Manage Bids

This playbook helps you change bids deliberately enough that you can still learn from the result.

Fast Path

Decide whether the bid move is defensive or scaling-oriented, change only as far as the evidence supports, and wait for enough post-change data before touching it again.

How To Use This

Every bid change should have a job:

• defensive moves are trying to reduce waste or protect margin
• scaling moves are trying to capture more profitable volume

Do not mix those intents in your head. The same numeric change can be smart in one context and reckless in another.

What To Watch For

Bid oscillation is usually a process problem, not a math problem. Teams often move bids again before the previous move has had enough time or traffic to teach them anything.

Large changes also hide learning. If you move too far too quickly, you cannot tell whether the result came from the direction of the decision or simply from the size of the shock.

FAQ

Should all bid changes be the same size?

No. The right change size depends on confidence, traffic quality, and how expensive it would be to be wrong.

Where do I verify that the bid actually changed?

Use Activity Log first, then judge performance after enough new data exists.

Next Steps

Go to Lower ACOS if the bid discussion is really an efficiency problem, or return to Work Command Center if you need Prism to prioritize the queue first.