Part 4 of 6
The Google Ads Bidding Ladder + Weekly Keyword Protocol
Direct answer
Bidding strategy should follow conversion volume, not calendar time: start on manual CPC or Max Clicks, move to Max Conversions at 15+ conversions, and graduate to target CPA/ROAS at 30+ conversions in a 30-day window. Keyword research is not a one-time task: a 15-20 minute weekly review of the search terms report is what compounds a 3x ROAS account into a 4.4x account over a few months, purely by eliminating waste.
This is part 4 of a 6-part series. Part 3 covered the campaign layers; this part covers how you bid and how you keep refining what you're bidding on.
The bidding graduation ladder
- 0-15 conversions: manual CPC or Max Clicks. The algorithm has no reliable pattern to learn from yet: give it structure, not freedom.
- 15-30 conversions: Max Conversions, without a target. Let the algorithm start optimizing, but don't constrain it with a CPA or ROAS target it doesn't have the data to hit reliably.
- 30+ conversions in a 30-day window: graduate to target CPA (search) or target ROAS (shopping/demand gen). The algorithm now has enough signal to optimize toward a specific goal rather than just volume.
The trigger is always conversion volume, never time. A campaign running 90 days with only 20 conversions should still be on Max Conversions. A campaign that hits 30 conversions in 15 days is ready to graduate immediately: don't wait out an arbitrary calendar window.
When you switch strategies: set your initial target above current actual performance, not at your ambition. If you're running 3.8x ROAS on Max Conversions, set your tROAS target at 3x — not 4.5x. Tighten in 10-15% increments over 2-week periods as performance stabilizes. Setting an aggressive target on day one starves the campaign before it has a chance to prove the new strategy out.
The one permanent exception: branded search stays on manual CPC, always, regardless of data volume: see Part 2 for why.
2026 update: what changed with broad match
The original ladder logic still holds, but Google made broad match the default keyword match type in mid-2024, and the platform is explicitly betting on modeled intent over rigid keyword matching. That changes the readiness bar for broad match specifically: it's no longer just "have I graduated to tCPA/tROAS," it's a signal-quality checklist:
- ~30+ conversions per month, sustained (not just a one-time spike)
- Enhanced Conversions implemented, not just standard tag tracking
- Automated bidding (tCPA/tROAS) already active and stable
- A working negative keyword framework already in place (see below) before you widen match types
Adopt broad match once those conditions exist, layered on top of proven exact/phrase performers: not as a starting point, and not just because Google defaults new campaigns to it.
The weekly keyword protocol
Keyword research isn't a setup task you finish once. It's a weekly intelligence operation, and it's one of the highest-ROI 20 minutes in the entire account.
The search terms report is your primary source of truth. It shows every query that actually triggered your ads, with full performance data: what's converting, what's wasting money, and what new themes are emerging that deserve their own campaign.
Weekly protocol (15-20 minutes, every Monday):
- Pull the search terms report for the last 7 days (or last 3 days if you're scaling fast and waiting a full week is too slow).
- Review the export for three things: converting queries that are not yet in your keyword list, high-spend zero-conversion queries that should be negatives, and new keyword themes that deserve their own campaign or ad group.
- Add converting queries as exact match to the relevant ad group.
- Add waste queries as negatives, with the appropriate match type.
- Flag any new query clusters for campaign expansion.
The compound effect is the whole point. Month 1, with zero negatives, at $10k spend, might generate $30k revenue (3x). Month 3, with 400+ accumulated negatives, might generate $44k at the same spend (4.4x). Budget didn't change. The waste got eliminated, one week at a time.
Real example
A supplement client was bleeding $12k/month on terms like "magnesium for fish tanks," "free magnesium samples," and "how to make magnesium at home" — $144k a year, paid to Google for zero-conversion traffic. One audit. One negative keyword list. $144k/year recovered.
FAQ
When should I switch from Max Conversions to target ROAS?
At 30+ conversions within a 30-day window: never based on how long the campaign has been running. A campaign with heavy spend but thin conversion volume should stay on Max Conversions until it clears that threshold.
Is broad match safe to use in 2026?
Yes, once specific conditions are met: sustained 30+ monthly conversions, Enhanced Conversions implemented, automated bidding already active, and an established negative keyword framework. Google made broad match the default for new campaigns in 2024, but "default" doesn't mean "ready": signal quality still has to come first.
How often should I review the search terms report?
Weekly, at minimum: 15-20 minutes is enough if you have a consistent process. Waiting longer lets waste accumulate and delays catching new high-intent query themes before competitors do.
What should my initial target CPA or ROAS be when I graduate a campaign?
Set it at or slightly better than current actual performance, not your goal. If a campaign is running 3.8x ROAS on Max Conversions, start the tROAS target around 3x and tighten by 10-15% every two weeks as performance holds.