Quick Answer
Google Ads budget leaks in six predictable places: location targeting that reaches people who are not in your market, search terms you never excluded, negative keywords that do not block what you think they block, Performance Max buying traffic you already own, conversions counted more than once, and ad schedules nobody has looked at since launch. This audit finds all six without a single third-party tool.
By Cody Wise, founder of Wise Media. Wise Media builds and manages paid search for Canadian service businesses, agencies and operators. Everything below is drawn from Google’s own published documentation, linked inline so you can verify it against your own account. Advertising spend is a commercial decision specific to your business, so treat this as a diagnostic method rather than financial advice.
Summary
- The biggest leak is a default setting. Google’s documented default for Search, Display and Shopping is “Presence or Interest”, which reaches people who have shown interest in your location as well as people in it.
- You cannot exclude by interest. Google’s location exclusion option is Presence only, and the API value for interest-based exclusion can no longer be used on Search, Display, Performance Max or Shopping campaigns.
- The search terms report is incomplete by design. Google omits terms without enough query activity for privacy reasons, so part of your spend has no visible query attached to it.
- Negative keywords do not match close variants. You must add plurals and synonyms yourself, and a negative appearing after the 16th word of a search may not block the ad at all.
- Performance Max brand exclusions apply to Search and Shopping inventory, and Google warns that brand settings limit reach and performance may decrease.
- Conversion double counting inflates your results and misleads Smart Bidding. The Count setting defaults to Every for website conversions and One for calls from ads.
Table of Contents
- Where does Google Ads budget actually leak?
- Is your location targeting billing you for the wrong country?
- How do you read the search terms report without being misled?
- Negative keywords: four rules most accounts get wrong
- Is Performance Max buying traffic you already own?
- Are you counting the same conversion twice?
- The audit, step by step
- Common mistakes
- Should you run this yourself or bring someone in?
- Frequently asked questions
Where Does Google Ads Budget Actually Leak?
In six places, and five of them are settings rather than strategy. The table below is the audit in miniature. Each row names the leak, where to look for it in the interface, and what the fix is. The sections that follow take each one in turn with the detail you need to actually act.
| Leak | Where to look | What good looks like |
|---|---|---|
| Interest-based location targeting | Campaign settings, Locations, Location options | Presence selected for any business with a physical catchment |
| Unqualified search terms | Insights and reports, Search terms | Reviewed weekly, irrelevant terms added as negatives |
| Negatives that do not block | Shared library, Negative keyword lists | Plurals and synonyms added explicitly, match types deliberate |
| Performance Max eating brand traffic | Campaign settings, Brand settings | Brand excluded from PMax if a brand Search campaign exists |
| Double counted conversions | Goals, Conversions, Count setting | One for leads, Every for sales, no duplicate sources |
| Dead ad schedule and device splits | Campaign, Ad schedule and Devices reports | Reviewed against actual booked work, not clicks |
Run this before you increase spend, before you change agencies, and before you conclude that Google Ads does not work for your industry. In our experience the account is usually not broken, it is just pointed slightly wrong, and the compounding effect of five slightly wrong settings is a budget that produces half of what it should. Paid search is one of the channels we run inside our paid advertising services, and this audit is the first thing we do on any account we inherit.
Is Your Location Targeting Billing You for the Wrong Country?
Probably, if you have never changed it. Google’s advanced location options documentation describes “Presence or Interest” as “this default and recommended option” for Search, Display and Shopping. It reaches people likely to be in your targeted location and people who have shown interest in it. For a Calgary plumber, that includes someone in Phoenix researching Calgary.
The asymmetry nobody mentions
Here is the part that makes this a structural leak rather than a preference. Targeting can be interest-based, but exclusion cannot. Google’s documentation names only one exclusion option, “Location exclusion (Presence)”, states that it is the default, and notes that the interest-based exclusion value can no longer be used for exclusions on Search, Display, Performance Max or Shopping campaigns.
Read those two facts together. You can be opted in to interest-based reach by default, and you have no interest-based lever to opt out with. Excluding the United States does not stop someone in the United States who has shown interest in Calgary from seeing a Calgary ad, because the exclusion only considers where they are. The only reliable control is to switch your positive targeting to Presence.
Which setting is right for you
| Business type | Setting | Why |
|---|---|---|
| Trades, home services, clinics, restaurants | Presence | The customer must be physically present to buy. Interest traffic cannot convert. |
| Local professional services with a catchment | Presence | Same logic. A lawyer in Lethbridge cannot serve a searcher in Lisbon. |
| Hotels, short-term rentals, tourism | Presence or Interest | The buyer is deliberately somewhere else. Interest is the entire market. |
| Relocation, immigration, education | Presence or Interest | The searcher is planning to be in the location, not in it yet. |
| National ecommerce shipping Canada-wide | Presence | Target Canada by presence. Interest adds spend without adding shippable orders. |
| B2B selling into a named market | Test both | Decision makers travel. Split by campaign and compare on qualified leads, not clicks. |
Google’s documentation adds two specifics worth knowing: hotel campaigns use Presence targeting only, and Demand Gen uses both presence and interest as its default and recommended setting. Performance Max defaults its exclusions to Presence.
How to audit it in four minutes
- Open each campaign, go to Settings, expand Locations, then Location options.
- Note which of the two targeting options is selected. If it says Presence or Interest and you sell to people who must be physically present, that is your leak.
- Open the Locations report and add the “Matched location” and “User location” columns. Matched location is what Google decided you were targeting. User location is where the person actually was.
- Sort by cost. Any meaningful spend where user location sits outside your service area is money you can recover today.
Change the setting rather than trying to exclude your way out of it. Exclusions are a blunt instrument here and, as above, they only act on presence.
How Do You Read the Search Terms Report Without Being Misled?
By accepting that it is incomplete. Google states that “some search terms that don’t have enough query activity are omitted from the search terms report in order to keep with our standards on data privacy.” A portion of your spend therefore has no visible query attached to it. The report is the best instrument available, but it is a sample, not a ledger.
Google’s search terms report documentation describes the rest of the behaviour you need to know. The report shows searches that triggered your ads and that a significant number of people used. Search terms insights handles the low-volume queries differently, grouping them into subthemes or an “other queries” group without showing individual queries. The Keyword column, which tells you which of your keywords matched, is hidden by default. And if you view search terms through the Report editor you cannot add keywords or negative keywords directly from the report.
The weekly routine
- Open Insights and reports, then Search terms. Set the date range to the last 7 days.
- Turn on the Keyword column and the Match type column. You are looking for broad match keywords pulling in searches with a different intent.
- Sort by cost descending, not by impressions. The expensive mistakes are at the top.
- Read down until the terms stop surprising you. For most accounts that is the first 30 to 50 rows.
- Classify each unqualified term as one of four types: wrong intent, wrong product, wrong location, or job seeker. Each gets a different negative.
- Add negatives at the right level. Wrong intent at the ad group, wrong product and job seeker at the account, wrong location at the campaign.
- Check Search terms insights for themes your term-level review missed, since that is where the hidden queries surface in aggregate.
Four categories catch nearly everything for Canadian service businesses: the word “free”, the words “job”, “jobs”, “hiring”, “salary” and “careers”, the words “DIY”, “how to” and “template” when you sell a done-for-you service, and competitor names if you have decided not to bid on them.
If your search terms look clean and the problem is downstream, the issue is the landing page rather than the account. The diagnostic in your website gets traffic but no enquiries covers that side, and it is worth ruling out before you spend another week on negatives.
Negative Keywords: Four Rules Most Accounts Get Wrong
Negative keywords behave differently from positive keywords in ways that are documented but rarely read. Google states plainly that “negative keywords don’t match to close variants”, so your ads can still show on close variations of the terms you excluded. Adding “job” does not exclude “jobs”. That one rule explains most of the negatives that quietly fail.
Rule 1: you add the variants yourself
Per Google’s negative keyword documentation, you must add synonyms and singular or plural forms yourself if you want them excluded. Casing and misspellings are handled for you. Practically, that means every negative needs a small family around it: job, jobs, career, careers, hiring, employment, salary, wage.
Rule 2: negative broad match requires every word
Negative match types do not mirror positive ones. Negative broad match, the default, blocks an ad only when the search contains all of the negative terms, in any order. Partial matches do not block it. Google’s own example: “running shoes” as a negative blocks “blue running shoes” and “shoes running”, but not “running shoe”.
| Negative match type | Blocks | Does not block | Use it for |
|---|---|---|---|
| Broad (default) | All terms present, any order | Partial matches, different word forms | Concepts you never want, such as free or DIY |
| Phrase | Terms in the same order, extra words allowed | Reordered terms | Product or service lines you do not offer |
| Exact | Only the exact terms, no extra words | Anything with extra words | A single query you want blocked without over-blocking |
Rule 3: the sixteen word limit
Google documents that if a search has more than 16 words, a negative keyword appearing after the 16th word may not block the ad. Long conversational queries are increasingly common, so a negative that works on a four word search can silently fail on a twenty word one. This is a good argument for excluding concepts early and broadly rather than relying on long tail blocks.
Rule 4: know your ceilings
Google’s published account limits set the architecture for you: 10,000 negative keywords per campaign, 5,000 keywords per negative keyword list, a maximum of 20 negative keyword lists, and a maximum of 1,000 negative keywords applied to Display Network and Video campaigns. Campaign-level negatives also count toward the one million campaign targeting items per account.
With 20 lists available, build them by theme rather than by campaign: Jobs and Careers, Free and DIY, Competitors, Out of Service Area, Adult and Irrelevant, Wrong Product. Apply the universal ones at the account level, the rest per campaign. That structure survives campaign rebuilds, which is the real reason to use lists at all.
The over-correction to avoid
It is possible to negative yourself into a corner. An aggressive broad negative on a common word can block a large volume of qualified searches without appearing anywhere as a problem, because blocked impressions do not show up in a report. If impressions fall sharply after a negatives session, audit what you added before concluding that demand dried up.
Is Performance Max Buying Traffic You Already Own?
If you run Performance Max alongside a brand Search campaign and have not applied brand exclusions, yes. Performance Max serves on Search and Shopping inventory, which means it can win impressions on searches for your own company name. Those conversions were already yours, so the campaign reports a flattering cost per acquisition for traffic that would have arrived anyway.
Google’s brand settings documentation explains the controls. Brand exclusions mean “your campaign won’t serve for queries associated with a brand that you want to avoid”, and for Performance Max they apply to Search and Shopping inventory. Brand inclusions, which restrict a campaign to brand traffic, are available for Search campaigns only.
Four things the documentation tells you that change how you use it
- Exclusions beat inclusions. If a campaign has both, exclusions take priority.
- Brand lists work with broad match to match a brand across languages and spellings, so you do not need to enter every variant of the spelling.
- Variants without a common word need separate entries. Google’s own example is Google, YouTube and Alphabet. If your company trades under more than one unrelated name, list each.
- Google warns about the cost. Brand settings “limit the reach of a campaign and performance may decrease”, and Google recommends using brand lists “only for campaigns that really need them”.
How to decide
Apply brand exclusions to Performance Max when you already run a dedicated brand Search campaign, because otherwise the two compete for the same query and you pay Performance Max prices for brand clicks. Leave brand traffic in Performance Max when you have no brand campaign and your brand volume is negligible, since excluding it then costs reach for no gain.
Measure the change on incremental non-brand conversions over a four to six week window, not on the Performance Max campaign’s own cost per acquisition, which will get worse by design once the easy brand conversions are removed. Our fuller treatment of the channel is in is Performance Max worth it for a Canadian service business.
Are You Counting the Same Conversion Twice?
This is the leak that does not cost you money directly. It costs you money by lying to Smart Bidding. If one enquiry registers as three conversions, every automated bidding decision is made against inflated data, and the algorithm buys more of whatever produced the inflation. Your cost per conversion looks healthy while your cost per actual customer climbs.
The Count setting
Each conversion action has a Count setting with two options. Per Google’s conversion counting documentation, Every counts every conversion after an ad interaction and is “a good choice if you’d like to track and improve your sales, because every sale likely adds value for your business”. One counts a single conversion per ad click and is “a good choice if you’re not interested in the number of sales, but instead whether or not a certain kind of lead was generated”.
The defaults matter because almost nobody changes them. Google documents Every as the default for website, in-app, Analytics transaction and import conversions, and One as the default for calls from ads, Analytics goals, website phone-number calls and app installs.
| Conversion action | Default | What it should usually be | Why |
|---|---|---|---|
| Contact form submission | Every | One | One person submitting three times is one lead, not three |
| Quote request | Every | One | Same enquiry, same job |
| Phone call from the website | One | One | Correct by default |
| Calls from ads | One | One | Correct by default |
| Ecommerce purchase | Every | Every | Each purchase is genuine additional value |
| Booking or appointment | Every | Every | Two bookings are two jobs |
| Newsletter signup | Every | One, and set as secondary | Not a sale. Should not drive bidding |
The three other ways one enquiry becomes three
- Double tagging. The Google Ads tag fires on the thank-you page and a GA4 goal for the same event is imported as a conversion. Pick one source per action.
- Stacked actions. A form submit, a phone click and a thank-you page view are all set as primary conversion actions. One visitor can trigger all three. Keep one primary per outcome and make the rest secondary.
- Thank-you page refreshes. With Count set to Every, a reload or a bookmarked confirmation page counts again.
Two related posts go deeper on this. If your conversion numbers disagree across platforms, read why Google Ads, Meta and GA4 never agree. If your cost per conversion has risen without an obvious cause, the 2026 diagnostic for Canadian advertisers works through the other explanations. And if your conversion volume dropped when consent banners changed, Google Consent Mode v2 in Canada is the place to start.
The Audit, Step by Step
Set aside 90 minutes. Use a 30 to 90 day date range depending on your volume, and work in this order, because fixing targeting before you read search terms changes what the search terms show you.
- Check location options on every campaign. Settings, Locations, Location options. Switch to Presence for any business requiring physical presence. This is usually the single biggest recovery in the account.
- Pull the Locations report with user location. Add the Matched location and User location columns, sort by cost, and quantify what you just recovered.
- Read the search terms report by cost. Last 30 days, Keyword and Match type columns on. Work down the first 50 rows and classify each unqualified term.
- Build or repair the negative keyword lists. Themed lists, plurals and synonyms added explicitly, match types chosen deliberately rather than left on broad.
- Check Search terms insights. This is where the queries omitted from the term-level report surface as themes.
- Audit conversion actions. Goals, Conversions, Summary. Check the Count setting on each, confirm one primary action per real outcome, and look for the same event arriving from two sources.
- Apply Performance Max brand exclusions if a dedicated brand Search campaign exists. Note the date so you can measure the change properly.
- Review the ad schedule. Segment by day and hour. Service businesses that only answer the phone during business hours frequently pay for clicks at 2am.
- Review devices. Segment by device and compare conversion rate, not click volume. Adjust bids only where the difference is consistent across a meaningful sample.
- Check for overlapping campaigns. The same keyword in two campaigns means you are bidding against yourself and paying for the privilege.
- Write down what you changed and the date. Without this you cannot attribute the improvement, and you will not know what to undo if something breaks.
Re-run steps 3 through 5 weekly. The rest quarterly, or whenever you launch a campaign, change a landing page, or inherit an account.
Common Mistakes
- Treating Google’s recommendations tab as an audit. It is an optimisation surface built to increase spend. Some recommendations are good, and the optimisation score is not a measure of account health.
- Auto-applying recommendations. This setting silently changes targeting, keywords and bidding. Turn it off before you audit anything, or you will be auditing a moving target.
- Pausing low-volume keywords. Low volume is not the same as low value. Judge keywords on cost per booked job, not on impressions.
- Adding negatives without checking what they block. Covered above. Broad negatives on common words do real damage invisibly.
- Optimising to conversions nobody verified. If your conversion count includes duplicates, every subsequent decision is compromised.
- Running one campaign for every service. A single budget shared across six services means the cheapest clicks win, not the most valuable jobs.
- Judging the account before the landing page. Half of what looks like wasted spend is qualified traffic reaching a page that does not ask for the business.
- Changing six things at once. You will have no idea which one worked.
Should You Run This Yourself or Bring Someone In?
Run it yourself at least once. Every step above uses reports Google gives you for free, and the understanding you get from reading your own search terms is worth more than the time it costs. Bring someone in when the account is large, when the same leaks keep reappearing, or when the audit finds nothing and performance is still poor.
Doing it in house
- For: No tooling cost. Everything is native to the Google Ads interface.
- For: You learn what your market actually searches for, which improves your website copy and your sales calls as much as your ads.
- For: The fixes are reversible. Nothing in this audit is destructive.
- Against: It is 90 minutes to audit and roughly 30 minutes a week to maintain, indefinitely.
- Against: Judging whether a structural change helped requires enough conversion volume to read a result, and small accounts often lack it.
Bringing in help
- For: Pattern recognition. Someone who has audited a hundred accounts finds the overlap and the duplicate conversion action in minutes.
- For: Structural work such as campaign restructures and bidding strategy changes needs someone who can judge the risk.
- Against: Management fees are typically quoted as a percentage of spend or a flat monthly amount, published rates vary widely across the Canadian market, and most providers do not publish a methodology for how the fee relates to the work. Ask what the fee buys in hours and deliverables, and get it in writing.
- Against: An agency that will not show you the search terms report is managing your perception rather than your account.
If you want this run for you, it is the first thing included in our paid advertising packages. We hand back the findings in writing whether or not you continue with us. For the budget-setting question that usually follows an audit, our Google Ads budget guide for Canadian small businesses is the companion piece. And if email is part of your follow-up, the 2026 email deliverability checklist covers the channel that catches the leads this one generates.
Frequently Asked Questions
How much Google Ads spend is typically wasted?
There is no credible universal figure, and the ones circulated online are usually vendor marketing without a stated methodology. The honest answer is that it depends entirely on account structure, industry and how long it has been since anyone reviewed the settings. Rather than trusting a percentage, run the audit above and quantify your own number from the Locations and search terms reports, which is a figure you can actually defend.
Why are my ads showing outside my target location?
Almost always because location targeting is set to Presence or Interest, which Google documents as the default and recommended option for Search, Display and Shopping. It reaches people who have shown interest in your location as well as people in it. Switch to Presence. Excluding other countries will not solve it, because Google’s location exclusion option considers presence only.
Why do some search terms not appear in the report?
Google omits search terms that do not have enough query activity, which it attributes to its data privacy standards. Those queries are not lost entirely. Search terms insights groups them into subthemes or an “other queries” group so you can see the pattern without seeing individual searches. Check both reports rather than assuming the term-level view is complete.
Do negative keywords block misspellings and plurals?
Partly. Google states that negative keywords do not match to close variants, and that you must add synonyms and singular or plural forms yourself. Casing and misspellings are handled automatically. In practice, add every plural and synonym explicitly and do not assume a single negative covers a family of related terms.
Should I exclude my brand from Performance Max?
Yes if you already run a dedicated brand Search campaign, because otherwise Performance Max competes for queries you were going to win anyway and reports the resulting conversions as its own. No if you have no brand campaign and minimal brand search volume, since Google notes that brand settings limit reach and performance may decrease. Measure the change on incremental non-brand conversions over four to six weeks.
How often should I run this audit?
The search terms and negatives portion weekly, which takes about 30 minutes once the lists exist. The full structural audit quarterly, and additionally whenever you launch a campaign, change a landing page, inherit an account from another provider, or notice cost per conversion moving without an obvious cause.
Is the Google Ads optimisation score worth paying attention to?
Treat it as a list of Google’s suggestions rather than a measure of account health. Some recommendations genuinely help, particularly on tracking and ad strength. Others increase reach and spend in ways that may not suit a business with a fixed service area or a limited capacity to deliver work. Review each one individually and keep auto-apply switched off.
Can I audit an account I do not have admin access to?
You need at least read-only access to run the reports in this article. If your current provider will not grant it, that is itself a finding. Ask for standard access to the account under your own Google account rather than a shared login, and confirm the account is owned by your business rather than by the agency, so you keep the history if you ever change providers.
The Bottom Line
Most Canadian Google Ads accounts are not failing because the strategy is wrong. They are leaking because a handful of defaults were never changed and nobody has read the search terms report in months. Every leak in this article is visible in a free report, and every fix is reversible.
Start with location options today. It takes four minutes per campaign and it is the change most likely to recover meaningful spend in the first week. Then book the weekly half hour for search terms, because that is the habit that keeps the account clean once you have fixed it.
If you would rather have the audit done and handed to you in writing, including the wasted spend figure for your own account, start with our intake form and tell us what you are spending and where.
