Retargeting only works if it lands when customers are actually ready to reorder. Too early and you pay to reach people who are not buying yet. Too late and they have already reordered or moved on. The Repeat Purchases page tells you the right window. This SOP turns that data into campaign settings.
When to use it: when setting up or reviewing retargeting and win-back campaigns.
Step 1: Find your peak reorder window
Open the LTV & Subscriptions report and go to the Repeat Purchases page. It shows the share of repeat orders that fall in each window: 0-7, 8-15, 16-30, 30-60, 60-90, 90-120 days, and beyond. Find the window with the largest share, for example 40% in the 30-60 day window.
Step 2: Check how many customers come back at all
Timing tells you when reorders happen. Retention depth tells you how many customers you can realistically win back. Open the Cohort Analysis page and use the Retention (percentage) view to see what share of a cohort has reordered by each month. If only 15% of customers ever come back, size your retargeting budget to that ceiling, not to your full customer base.
Step 3: Match your retargeting lookback to the window
Amazon's native retargeting does not target an exact range of days since purchase. Sponsored Display purchase remarketing uses rolling lookback windows: everyone who bought in the last 7, 14, 30, 60, 90, 180, or 365 days. The goal is to pick the lookback that covers your peak reorder window.
Cover the window: if most reorders happen at 30 to 60 days, a 60-day lookback covers it. A 14-day lookback reaches customers before they are ready to reorder, and a 365-day lookback spends most of the budget on customers who have long since reordered or moved on.
Know the overlap: a rolling window also includes very recent buyers, since a 60-day lookback includes someone who bought yesterday, and Sponsored Display cannot exclude them. If you want the exact window at scale, Amazon DSP can build it: for example, customers from the last 60 days with buyers from the last 30 days excluded.
Step 4: Use the window to judge Brand Tailored Promotions
Brand Tailored Promotions audiences come with fixed time definitions set by Amazon: brand cart abandoners cover the last 90 days, repeat customers and high spenders the last 12 months. There is no days-since-purchase setting to adjust. Use your reorder data to decide whether the promotion is worth the discount instead: if most customers who come back do so within 60 days, then customers beyond that window are genuinely lapsed, and a promotion to them is a true win-back rather than a discount on orders that would have happened anyway.
Step 5: Adjust per product and order type
Reorder timing differs by product, so read the Repeat Purchases page per product rather than once for the whole account. A consumable and a durable will have very different reorder curves. Both tools can follow per-product windows: Sponsored Display remarketing is scoped to the products you advertise, and Brand Tailored Promotions can be run at the ASIN level.
Also separate Subscribe and Save from non-subscribe cohorts when reading the data. Subscription reorders follow their own schedule, so a blended average can put your window in the wrong place. Base retargeting decisions on the non-subscribe behavior.
Step 6: Verify with cohorts
After adjusting timing, come back to the Retention (percentage) view and confirm that retention improves in the cohorts acquired after the change.
Tip: revisit this when your product mix or reorder behavior changes, the right window can shift over time. The same reorder prediction also feeds inventory planning: if 40% of repeat orders arrive 30 to 60 days after the first one, your stock planning should anticipate them.
