Broad, Phrase Or Exact? The Google Ads Setting Most Owners Get Wrong

There are dozens of guides explaining Google Ads match types. Almost all of them stop at the definitions.

You finish reading, you can now describe all three, and you still don’t know which one to use on Monday morning.

That’s the gap this post is for. Not what the match types are. Which one belongs in your account, given your budget, this month.

We’ll use the largest public dataset on the question to settle the argument, then give you a rule you can apply in ninety seconds.

There Are Three. There Used To Be Four.

Start here, because a surprising number of guides still ranking on page one get this wrong.

Google Ads has three keyword match types: broad, phrase and exact.

Until 2021, there was a fourth called broad match modifier. You put a plus sign in front of the words that had to appear, and Google had to honour it.

Google retired it and folded its behaviour into phrase match. If a guide tells you there are four match types, it’s teaching a system that stopped existing five years ago.

This matters practically, not just pedantically. When broad match modifier went, everything loosened by a notch. Phrase match took on some of its behaviour and broad match got wider still.

So if you built your campaigns before 2021 and haven’t rewritten the keywords, your account isn’t doing what you told it to. The words are identical. The rules underneath moved.

The Three, In Plain English

Forget the jargon for a minute. The only real difference is how much control you keep.

Exact match is written in square brackets, like [emergency plumber]. Your ad shows on searches with the same meaning as your keyword. You keep the most control.
Phrase match is written in quotation marks, like “emergency plumber”. Your ad shows on searches that contain the meaning of your keyword, with extra words allowed before or after. Middle ground.
Broad match is written as plain text, like emergency plumber. Your ad shows on anything Google judges to be related. You keep the least control.

Notice what’s happening across those three. You aren’t choosing how many people see your ad. You’re choosing how much of the decision you hand to Google.

That’s the actual trade, and it’s why the right answer depends entirely on how much data Google has to make that decision well.

A Worked Example Of The Same Keyword

Take one keyword and follow it through all three settings. Say you’re a plumber bidding on emergency plumber in Luton.

On exact match, [emergency plumber luton], you’d show for “emergency plumber luton”, “emergency plumbers luton”, “luton emergency plumber”. Same meaning, different arrangements.

On phrase match, “emergency plumber luton”, you’d add “24 hour emergency plumber luton”, “cheap emergency plumber luton near me”, “emergency plumber luton bedfordshire”. Your phrase is in there, with words around it.

On broad match, emergency plumber luton, you’d add “boiler leaking who to call”, “plumber near me open now”, “how to stop a burst pipe”, “plumbing jobs luton”.

Look at that last list carefully. The first two are excellent. You’d have paid for them happily and might never have thought to write them down.

The third is somebody fixing it themselves. The fourth is looking for work.

That’s the whole thing in one example. Broad match finds you customers you couldn’t have guessed, and it charges you for people who were never going to call.

What The Data Actually Says

This is the part the definition guides leave out, and it’s the part that should decide your settings.

Optmyzr ran the largest public study on this question, analysing 992,028 keywords across 15,491 Google Ads accounts.

The headline results are one-sided.

Exact match delivered the highest return on ad spend at 415 per cent, and the highest click-through rate at 21.6 per cent.

In roughly 70 per cent of accounts, exact match produced a lower cost per acquisition than broad match.

In roughly **72 per cent of accounts**, exact match produced better return on ad spend.

There’s one finding that goes the other way, and it’s the one Google’s guidance leans on. In 44 per cent of accounts, broad match did produce a lower cost per click.

Read those two results together, because that’s where the lesson is. Broad match often buys cheaper clicks. Those cheaper clicks convert at a lower rate. Cheaper traffic, more of it, less of it useful.

An earlier Optmyzr study found the same shape from a different angle: in 85.6 per cent of accounts, click-through rate was higher on exact than on broad. People clicked the exact match ads more because they were more relevant.

One more result worth knowing, because it’s the one nobody expects. In the ecommerce segment, phrase match posted the highest conversion rate of any match type.

Phrase match catches shoppers using slightly different product language while they’re still in buying mode. It’s the most underrated setting in Google Ads.

Why Cheap Clicks Are The Expensive Ones

If you take one idea from this post, make it this one.

Cost per click is not a measure of value. It’s a measure of competition.

A click costs less when fewer advertisers want it. Fewer advertisers want it because it converts less often. That’s not a coincidence; it’s the auction working correctly.

So when broad match delivers a cheaper click, it’s usually telling you something true: this search is worth less. The market has priced it.

The number that matters is cost per enquiry. Cost per customer, if you can track that far.

Run the sums on a real example. Say broad match buys clicks at £2 and converts at 2 per cent. That’s £100 per enquiry.

Exact match buys clicks at £4 and converts at 6 per cent. That’s £67 per enquiry.

Exact match costs twice as much per click and a third less per customer. If you judged those two on cost per click you’d pick the worse one, confidently, every time.

The Decision Rule

Here’s the ninety-second version. Find your line and act on it.

Under 30 conversions a month. Phrase and exact only. Don’t touch broad match yet.

Smart Bidding needs roughly thirty conversions a month to learn properly. Below that, broad match hands the decision to a system that doesn’t have enough information to make it. You’re paying for the algorithm’s education.

Between 30 and 100 conversions a month. Phrase and exact as your core. Test broad match in a separate campaign with its own capped budget, ten to twenty per cent of the total.

Now broad match spend is visible and contained. You can measure exactly what it costs and what it returns without it eating the budget funding your proven keywords.

Over 100 conversions a month. Broad match with Smart Bidding becomes a genuine growth lever. You have the data to steer it and the volume to absorb its mistakes.

Even here, keep your exact match keywords running. They’re your floor.

One override, whatever your volume. If your niche is genuinely small and your exact match keywords are starved of impressions, broad match is how you find demand you couldn’t guess the wording of. Cage it, but use it.

The Structure That Runs All Three

Match type isn’t really one choice. Most good accounts run all three at once, doing different jobs.

Think of it as three tiers.

Tier one, exact match. Your proven money keywords, the searches you know convert. Highest bids, tightest control, most of the budget. This is the engine.

Tier two, phrase match. The same themes with room to breathe. Catches the variations and the longer, more specific searches. Middle bids. This is where a lot of quiet profit sits.

Tier three, broad match. Its own campaign, capped budget, treated as research. Its job isn’t to deliver your enquiries. Its job is to find search terms you didn’t know existed.

Then the loop that makes it work. Every week, read the search terms report from tier three. Anything converting gets promoted to phrase or exact. Anything wasting gets added as a negative.

Broad match becomes a keyword discovery machine that feeds your tight campaigns. That’s what it’s genuinely good at.

Without that loop, tier three is just a hole in the budget.

What The Three Tiers Look Like In A Real Account

The tiers sound abstract until you see them laid out. Here’s a plumbing business on £900 a month, structured properly.

Campaign one, Emergency, exact match, £450 a month. Keywords like [emergency plumber luton], [24 hour plumber luton], [burst pipe repair luton]. Highest intent, highest bids, half the budget. These people are ringing someone in the next twenty minutes.

Campaign two, Services, phrase match, £315 a month. Keywords like “boiler repair luton”, “bathroom plumber luton”, “leaking tap repair”. Real intent, longer decision, more variation in how people phrase it.

Campaign three, Discovery, broad match, £135 a month. Two or three broad keywords, hard budget cap, reviewed weekly. Its only job is to surface search terms nobody wrote down.

Fifteen per cent of budget in tier three is the number we’d defend. Enough to learn something. Not enough to hurt.

Now the maintenance, which is the part that actually creates the value.

Each Monday, open the discovery campaign’s search terms report. Anything that converted gets added to campaign one or two as an exact or phrase keyword. Anything obviously wasteful gets added to the shared negative list.

Over three months, that discovery campaign quietly rewrites your keyword list using your customers’ own vocabulary rather than your guesses about it.

Compare that to the default setup, which is one campaign, four broad keywords, no negatives and no review. Same budget, completely different business.

How Match Types Change What You Pay Per Click

There’s a second-order effect here that most guides skip, and it compounds quietly in your favour.

Google gives every keyword a Quality Score out of ten, built largely on expected click-through rate, ad relevance and landing page experience. Higher Quality Score means you pay less for the same ad position.

Match type feeds directly into that.

A tight exact match keyword lets you write an ad that speaks precisely to that search. The searcher sees their own words in your headline, so they click more often. Click-through rate rises, Quality Score rises, and your cost per click falls.

A broad match keyword has to serve hundreds of different searches with one set of ads. Those ads can only ever be generically relevant. Click-through rate suffers, Quality Score suffers, and you pay a premium for the same position.

This is why the Optmyzr click-through figures matter beyond the headline. Exact match hitting a 21.6 per cent click-through rate isn’t just a nicer number. It’s cheaper clicks arriving through the back door.

So the account that tightens its match types often sees cost per click rise at first, because it’s competing for better searches, then drift back down over a couple of months as Quality Score catches up.

That second movement is the one nobody warns you about, and it’s the reason six weeks is the minimum honest review window.

Match Types And Negative Keywords Are One Decision

Treat these as separate jobs, and you’ll do both badly. They’re two halves of the same control.

Match types decide how wide the net is. Negative keywords decide what falls through it.

Loosen the match type without building the negative list, and you’ve widened the net with no filter. That’s the default account, and it’s why 2357 in this series exists.

Tighten the match type, and you still need negatives, just fewer of them. Close variants guarantee some drift even on exact match.

The practical rule is simple. Every time you loosen a match type, add negatives first. Not afterwards, when you’ve already paid for the lesson.

Build one shared negative keyword list at account level and apply it to every search campaign. Start with the universal blockers: free, cheap, DIY, how to, jobs, salary, course, training, wholesale, second hand. Then add your trade’s own traps.

Block patterns rather than individual searches. If you can imagine five versions of the same bad search, find the word they share and block that as a phrase negative. One entry, permanent effect.

Where Match Types Don’t Apply At All

Worth knowing, because it explains a gap that confuses a lot of owners looking at their accounts.

Match types only exist in Search campaigns. That’s it.

Performance Max has no keywords and no match types. You give Google assets, an audience signal and a goal, and it decides where and when to show. You can add negative keywords at account level, and you can exclude search themes, but you cannot bid on a keyword in a chosen match type.

Demand Gen, Display and Video work the same way. No keywords, no match types.

Shopping campaigns have no keywords either. Google matches your products to searches using your product feed, which is why feed quality does the job keywords do in Search.

So if most of your budget sits in Performance Max, this post changes less for you than you’d hope. The equivalent lever there is your negative keyword list, your search theme exclusions and the quality of your conversion data.

It’s also a reason to keep a proper Search campaign running alongside Performance Max. Search is where you still get to make the decisions.

Close Variants, The Thing That Breaks All Three

Here’s the mechanism almost no beginner guide mentions, and it explains a question we get constantly.

Why is my exact match keyword showing for a search that isn’t exact?

Because exact match hasn’t meant exact since 2018, it means the same thing.

Close variants cover plurals, misspellings, abbreviations, reordered words, added function words, and now synonyms and same-meaning phrasings.

So [emergency plumber] can serve “emergency plumbers”, “plumber emergency”, “urgent plumber”, and sometimes things you’d argue with.

This applies to all three match types. Every one of them is looser than its name suggests, and they’ve all been getting looser.

Two practical consequences.

First, you cannot set match types once and assume they hold. The behaviour drifts underneath you as Google’s language models change. Review quarterly at minimum.

Second, negative keywords matter even on exact match. People assume exact match doesn’t need protecting. It does, just less.

How To Change Match Type Without Losing Your Data

Small but important, because doing this wrong throws away months of performance history.

You can’t edit a keyword’s match type. Google treats each match type as a different keyword.

So don’t delete and replace. Add the new one alongside the old.

Add [emergency plumber] while “emergency plumber” is still running. Let both run for a fortnight. Then pause the loser one.

Why it matters: your keyword’s performance history feeds Quality Score and Smart Bidding. Delete the keyword, and you delete the history. Add alongside, and the account keeps what it learned.

One caution. Running the same term on two match types means they compete for the same searches. Google generally serves the more specific one, but check your search terms report to confirm which is actually picking up the traffic before you pause anything.

What To Expect In The Six Weeks After

Set expectations properly, because tightening an account looks like a mistake before it looks like a fix.

Weeks one and two. Impressions fall, sometimes sharply. This is correct. You’ve stopped showing ads to people who were never buying.

Weeks two and three. Cost per click rises. Also correct, and it catches people out badly. You’re competing for searches other real businesses want.

Weeks three and four. Click volume settles at a lower level. Enquiry volume holds roughly flat.

Weeks five and six. Cost per enquiry starts falling. This is the payoff.

Watch traffic, and this whole sequence reads as failure. Watch cost per enquiry, and it reads as exactly what you paid for.

Give any match type change six weeks before you judge it. Two weeks tells you nothing except that the numbers moved.

The Mistakes We See Most

Drawn from accounts we’ve taken over this year.

Judging match types on cost per click. The single most expensive habit in paid search.
Using only exact match and wondering why volume is thin. Exact match is a floor, not a whole strategy.
Deleting a keyword to change its match type, and binning the history with it.
Assuming exact match needs no negative keywords. Close variants say otherwise.
Running broad match in the same campaign as proven keywords, where it quietly consumes their budget.
Reading a guide that describes four match types and hunting for a plus sign that hasn’t existed since 2021.
Changing match types and judging the result after ten days.

Your Twenty Minute Job This Week

One sitting, no tools.

Minutes one to five. Open your keywords tab. Count how many are plain text, how many are in quotes, how many are in brackets. Most owners are surprised.

Minutes six to ten. Find your monthly conversion count. Apply the decision rule above and write down which tier you’re in.

Minutes eleven to fifteen. Take your five best-performing keywords. Add each one in exact match alongside its current version. Don’t delete anything.

Minutes sixteen to twenty. Check whether broad match is sharing a campaign with your proven keywords. If it is, that’s your next job: move it into its own campaign with a capped budget.

Diarise a review for six weeks. Judge it on cost per enquiry.

The Scorecard: Are Your Match Types Working?

Six weeks later you’ll want to know whether the change worked. Here’s how to tell, because the obvious metrics will lie to you.

Score your account on these five, using a 30-day window against the 30 days before you changed anything.

Irrelevant spend. Run the search terms report, sort by cost, and total the spend on searches you’d never have chosen. Under 10 per cent is healthy. Over 25 per cent means the net is still too wide.

Cost per enquiry. The only number that matters. It should be flat or down. If it’s up after six weeks, you’ve tightened too far and starved the account.

Enquiry volume. Should be flat or up. Down sharply means your exact match keywords don’t have the search volume to carry the account alone, and you need more phrase match.

Impression share lost to budget. If this has jumped, your tighter keywords are now competing well, and you’re running out of money before the day ends. That’s a good problem. It means the next sensible move is more budget, not more keywords.

Search terms per keyword. Pick one keyword and count how many distinct searches it served. An exact match keyword serving forty different searches is telling you close variants have drifted a long way.

Two of those five moving the right way is a decent result. Four is a good quarter.

One warning about the trap most owners fall into here. Clicks and impressions will both be down, and they’ll be down in a way that looks alarming on the graph.

That’s the change working, not failing. You deliberately stopped buying traffic that never became revenue. The graph can’t tell the difference between traffic you lost and traffic you refused.

Your bank account can.

Frequently Asked Questions

Which Google Ads match type should I use?

Under 30 conversions a month, run phrase and exact match only. Between 30 and 100, keep phrase and exact as your core and test broad in a separate capped campaign. Over 100, broad match with Smart Bidding becomes a reasonable growth lever. The rule tracks conversion volume because that’s what Smart Bidding needs to steer broad match well.

What is the difference between phrase match and exact match?

Phrase match shows your ad on searches containing the meaning of your keyword, with extra words allowed before or after it. Exact match shows it only on searches with the same meaning as your keyword. Phrase buys you reach, exact buys you control. Most good accounts run both.

Does exact match perform better than broad match?

In most accounts it does. Optmyzr analysed 992,028 keywords across 15,491 accounts and found exact match delivered a lower cost per acquisition than broad in roughly 70 per cent of accounts, and better return on ad spend in roughly 72 per cent. Broad match won on cost per click in 44 per cent, which is the trap.

What happened to broad match modifier?

Google retired it in 2021 and folded its behaviour into phrase match. Any guide describing four match types is out of date. If your account predates 2021 and the keywords haven’t been rewritten, they aren’t behaving the way you originally set them.

Why is my exact match keyword showing for different searches?

Close variants. Since 2018, exact match has meant the same meaning rather than the same words, and it now covers plurals, misspellings, synonyms, reordered words and equivalent phrasings. Some drift is normal. Only your search terms report will tell you how far it has gone in your account.

How do I change match type in Google Ads?

You can’t edit a keyword’s match type directly, because Google treats each match type as a separate keyword. Add the keyword again in the new match type, let both run for a fortnight, then pause the old one. Adding rather than replacing preserves the performance history that feeds Quality Score.

Is phrase match underrated?

We think so, and the data supports it. In the ecommerce segment, phrase match posted the highest conversion rate of any match type. It catches buyers using slightly different product language while they’re still in buying mode, which is a large and profitable group most accounts under-serve.

Do match types still matter now that Google uses AI?

They matter more, not less. Google’s language models have loosened every match type, so the gap between what you typed and what you serve on is wider than it was. Match type is now one of the few levers you still control.

Pick The Tier, Not The Setting

The question isn’t broad, phrase or exact. Framed that way, it has no answer.

The question is how much of the decision you can afford to hand to Google right now, and that comes down to how much conversion data you’re giving it to decide with.

Thin data, keep control. Rich data, share it. That’s the whole framework.

So find your conversion count, pick your tier, and give it six weeks before you judge anything.

And whatever you do, stop judging match types on cost per click. The cheapest clicks in your account are almost always the most expensive customers you’ll never get.

If you’d like someone to look at your keyword list and tell you which tier you’re actually in, that’s what we do at SF Digital.

Did You Enjoy This Blog Post?

I hope you enjoyed this blog post, and thank you so much for being here. We also upload videos to our YouTube channel every weekday. Please subscribe so you are one of the first to be notified.

If you enjoyed this blog, you may also like:

SF Digital Studios

Search This Blog