Every publisher with a consent banner faces the same sequencing question: do you wait for the visitor’s choice before requesting ads, or do you request immediately and sort it out afterwards? Both extremes cost money, and the best answer is usually somewhere in between.

The Core Tension

Ad auctions are fast and unforgiving. A bid request that goes out before consent is known can only carry limited signals, and in regions where consent is required for personalised advertising, you cannot send personalised requests until the user has agreed. But waiting for a choice delays the first auction, and a slow first ad often means a lower viewability score or a user who has already scrolled away.

Strategy 1: Block Until the Choice Is Made

Nothing ad-related fires until the visitor interacts with the banner.

• Pros: the simplest to reason about; every request carries a definitive consent state.
• Cons: visitors who never interact generate no ad revenue at all, and the first ad slot loads late for everyone who does.

This is the safest legal posture but the most expensive in lost impressions.

Strategy 2: Request Immediately With Restricted Signals

Ad calls fire straight away using non-personalised or limited requests, and upgrade to personalised once consent arrives.

• Pros: every page view can monetise from the first moment; no latency penalty.
• Cons: early requests fetch lower-value demand, and the setup must be configured so that nothing personalised leaks before consent is granted.

This is the approach most compatible with a no-choice-yet visitor, provided your tags honour the restricted state.

Strategy 3: Short Wait With a Timeout

The ad stack waits a short, fixed period for a stored or fresh consent decision, then proceeds with whatever state is known.

• Pros: returning visitors, whose choice is already stored, get personalised requests with almost no delay; new visitors fall back to restricted requests.
• Cons: the timeout needs tuning, and a value that is too long hurts viewability while one that is too short wastes the stored-consent advantage.

For most sites with a sizeable share of returning users, this hybrid is the strongest starting point.

What to Watch in Your Data

1. Time to first ad request on first visits versus returning visits.
2. Viewability of the above-the-fold slot in each strategy.
3. Share of impressions served as restricted versus personalised.
4. Page layout shift. A banner that pushes content, or ad slots that resize when the banner closes, hurts both experience and ad performance.

Practical Tips

• Reserve space for ad slots so late-loading ads do not shift the layout.
• Load the CMP script early and with high priority so the stored choice is read quickly.
• Test on slow mobile connections, not only on a fast office network.
• If you use header bidding, make sure the consent state is available to each bidder adapter before the auction starts, not afterwards.
• Re-run the comparison after any banner redesign, since banner size and position change how quickly people decide.

Where a CMP Fits In

FlexyConsent is a CMP certified by Google (CMP ID 317), IAB Europe and IAB Canada, with IAB TCF 2.4, IAB GPP v1.1 and Google Consent Mode v2 support, and a banner available in 58 languages. Its banner stores the visitor’s choice so returning visitors do not see it again, which is what makes the short-timeout approach work well.

Key Takeaways

• Blocking all ads until a choice is made is simple but loses revenue from visitors who never interact.
• Requesting immediately with restricted signals protects speed but earns less on first requests.
• A short timeout that reads stored consent is a good default for sites with many returning users.
• Measure time to first ad, viewability and restricted share before and after any change.
• Reserve ad space and load the CMP early to avoid layout shift and delays.