Broad Match Plus Smart Bidding: Does It Work In 2026?

Broad match has a reputation problem and most of it was earned. Anyone who ran it on manual CPC five years ago watched a month of budget vanish into searches that had nothing to do with the business, and they have not touched it since. I still meet advertisers who treat the unquoted keyword as a trap.

Here is the honest answer for 2026. Broad match paired with Smart Bidding works, and it regularly beats a tight phrase and exact list, but only when three things are already true: your conversion tracking counts things worth money, your negative keyword hygiene is a routine rather than a memory, and your budget is big enough for the algorithm to learn on. Miss one and broad match is just a way of paying Google to guess. Below: what changed in the matching engine, when broad plus target CPA or target ROAS wins, the guardrails to fit first, how to test it properly, and the September 2026 change that catches campaigns using the broad match setting.

What actually changed in the matching engine

The keyword stopped being the whole instruction.

Google’s documentation lists what broad match reads when deciding whether to match a query: the user’s recent search activity, the content of your landing pages and assets, and the other keywords in the ad group. That last one matters more than people realise. Your keyword list is now context, not just a list of triggers. A tight, thematically consistent ad group gives the matching engine a clean signal. A bin bag of unrelated terms gives it noise, and noise produces the search terms that make advertisers swear off broad match forever.

The second change is prioritisation, and it calms most nerves. If a search is identical to an eligible exact match keyword, that keyword is prioritised. Phrase and broad keywords identical to the term share equal priority, and where priority is level, Ad Rank decides. Broad match does not hijack the exact terms you spent two years building. It fishes in the water around them.

What has not changed: broad match runs on Google’s reading of intent, and Google is wrong sometimes. Policing the output is still your job.

When broad match plus Smart Bidding actually wins

Google publishes a figure for this. Advertisers who move phrase keywords to broad match can see around 25% more conversions in target CPA campaigns and around 12% more conversion value in target ROAS campaigns, while still meeting their targets.

Read the small print next to it, because that is where the lesson sits. Google also states that additional budget may be required, and the interface shows a “New budget” column telling you how much more to spend to reach the extra traffic. The honest translation: more conversions at your target cost, for more money. That is a scale lever, not a free efficiency win. If you are already capped at your target and cannot fund more spend, broad match is solving a problem you do not have.

Where I reach for it:

• Accounts with real conversion volume and clean tracking. Thin data means the algorithm learns from noise.
• Ecommerce on target ROAS with genuine value tracking. Values let the system sort winners from junk without you refereeing every query.
• Accounts out of keyword ideas. When the search terms report has stopped surprising you, broad match is the cheapest discovery tool available.

Where I leave it alone:

• Manual CPC. Google strongly recommends using broad match only with Smart Bidding. Broad match on manual bidding is the 2019 horror story, still available to anyone who wants to relive it.
• Under roughly 30 conversions a month. Google’s floor for target ROAS on Search and Shopping is 15 conversions in 30 days. Fifteen is permission, not readiness.
• Niche B2B with ambiguous terminology. If your product name is also a common word in another industry, broad match will find that industry with enthusiasm.

The guardrails to fit before you turn it on

Fit all four. Not three.

Guardrail	Why it matters	What good looks like
Tracking that counts money	Query selection moves to the bidding system, so its definition of success becomes your targeting	One primary conversion action per campaign goal, deduplicated, values passed where they vary
A working negative layer	Broad match finds the edges of your category, including the free, DIY and jobs edges	Account-level lists for jobs, free, cheap, DIY and competitors, plus a weekly campaign sweep
Enough budget to learn	A starved campaign samples badly and concludes badly	Daily budget that absorbs 10 to 15 clicks a day in the test campaign
Brand controls	Broad match otherwise wanders into brand terms and flatters your CPA with traffic you already own	Brand exclusions on the prospecting campaign, brand terms in a dedicated exact match campaign

Brand settings deserve a note. Brand exclusions apply across match types and stop your ads serving on the brands you list, including misspellings and variants. Brand inclusions do the reverse, restricting a campaign to searches related to the brands you specify, which is how you run a branded or conquesting campaign on broad match without drift. Where both are applied, exclusions take priority. Keep brand and generic separate, the same logic behind the two-campaign feeder structure we use.

How to test it without torching spend

Do not flip the account. Test it like a change you might have to reverse.

1. Pick one campaign with volume, not your worst performer. Aim for 30 or more conversions a month so the result is readable.
2. Isolate the test. Run a Search campaign experiment against the existing campaign, or build a separate campaign with its own budget. A separate campaign is easier to kill; an experiment gives the cleaner read.
3. Keep the bid strategy and target identical to the control. Change one thing. Loosen the target at the same time and you will never know which move produced the numbers.
4. Match the ad group theme tightly. The other keywords are a matching signal. Five closely related broad keywords beat thirty scattered ones.
5. Run it for four weeks or 30 conversions, whichever comes later, and judge it on incremental volume at target rather than CPA alone. CPA held and conversions rose, broad match won. CPA held and volume was flat, you spent a month proving nothing changed.

Through the test, read the search terms report twice a week. The first fortnight is when the ugly queries surface, and a negative added on day four saves three weeks of spend. We have written up a repeatable routine for that in mining the search terms report for wasted spend.

The broad match campaign setting, and the September 2026 catch

There is a campaign-level setting that makes a Search campaign use broad match exclusively. Switch it on and every phrase and exact keyword converts to broad, and new keywords default to broad. It is only available on campaigns using conversion-based Smart Bidding, which tells you plainly what Google thinks of broad match without it. Keywords in that campaign are treated as if they were both exact and broad, so an identical search still gets exact match prioritisation. The setting is reversible, though you reassign match types by hand afterwards.

Now the part to diarise. From September 2026, campaigns using that setting are being automatically upgraded to AI Max for Search, with search term matching enabled by default. This is not a rename. It changes how queries are matched and what assets can be generated, so a campaign you built as a controlled broad match test can become something else while you are not looking. If you are running the setting today, open the campaign this month, decide whether you want AI Max, and set brand controls and negatives accordingly rather than finding out from a search terms report in October.

What to check, and how often

Once broad match is live, the maintenance is light but non-negotiable.

• Weekly: search terms report, add negatives, promote any repeat converter to its own exact match keyword.
• Fortnightly: check the share of spend going to brand or near-brand queries. If it is climbing, your brand exclusions need work.
• Monthly: compare broad against the control on conversions and cost per conversion, never on impressions. Impression growth from broad match is guaranteed and meaningless.

Broad match in 2026 is a scaling tool with a tracking prerequisite. Treat it that way and it earns its place. Treat it as a shortcut around keyword research and it will find you every tyre-kicker in the country.

Want a second pair of eyes before you switch it on

Most of the broad match disasters I get called into were not caused by broad match. They were caused by conversion tracking that counted the wrong thing, and broad match made the problem expensive faster. If you are not certain what your account is optimising to, get that checked first. Our free Google Ads audit covers tracking, structure, wasted spend and bid strategy, and you get a straight answer on whether your account is ready for broad match. If you would rather hand it over, that is what our Google Ads management and wider paid search work is for.

FAQ

Is broad match better than phrase match in 2026? Not universally. Broad match wins where you have clean conversion tracking, Smart Bidding, budget headroom and a need for more volume. Phrase and exact still win where budget is tight, conversion volume is low, or your category shares vocabulary with an unrelated industry. Test, do not assume.

Do I need Smart Bidding to use broad match? Technically no, practically yes. Google strongly recommends using broad match only with Smart Bidding, and the campaign-level broad match setting is only available on campaigns running conversion-based Smart Bidding. Broad match on manual CPC is how broad match got its bad name.

How many conversions do I need before trying broad match? Google’s stated minimum for target ROAS on Search and Shopping is 15 conversions in 30 days. My floor is roughly 30 a month before adding broad match on top, because below that a couple of unusual orders swing the bidding and you misread the test. Our guide to Google Ads bidding strategies covers what each strategy needs.

Will broad match steal traffic from my exact match keywords? Not for identical searches. If a query is identical to an eligible exact match keyword, that keyword is prioritised, and where priority is equal Ad Rank decides. The real risk is broad match picking up brand queries you would have won for pennies, which is what brand exclusions are for.

What happens to my broad match campaigns in September 2026? Campaigns using the campaign-level broad match setting are being auto-upgraded to AI Max for Search from September 2026, with search term matching on by default. Review them now, confirm negatives and brand controls are in place, and decide whether you want AI Max before the upgrade decides for you.