Changing what the bidder optimises restarts learning. Changing how much you spend usually does not — but a large budget jump disrupts delivery in a way that looks identical to a learning reset, which is why the myth persists.

On this page

• What actually resets the Google Ads learning phase?
• What disrupts delivery without resetting learning?
• How big a budget change is safe in one step?
• When can I change a budget in one move instead of steps?
• Should I pause a campaign or turn the budget down?
• Why should I never stack two changes in one window?

What actually resets the Google Ads learning phase?

• Changing the conversion actions the campaign bids on.
• Changing the bid strategy type.
• Adding or removing biddable conversion actions, including inert micro-events.

That last one is worth knowing before you tidy up: removing biddable actions re-enters learning, so deleting inert placeholder conversions costs a reset in exchange for cleaner reports. Usually not worth it.

What disrupts delivery without resetting learning?

A large budget increase pushes real incremental spend into less efficient inventory faster than the algorithm can re-find efficient placements. Cost per acquisition spikes. It reads as a learning reset. It is not one — but the practical effect on your week is the same.

How big a budget change is safe in one step?

Move budgets on Smart Bidding campaigns in roughly 15% steps, gated on cost per acquisition at each rung. No pre-set endpoint. Step up while the cost holds, stop the moment it degrades, let the data pick the ceiling.

A proposed $25 → $50 increase on a software affiliate account was flagged as too aggressive, and that was the right call — those campaigns were already spending to cap, so the increase was true new delivery, not unused headroom.

When can I change a budget in one move instead of steps?

The 15% cap does not apply to:

Should I pause a campaign or turn the budget down?

Pausing and resuming does not wipe a campaign's history. Long pauses hurt because the auction, your competitors and seasonality all moved while you were away — the campaign is re-entering a different market, not a blank one.

For a slow season, turn the budget down rather than pausing, in steps, and keep the highest- intent campaigns alive. That preserves signal continuity and avoids re-entering cold.

Why should I never stack two changes in one window?

Never stack changes inside a gate window. If you change the budget and the target and the creative in the same week, you have no read on any of them. One change, one read date, written down before you start.

Related questions

My Google Ads campaign says "limited by budget" but it's already beating its target — should I raise the budget?

Yes, and it is the clearest scaling signal in the platform — but raise it in roughly 15% steps gated on cost per acquisition, not in a jump. Read the full answer.

How do I stop Performance Max from spending my budget on YouTube?

You cannot switch YouTube off inside a Performance Max campaign. What you can do is see exactly how much it is costing you, fix the signal that is sending it there, and decide whether to split the campaign. Read the full answer.

How do I work out what target ROAS to actually set for my store?

Derive it from contribution margin and your fixed per-order costs, not from a benchmark. The formula everyone uses — 1 ÷ gross margin — is a breakeven floor, not a target, and it ignores the per-order costs that decide whether an order is actually worth having. Read the full answer.

Written by Usama Khan, who runs Google Ads for e-commerce brands and home-services companies at AdventurAds. Figures are from accounts under active management and are anonymised. Last reviewed 2026-08-26.