Broad vs Phrase vs Exact Match: Which Wastes Budget

Match type is the single setting most likely to quietly drain a small Google Ads budget. It decides how far Google is allowed to wander from the words you actually typed into the keyword box.

The bad news: all three match types are looser than their names suggest, and Google's public advice — put everything on broad match and let Smart Bidding sort it out — is written for advertisers with far more conversion data than a plumber spending $3,000 a month has.

This post gives you the volume threshold, the escalation ladder, and the search-terms audit that keeps broad match from eating your month.

What do broad, phrase and exact match actually mean in 2026?

Google defines three match types by how much meaning they are allowed to stretch. Per Google's keyword matching options documentation: broad match ads 'may show on searches that are related to your keyword, which can include searches that don't contain the direct meaning of your keywords'; phrase match ads 'may show on searches that include the meaning of your keyword'; exact match ads 'may show on searches that have the same meaning or same intent as the keyword.'

Read those three sentences again. Not one of them says 'the words you typed.' All three are meaning-based. The difference is only how wide the meaning net is thrown.

Verdict for a service business under $5,000/month: exact and phrase win, broad loses. Not because broad match is broken, but because broad match is a bet on an algorithm that needs conversion data you do not have yet.

Why is exact match no longer exact?

Because Google matches exact-match keywords to close variants, and per Google's close variants documentation, 'By default, all keyword match types are eligible to match to close variants. There's no way to opt out.'

For exact match, Google lists close variants that include misspellings, singular and plural forms, reordered words, added or removed function words, implied words, 'synonyms and paraphrases,' and — the big one — 'same search intent.' Google's own example: the exact keyword [images royalty free] can show ads on searches for 'free copyright images.'

Translate that to a trade. Your exact keyword [emergency plumber] is eligible for anything Google's models judge to carry the same intent. That is a judgment call made by a machine, on your money, and you cannot switch it off. Exact match in 2026 means tightest, not literal.

This is why a keyword-level bid cut is a weak lever and a well-built negative keyword list is a strong one. Negatives are the only place you still get literal control.

At what conversion volume does broad match start working?

Use 30 conversions in the last 30 days, per campaign, as the line. That number is not ours — it is Google's own evaluation guidance. Google's Target CPA documentation says: 'For evaluation, we recommend you measure performance for the last 30 days, including at least 30 conversions.'

That sentence is the whole argument. Broad match only behaves when Smart Bidding can price each auction well. If you cannot even evaluate the bid strategy at your conversion volume, you certainly cannot trust it to spend your budget on searches you never chose.

So the threshold is simple. Below 30 conversions a month in a campaign, broad match is malpractice for a small local advertiser: you are paying for the model's training data. Above it, broad becomes a controlled experiment instead of a donation.

Count real conversions, not clicks and not form views. If your conversion action is 'phone call over 60 seconds' and you get 14 of them a month, you have 14 — not 40. Getting that number honest is the first job in any bid strategy automation decision.

Why does Google recommend broad match to accounts that can't support it?

Because Google's documentation is written for every advertiser at once, and because more matched searches means more auctions. Google states in the same Target CPA doc that 'Advertisers can start using Target CPA with no conversion history, and Target CPA is effective for campaigns of all sizes' — while, two sections later, telling you to evaluate it over at least 30 conversions.

Both statements can be true. You can start. You just cannot tell whether it worked. That gap is where a $3,000 monthly budget disappears.

It gets worse for a small account when you read what broad match uses to decide. Per Google's matching documentation, broad match 'may also take into account' the user's recent search activities, the content of your landing pages and assets, and other keywords in the ad group.

• Your landing page copy is now a targeting input — a page that mentions 'commercial HVAC' can pull commercial searches into a residential campaign
• Other keywords in the same ad group shift what broad match thinks your keyword means, so a sloppy ad group widens every keyword in it
• The searcher's recent activity is a signal you cannot see, cannot audit, and cannot negative out
• Google explicitly says 'It's critical to use Smart Bidding with broad match' — which is an admission that broad match without a trained bidding model is uncontrolled

None of that is a conspiracy. It is an incentive. Read the docs as a vendor's product guidance, not as advice tailored to your P&L, and the recommendation makes sense — for someone else's account.

What is the escalation path from exact to phrase to broad?

Four stages, each with a numeric trigger to move up. Do not skip a stage because a rep told you to consolidate.

Stage 4 is a test with a kill switch, not a migration. If the broad ad group's cost per booked job is worse than phrase after 30 days, you pause it. That is the same 90-day logic we apply to a whole channel: no qualified leads, no budget.

If you are still deciding what the budget should be in the first place, start with how much Google Ads budget a service business needs before you touch match types. Match type strategy on a budget too small to generate 15 conversions is rearranging deck chairs.

How do you audit the search terms report for wasted spend?

Weekly, at 15 minutes, with three columns turned on: Search term, Match type, and Keyword. Per Google's search terms report documentation, the Keyword column does not show by default — you add it under Columns, then Attributes.

The audit itself is mechanical. Sort by cost descending. Read the top 50 terms. For each one, ask a single question: would you answer that phone call? If the answer is no, it becomes a negative. If the answer is yes and it is converting, promote it to its own exact-match keyword.

• Sort by cost, not by clicks — the expensive junk term is the one bleeding you
• Segment by match type: if broad match terms carry most of the spend and none of the conversions, you have your answer
• Look for the three classic service-business leaks — job seekers ('plumber jobs'), DIY ('how to fix a leaking tap'), and competitor brand names you never meant to bid on
• Promote converting search terms to exact match in their own ad group, so you control the bid instead of the algorithm
• Add negatives at the right level — campaign for structural exclusions, shared list for the DIY and jobs leaks across every campaign

One catch nobody mentions: the report is incomplete by design. 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.' You are shown the searches that hit a volume threshold — not all of them. On a broad-match campaign in a small account, a real share of your spend is invisible.

That is the strongest practical argument against broad match at low budget. You cannot negative out what Google will not show you.

What does broad match cost a $3,000/month local advertiser?

Run the arithmetic on your own numbers and the answer is usually 30–50% of the budget, spent before you can measure it. Here is the worked example with the assumptions stated out loud, because anyone quoting you a hard 'broad match wastes X%' figure is inventing it.

Broad match buys you more clicks and fewer leads. It looks better on the impressions and clicks columns your agency puts in the monthly PDF, and worse on the only column that pays your staff.

Swap in your own CPC and your own conversion rate — the shape does not change. The cheaper the click, the wider the intent, and intent is what you are actually buying. That is the entire reason we report booked calls rather than click volume in our paid ads work.

And if a campaign at 40 conversions a month makes broad match viable? Test it. The threshold is a threshold, not a religion. Just cap it, measure it against phrase on cost per booked job, and kill it if it loses.

If you want a second pair of eyes on a search terms report before you sign another quarter with whoever built the account, we will read it for you and tell you what we would cut. Get my free audit.