On May 18, 2026, Meta raised the maximum retention window for Purchase event custom audiences from 180 to 730 days. If you had an existing custom audience set to 180 days, Meta automatically expanded it. No notification. No opt-in. The audience got bigger and kept running.
Depending on how your retargeting stack is built, this is either a meaningful upgrade or a problem you need to fix today.
What changed and how it happened
Custom audiences built from website or app Purchase events could previously hold users for up to 180 days. Starting May 18, you can now set that window to up to 730 days — just over two years.
The auto-expansion of existing audiences is the part that deserves attention. Meta's stated logic: advertisers who set 180 days presumably wanted the widest window available. So when the ceiling moved, existing audiences moved with it.
That logic holds for some accounts. For others, 180 days was a deliberate choice — not a default they'd never revisited. If you built a purchase exclusion audience to suppress recent buyers from seeing prospecting ads, that audience just got a lot larger. People who bought in months 7 through 24 are now in your exclusion pool alongside people who bought last week.
Where the 730-day window actually makes sense
For high-ticket, long-cycle categories, 180 days was always too short. Someone who buys outdoor furniture, a mattress, a standing desk, or a luxury bag is not in-market again within six months. Running winback ads to them at month five made little sense — and excluding them entirely at month seven made even less.
730 days gives you room to build proper lifecycle campaigns. A 30-day post-purchase suppression to let the order arrive and settle. A 90-day window for accessories and add-ons. A 365-day window for category re-entry. A 730-day window for complete repurchase cycles.
Jewelry brands with seasonal gifting cadences. Apparel brands with annual wardrobe refreshes. Any subscription with a natural annual renewal cycle. These categories have always needed a longer audience memory than Meta allowed.
The actual risk for performance-focused accounts
The auto-expansion creates two specific problems worth checking.
First, exclusion drift. If you exclude recent purchasers from acquisition campaigns to protect your CPAs, your exclusion list just grew substantially. Depending on your purchase volume and the 13–24 month window, that could be tens of thousands of additional users excluded from prospecting — shrinking your addressable audience for no strategic reason.
Second, retargeting audience dilution. If you built a purchase retargeting audience to run loyalty or repurchase campaigns, you're now serving ads to people who bought 20 months ago alongside people who bought 20 days ago. Those two groups behave differently. The 20-month buyer might not even remember your brand. The ad that works for one won't perform the same for the other.
Neither problem is catastrophic. But both affect performance in ways that will be hard to diagnose if you don't know the audience definition changed underneath you.
What to do in your account
Open Audience Manager and filter for audiences using Purchase events. Look at the retention window on each one.
For any exclusion audiences, decide whether 730 days is right for your business. If you're suppressing recent buyers for 180 days, you may want to keep that explicit — not let it drift to two years. Rebuild those audiences with a manual window if needed.
For retargeting audiences, consider whether a single combined audience still makes sense or whether you want to segment by recency: 0–90 days, 91–365 days, 366–730 days. You'd target those groups differently anyway — the messaging, offer, and creative approach for someone six weeks post-purchase is not the same as for someone who bought 18 months ago and probably forgot about you.
For seasonal brands, the 730-day window is worth building into your strategy now. Create an audience of buyers from last year's relevant season and you'll have it ready when that period returns.
The bigger pattern here
This is the fourth major account-level change Meta has made in the past six months that moved silently. The attribution redefinition in March. The auto-catalog sync. The Advantage+ audience expansion defaults. And now this.
Meta is consistently nudging accounts toward broader, looser data pools that its algorithm can optimize against. Whether that's good or bad depends on your business model. For brands with strong creative and healthy pixel data, larger audiences give the algorithm more room to find buyers. For brands doing precise lifecycle segmentation, each of these changes erodes the signal quality of the segments you've spent years building.
The answer isn't to fight the direction. It's to stay close enough to your account structure that you notice when something changed — and decide deliberately whether to go with it.
If you want to see how your current audience setup and attribution are actually performing, the free Gromerce audit gives you a read on where the gaps are in under three minutes.
Your retargeting audiences just changed. Whether they changed in a useful direction is worth finding out.
Sources: Meta for Business, Search Engine Journal, Vizup, Jon Loomer Digital, May 2026
