold, or close | Cost per real sale, taken from your sales system |
The daily budget is not a daily cap. Google can spend more on a high-demand day and compensate on another, balancing out over the month. An owner checking the account on a Tuesday evening sees a number that worries them for no reason. It is the monthly total that should be read.
Where the budget goes
The search terms report shows what people actually typed, as opposed to the keywords you chose. It is the first place to look on any account that disappoints.
| Leak | Symptom | Fix |
|---|---|---|
| Off-topic queries | Unrelated terms in the report | Negative keywords |
| Job seekers | "job", "salary", "career" | Systematic exclusions |
| Looking for free | "free", "DIY", "yourself" | Exclusions, unless it is a lead offer |
| Competitors and comparison sites | Third-party brand names | Case-by-case decision |
| Out-of-area traffic | Clicks coming from elsewhere | Target by presence, not by interest |
| Networks enabled by default | Clicks from sites and apps you never chose | Search only, at the start |
| Automatically created assets | Ad headlines nobody on your side approved | Turn them off until the message is settled |
| Unsuitable landing page | High bounce rate, low conversions | A dedicated page per ad group |
The location setting includes by default people who show an interest in your region without being in it. On a local service, that can represent a significant share of the spend. The setting to keep is actual presence.
What stays in-house is the list of what you do not sell. The terms adjacent to your offer, the products you stopped carrying, the requests you turn down: that list is what feeds the exclusions, and it comes from sales, not from the ad account. What gets delegated is everything else: campaign structure, bidding strategy, reading the search terms, the match between ad and landing page. An account run by someone who does not know what the business refuses to sell spends in the wrong place every month.
Choosing a bidding strategy
Automated bidding works well, on one condition: it needs conversions that are correctly measured and valued to know what to optimize toward.
| Strategy | When to use it | Condition |
|---|---|---|
| Manual | Startup, very low volume | Management time available |
| Maximize clicks | Data-gathering phase | A limited, watched budget |
| Maximize conversions | Reliable tracking, steady volume | At least a few conversions per week |
| Target cost per acquisition | Known, stable acquisition cost | Enough conversion history |
| Target return on ad spend | Online store with values passed through | Accurate conversion values |
The last two rows changed nature on August 17, 2026. In a campaign limited by its budget, the target you entered is no longer a safety ceiling: it is the return the system optimizes toward.
Switching to automated bidding without valued conversions amounts to asking the system to optimize toward a target it cannot see. It will then optimize toward what it does see, usually the click, which produces volume without sales.
What to measure
The cost per click is the most watched indicator and the least decisive. It can rise while profitability improves, if the clicks you buy convert better.
| Rank | Indicator | What it tells you |
|---|---|---|
| 1 | Cost per acquisition | Whether the channel is profitable |
| 2 | Return on ad spend | Whether scaling is possible |
| 3 | Conversion rate by group | Where the gap is |
| 4 | Budget lost impression share | Whether there is volume left to capture |
| 5 | Quality Score | Where the cost can come down |
| 6 | Cost per click | A diagnostic, not a target |
On Google Ads, you do not win by paying more. You win by paying less for the same clicks, and by ceasing to pay for the others.
Falia analysis gridQuick account diagnostic
Building landing pages is detailed in what makes a landing page convert. The difference from social buying is covered in how to choose an advertising platform. Execution is covered on the Google Ads and search engine marketing pages.
Our test of ads inside ChatGPT, with the numbers, is in our test of ads inside ChatGPT.
Google Ads is one lever of the Optimize the profitability of your digital campaigns goal.
Already running a marketing team? See how we plug in as reinforcement on Google Ads.
Frequently asked questions about Google Ads
What is Quality Score on Google Ads?
It is Google's rating of how relevant an ad is, scored from 1 to 10. It combines the expected click-through rate, the relevance to the query and the experience offered by the landing page. Together with the bid, it determines the position and the cost actually paid.
How long before Google Ads shows results?
The first clicks arrive the same day. The first reliable conclusions take six to twelve weeks: the time for the campaign to gather enough conversions for automated bidding to learn, and for negative keywords to have cleaned out the irrelevant queries.
What budget do you need for Google Ads?
It depends on the cost per click in your sector and the number of conversions needed to steer by. The ceiling matters as much as the floor: on a narrow market, doubling the budget does not double the results and drives the cost per acquisition up.
Why do my ads attract irrelevant traffic?
Because keywords trigger adjacent queries you did not choose. The search terms report shows what was actually typed, and it feeds the negative keyword list. It is the most profitable work on the channel.
Should you use automated bidding?
Yes, provided conversions are correctly measured and valued. Without that, the system optimizes toward what it sees, usually the click, which produces volume without sales.
Can you send ads to the homepage?
It is possible but costly. The homepage does not echo the ad's promise, which degrades the page experience, therefore the Quality Score, therefore the cost per click. Conversion falls at the same time as the cost rises.
Google Ads or organic search?
Both answer the same intent by different means. Google Ads delivers a result in a few days but stops when the budget stops. Organic search takes months and keeps producing afterwards. Together, they form SEM.
Can a small business manage Google Ads in-house?
Yes, on two conditions: that someone reads the search terms report every week, and that cost per sale is tracked outside the platform. Without those two habits, an account left alone drifts toward broader and broader queries.
- Google Ads Help, About Quality Score, accessed July 2026.
- Google Ads Help, About campaign location targeting, accessed July 2026.
- Google Ads Help, About Smart Bidding strategies, accessed July 2026.
- Google Search Central, Core Web Vitals, accessed July 2026.
Gabriel almost always takes your first call and carries out your audit. He builds the strategy starting from your growth goal: where to put your budget, which market to test and how to connect each lead to a real sale in your CRM. He mainly leads engagements for three goals: Optimize the profitability of your digital campaigns, Develop a new market, and Generate demand and growth. With Geneviève, he also works on organic search (SEO), AI visibility (GEO) and conversion rate optimization (CRO). The sales a Google Ads or Meta Ads campaign brings in depend on the page that receives the click. He writes mainly about marketing strategy, paid advertising and measurement.
About Falia →
SEO
Paid advertising
Conversion
AI visibility
Web design