Meta Ads Retargeting: Window and Suppression Design

For the surrounding account decisions, compare Meta Ads Agency Handover: Access and Control and use Meta Ads Automation: A Safe Workflow Design as the next diagnostic.

In short

This guide owns one decision artifact: the filled, auditable structure below. Reader-supplied thresholds stay explicit; missing evidence stays missing.

Event-recency worksheet

Observed purchase-lag band	Cohort calculation	Window band	Message sequence	Suppression
Same-session / earliest observed band	Purchases whose first eligible event and recognized order fall in band ÷ eligible recognized purchases	Reader sets boundary from its own cumulative lag distribution	Reminder of viewed product/offer continuity; no invented urgency	Recognized purchasers and invalid-consent records
Consideration band	Purchases whose observed lag falls after band 1 and inside band 2 ÷ eligible recognized purchases	Boundary chosen where the reader sees a meaningful change in cumulative purchase share	Evidence, comparison, or objection handling supported on destination	Purchasers, expired offers, unavailable items
Long-lag band	Purchases beyond band 2 but inside justified maximum ÷ eligible recognized purchases	Maximum follows observed data, purpose, and retention rule	Re-entry message appropriate to stale intent, with current terms	Purchasers, deleted/withdrawn identifiers, stale catalog items

Cumulative purchase-lag share through day d = (eligible recognized purchases occurring within d days of the declared first event) ÷ (eligible recognized purchases old enough to reach day d). Source: consented event/order join by cohort age. Limitation: unobserved paths and selection remain. A holdout supports a causal read; a matched-period comparison is observational and must state concurrent changes.

Interpretation boundary

Use the purchase-lag recency bands only for its stated decision. Choose early, consideration, and long-lag boundaries from cohort-age purchase shares, then attach current messages and purchaser suppression. A matched-period result remains observational; use a suitable holdout for lift and exclude cohorts too young to reach each lag. Reader-supplied thresholds remain inputs, not universal standards.

Can software help?

Bach.ai audits your connected Meta account against 100+ checks, ranks what it finds by estimated impact, and proposes specific fixes. It stays read-only until you approve a change, then executes the approved change on Meta; connected Google Ads data is used for intelligence only. Think of it as an automated audit layer that surfaces issues and proposed fixes for your review — not a replacement for your team’s judgment, and it does not generate your creative.

FAQ

How do purchase-lag cohorts determine retargeting windows?

Choose early, consideration, and long-lag boundaries from cohort-age purchase shares, then attach current messages and purchaser suppression.

Why can’t matched-period retargeting results establish lift?

A matched-period result remains observational; use a suitable holdout for lift and exclude cohorts too young to reach each lag.

What can recency bands support when attribution is not incrementality?

It supports the bounded operating choice encoded by the purchase-lag recency bands. It cannot replace missing source records or turn platform credit and observed association into incremental impact.