Broad match lets your ad show for any search related to the meaning of your keyword; exact match restricts it to searches with the same intent. In 2026 the right choice is neither one alone: accounts with enough conversion volume should discover with broad match plus Smart Bidding, then promote proven terms into phrase and exact. This guide covers the thresholds and a sample budget split for building that system.
What are keyword match types and how do they work?
Google Ads has three match types: broad match, phrase match and exact match. The match type controls which search queries your keyword can trigger. In 2026 matching is intent-based, not string-based: even exact match targets the same search intent rather than the literal spelling.
- Broad match: 'running shoes' can show for 'best footwear for a marathon'. Widest reach, lowest control.
- Phrase match: the query must include the meaning of your keyword; 'running shoes price' matches, 'casual sneakers' mostly does not.
- Exact match: only queries with the same intent; 'running shoes' and 'shoes for running' match. Narrowest reach, highest control.
When is broad match the right choice in 2026?
Broad match works well when three conditions hold at once: Smart Bidding (tCPA or tROAS) is active, the campaign generates roughly 30-50 conversions per month, and a negative keyword list exists from day one. Without those, broad match spends on irrelevant searches due to weak signals. If your bidding setup is not settled yet, see our tCPA vs tROAS guide.
How do you build a tiered match type strategy?
In a tiered system each match type has a job: broad match discovers new queries, phrase match validates proven terms, exact match locks in your most profitable ones. It works like a one-way conveyor belt, fed by a weekly search terms review.
- Discover: collect new queries in a low-budget campaign running broad match with Smart Bidding.
- Validate: after 2-3 weeks, add converting queries to your main campaign as phrase match.
- Lock in: move terms converting below target cost into exact match, where bid control is highest.
- Clean up: add non-converting queries to the negative list every week so discovery does not accumulate waste.
Want the search terms report read for you?
Ads Sensor analyzes your campaigns with AI, catches wasteful queries and delivers negative keyword suggestions with reasoning attached.
How should budget be split across match types?
There is no fixed formula, but a healthy starting point is ~50% broad for discovery, ~30% phrase for validation and ~20% exact for locked winners. As the account matures, the exact share grows; when entering a new market, the discovery share grows. Keep brand searches out of this split in a separate brand campaign.
Do match types still matter in the AI Max era?
Yes. With keywordless features like AI Max for Search, Google expands campaigns to new queries based on your landing pages and ads; according to platform data this can add conversions at similar cost. But AI Max builds on top of your existing keywords and negatives: automation amplifies mistakes in an undisciplined account. For choosing the keywords themselves, see AI keyword research.
In short: do not enable broad match before the conditions are met, reserve exact match for winners, and keep the system clean with weekly negative feeding. If running this loop by hand takes too much time, join the Ads Sensor beta: it surfaces search term waste, match-driven drift and bid opportunities through AI analysis.