Remarketing targeting pairs each mode (views or purchases) with a selectable look-back window — commonly 30, 60, 90, or 180 days. Rather than picking one window based on an assumed product usage cycle (e.g. a 90-day supplement bottle → 90-day window), run multiple look-back windows simultaneously as separate targets within the same ad group and let performance data reveal which window actually converts. This treats the window choice as an empirical question rather than a category-based guess.
The best-performing look-back window is frequently not the intuitive one. A product with roughly a 60-day usage/repurchase cycle might actually perform best on a 90- or 180-day look-back rather than the matching 60-day window. Practical method: add several look-back periods (e.g. 30/60/90/180 days) as separate targets within the same ad group, let them run, then compare performance directly in the targeting tab rather than pre-selecting one window based on assumption.
Customer Loyalty Analytics (Brand Analytics Dashboard) displays an average repeat purchase interval for the brand. Use that interval as the starting hypothesis for a first-purchase retargeting campaign's look-back window, then split-test other look-back periods from there.
Rather than testing look-back windows in the abstract, tune the remarketing look-back window per loyalty segment: a shorter window for At Risk customers (catch them before they lapse), a longer window for Hibernating customers (they've already lapsed, so the window needs to reach back further).
Apply: Use the average repeat-purchase interval (available from the LTV spreadsheet output) as the starting-point number for both windows, rather than guessing.