You added the keyword "bookkeeping services" — so why did your ad show for "free bookkeeping course"? Match types. Every keyword carries a match type that tells Google how loosely to interpret it, and the modern versions are looser than most tutorials admit.
Understanding the three types — and the two tools that police them — is the difference between a campaign that targets and one that leaks. This guide assumes a campaign exists (start here if not) and feeds the weekly routine that keeps spend honest. Budgets and bids have their own guide.
Quick Facts
| Detail | Information |
|---|---|
| Provider | Fredeveloper Academy |
| Topic | Keyword match types, the search terms report, and negatives |
| Best for | Anyone running or about to run search campaigns |
| The three types | Broad (loosest), phrase (moderate), exact (tightest — still not literal) |
| The control tools | Search terms report + negative keywords |
| Beginner default | Phrase and exact first; broad only with data and guardrails |
| Coming soon | Google Ads course — join the waitlist |
| Last updated | 27 March 2026 |
The ContextWhy Match Types Are Where Budgets Leak
The same keyword can be a sniper or a sprinkler depending on one setting. You're probably here because:
| Situation | Why it matters |
|---|---|
| Your ads show for bizarre searches | That's match-type looseness — diagnosable in the search terms report in minutes |
| Broad match spent your budget in a day | Broad casts the widest net by design; without negatives and conversion data it's a leak |
| Exact match isn't showing much | Exact trades volume for precision — normal, and fixable by adding close variants deliberately |
| Tutorials contradict each other | Match behavior has loosened over the years; old guides describe rules that no longer hold |
The ThreeWhat Each Match Type Actually Does
Infographic 01 · The Spectrum
One keyword, three behaviors
The pattern: broad buys reach and discovery at the cost of precision; exact buys precision at the cost of volume; phrase sits usefully between. None is "best" — they're different tools, and the mix should shift as your data grows.
The PoliceThe Search Terms Report & Negative Keywords
Match types propose; two tools dispose. The search terms report shows the actual searches that triggered your ads — not your keywords, the real queries. Read it like a bank statement. Negative keywords are the block list: add "free", "jobs", "course", "DIY" (or whatever your waste pattern is) and those searches stop costing you. Negatives are the highest-ROI five minutes in Google Ads — every one permanently plugs a leak.
The StrategyA Beginner Mix That Doesn't Leak
Infographic 02 · The Progression
Match strategy as data grows
The harvest loop in step 3 is the quiet engine: real searches that converted get promoted into exact-match keywords, so proven spend gets prioritized. Over months, your best keyword list writes itself out of the search terms report.
The JudgmentWhen Each Type Earns Its Place
| Use this | When |
|---|---|
| Exact match | Your proven converters and core service terms — protect and prioritize them |
| Phrase match | The launch workhorse: meaningful reach with the intent still attached |
| Broad match | Discovery, once conversion tracking has data to guide it and negatives guard the flanks |
| Negative keywords | Always, everywhere, reviewed weekly — the list should grow for the life of the account |
Guided PassReading a Search Terms Report: A Guided Pass
The skill is classification: every row in the search terms report falls into one of five buckets, each with a fixed action. Here's a guided pass over a typical week's report for a campaign selling bookkeeping services.
| The actual search | The bucket | The action |
|---|---|---|
| “small business bookkeeping services” — converted | Proven winner | Promote to exact match in its own right; protect its budget |
| “bookkeeping packages for cafes” — clicked, no conversion yet | Promising variant | Leave running; watch. If it converts, promote; consider a cafe-specific ad later |
| “free bookkeeping templates” | Wrong intent | Negative: 'free' and 'templates'. This searcher will never buy |
| “bookkeeping jobs near me” | Wrong audience | Negative: 'jobs', 'careers', 'hiring' as a standing set |
| “what does a bookkeeper do” | Researcher, not buyer | Negative for THIS buyer campaign — but a content idea for the website's SEO |
Run every row through those five buckets and the report stops being noise. The cadence: daily glances in week one of any campaign (waste shows up fast), then a weekly fifteen-minute pass forever. Each pass should produce two artifacts — new negatives, and any promotions to exact — and a pass that produces neither means the targeting has matured, which is its own useful signal.
Note the last row's second life: wrong-intent searches for ads are often right-intent topics for content. The terms report quietly doubles as free keyword research — real queries, real volume, zero tools — which is one more reason it's the most valuable screen in the platform.
Worked ExampleReading One Search Terms Report — Line by Line
Illustrative excerpt from a bookkeeping-service campaign's first fortnight, keyword “bookkeeping services” (phrase match). The report shows: “affordable bookkeeping services” — relevant, converting → promote to exact match. “bookkeeping services price list” — relevant, no conversions yet → keep, watch. “free bookkeeping software” — wrong intent → negative “free” AND “software”. “bookkeeping jobs work from home” — job seekers → negative “jobs”, “hiring”, “salary”. “what is bookkeeping” — research intent, wrong campaign → negative here (it might suit a future awareness campaign, not this one). Ten minutes of reading produced two permanent leak-plugs, one promotion, and a cleaner account than most run after a year. This exact ritual, weekly, is what “managing” a search campaign mostly is.
Infographic 03 · The Triage
Every search term gets one of four verdicts
Field NotesMatch-Type Mistakes in the Wild
| The mistake | The symptom | The fix |
|---|---|---|
| All broad match at launch | Budget gone, terms report full of strangers | Phrase + exact until data exists |
| Negatives added once, never again | New leaks open as Google explores | The weekly triage — leaks are continuous |
| Negating too aggressively | Volume collapses; good terms blocked | Negative PATTERNS you're sure of, not every odd term |
| Same keyword in multiple ad groups | You bid against yourself; data splits | One home per keyword theme |
| Ignoring close variants on exact | ‘Exact’ surprises in the report | Triage them too — exact is tight, not literal |
Build a shared negative list of universal wrong-intent words for your business — free, jobs, salary, course, DIY, template — and apply it account-wide on day one. It pre-plugs the leaks every campaign would otherwise pay to discover.
ToolboxThe Targeting Toolkit
Everything match-type management needs: the search terms report (the truth screen — calendar the weekly read); shared negative lists (universal wrong-intent words applied account-wide, plus campaign-specific lists); a promotion habit (converting terms become exact-match keywords in their themed ad group); and a simple log of what you negatived and promoted, so the account's evolution is explainable — to a client, or to yourself in six months. This system feeds everything downstream: cleaner traffic makes copy tests readable and budget math honest, and it's the first place to look when conversions go missing.
GlossaryTerms You'll Meet Around Targeting
| Term | Plain-English meaning |
|---|---|
| Match type | How loosely Google may interpret your keyword — broad, phrase, exact |
| Close variant | Plurals, misspellings, same-intent rephrasings — allowed even on exact |
| Negative keyword | A word/phrase that blocks your ads from showing |
| Shared negative list | One negative list applied across campaigns |
| Search term | What the person ACTUALLY typed — vs the keyword you bid on |
| Query sculpting | Steering traffic between keywords/groups using negatives |
Negative ArchitectureBuilding a Negative List That Scales
As accounts grow, negatives need structure or they sprawl into an unmanageable pile. The architecture that holds: three tiers. Account-level universals — the words that are wrong for your business everywhere, always: "free", "jobs", "careers", "salary", "course", "DIY" for most service businesses. Set once, they guard every campaign you'll ever build. Campaign-level negatives — wrong for this goal but maybe right for another: "cheap" might be negative for your premium campaign and acceptable for a budget-line one. Cross-group negatives — the traffic-cop tier that stops your own ad groups competing: if one group targets "bookkeeping services" and another "catch-up bookkeeping", each gets the other's term as a negative, so every search triggers the ad group whose copy actually mirrors it.
Maintain the tiers with a simple ritual: every weekly terms-report pass, each new negative gets placed deliberately — universal, campaign, or cross-group — rather than dumped wherever the screen happens to be open. A negative in the wrong tier either blocks too much or guards too little; thirty seconds of placement thinking per negative keeps the architecture true. Within a few months, the universal list alone becomes a portable asset — a battle-tested file you carry into every new campaign and, if you freelance, every new client account, where it routinely saves real money in week one.
ScenariosFour Account Situations, Four Match Moves
| Situation | The move |
|---|---|
| Brand-new account, no conversion data | Phrase + exact only, tight negatives from day one — broad has nothing to steer by yet |
| Budget vanishing on junk searches | Emergency triage: pause broad keywords, read 30 days of terms, rebuild the negative list, relaunch tighter |
| Exact-match campaign, volume too low | Add phrase variants of your proven terms and mine the terms report for close cousins — expand from strength |
| Mature account, conversions flowing | Now broad earns its place: one scouting campaign, separate budget, conversion-based bidding, heavy negatives |
Match strategy is situational, not ideological. The same broad match that wrecks a new account is a legitimate discovery engine in a data-rich one — the difference is entirely what the system has to steer by.
Deeper DiveNegative Match Types: The Detail That Bites
Here's the asymmetry that catches even intermediate advertisers: negative keywords don't use close variants. A regular keyword forgives plurals and misspellings; a negative blocks only what it literally says. Add “job” as a negative and “jobs” still triggers ads; block “free” and “for free” may still slip through in some structures. The practical fix: add the variants deliberately — singular and plural, common misspellings of your worst offenders — and check the terms report for leaks around your existing negatives, not just new waste.
Structure matters too. Negatives can live at three levels: account-wide lists (universal junk — “free”, “jobs”, “diy”, “salary” — built once, applied everywhere), campaign negatives (terms wrong for this goal), and ad-group negatives for traffic-shaping — blocking a term from one group so it routes to the group with the better ad. That last technique, called sculpting, is how tidy accounts make sure each search meets its best-matching ad. Build the universal list once; it protects every campaign you'll ever launch.
ProgressMatch Health, Measured Monthly
| Metric | Healthy direction |
|---|---|
| Wasted-spend share (junk terms ÷ total spend) | Falling month over month as negatives accumulate |
| Negative list size | Growing — a static list means the terms report isn't being read |
| Share of spend on exact-match proven terms | Rising — the harvest loop concentrating money on winners |
| New converting terms found per month | Steady trickle — discovery still working, not just defense |
The first metric is worth computing by hand quarterly: tag a month of search terms as relevant or junk, total the junk spend, and watch that percentage fall as the system tightens. It's the single clearest proof — to yourself or a client — that match-type discipline is paying.
RisksChallenges & Misconceptions
| Misconception | The honest version |
|---|---|
| “Broad match is just bad” | Unsupervised broad is bad. With conversion data and a strong negative list, it's a legitimate discovery tool |
| “Exact match means literally exact” | Close variants apply — plurals, misspellings, same-intent rephrasings. Tighter, not literal |
| “Set match types once and move on” | The mix should evolve with data: harvest converters into exact, expand scouting as tracking matures |
| “Negatives are optional housekeeping” | They're the control system. A campaign without growing negatives is a sieve |
Next StepsTighten Your Targeting Today
- Open the search terms report for the last 30 days.
- Add every irrelevant search pattern as a negative keyword.
- Promote any converting search terms into exact-match keywords.
- Check your match mix: launching? Phrase + exact. Data-rich? Scout with broad.
- Calendar the weekly terms-report review. Full optimization system: course waitlist.
FAQFrequently Asked Questions
What are the Google Ads match types?
Broad match (widest interpretation of your keyword), phrase match (the meaning must be included), and exact match (same intent plus close variants). They control which searches trigger your ads.
Which match type should beginners use?
Phrase and exact at launch — meaningful reach with intent attached. Introduce broad match later, once conversion tracking has data and a negative list guards against waste.
Why is my ad showing for irrelevant searches?
Match-type looseness — especially broad match. Open the search terms report to see the actual queries, then block the irrelevant patterns with negative keywords.
What are negative keywords?
Words and phrases you never want to trigger your ads (like 'free' or 'jobs' for a paid service). Each one permanently stops a category of wasted clicks.
Does exact match still allow variations?
Yes — plurals, misspellings, and same-intent rephrasings can trigger exact-match keywords. It's the tightest option but no longer perfectly literal.
How often should I check the search terms report?
Weekly in a campaign's first months — leaks are most expensive early. Mature, stable campaigns can stretch to fortnightly, but the ritual never fully retires.
Can negative keywords block my good traffic?
Over-aggressive ones can — negating whole words you're unsure about blocks every phrase containing them. Negative the patterns you're confident are wrong-intent; watch volume after additions.
What's the difference between negative phrase and negative exact?
Same logic as positive types, applied to blocking: negative phrase blocks any search containing the phrase; negative exact blocks only that specific search. Most beginner negatives work fine as phrase.
Do negative keywords block plurals and misspellings automatically?
No — negatives match more literally than regular keywords. Add the variants yourself (singular, plural, common misspellings) and check the search terms report for leaks around existing negatives.
What is a negative keyword list?
A reusable set of negatives applied across campaigns — typically universal junk like 'free', 'jobs', and 'diy'. Build it once at account level and every current and future campaign inherits the protection.
How do I use the search terms report effectively?
Classify every row into five buckets — proven winner, promising variant, wrong intent, wrong audience, researcher — each with a fixed action. A weekly 15-minute pass producing negatives and exact-match promotions is the whole discipline.
Your Budget Deserves a Bouncer at the Door
The upcoming Google Ads course covers the full targeting system — match strategy, the harvest loop, negative architecture. Join the waitlist, or bring your search terms report to the community.
No income promises. No hype. Just the path.