Meta 730-Day Purchase Audiences: What Malta Advertisers Must Change in 2026

On 18 May 2026, Meta quietly raised the maximum retention window on Purchase-based custom audiences from 180 days to 730 days, and auto-migrated existing 180-day purchase audiences unless advertisers manually opted out, ad account by ad account. Most Malta advertisers never saw the notification. Many are now running prospecting campaigns that silently exclude two years of buyers instead of six months.

That is not a small reporting quirk. In a market of roughly half a million people, the difference between a 180-day and a 730-day exclusion pool is often the difference between a campaign that can still find people and one that has quietly run out of country. If you manage Meta Ads in Malta, this is the single highest-leverage setting to audit this quarter.

This guide covers what actually changed, why the small-market maths makes it more consequential here than in the UK or Germany, the exclusion trap that is costing local accounts real revenue, and the five plays worth running now that two years of purchaser data is addressable.

What Meta actually changed

Custom audiences built from website and app activity using the Purchase event can now be set to a retention window of up to 730 days. Previously the ceiling was 180 days for most event-based audiences.

Three details matter more than the headline:

• It applies to Purchase events specifically, not ViewContent, AddToCart, Lead, or CompleteRegistration, which remain on their existing shorter windows.
• Existing audiences were migrated automatically. Audiences sitting at 180 days were rolled up to 730 days on the switchover date. Opting out required manual action in each ad account before the deadline.
• It flows downstream. Anything built on top of a purchase audience inherits the change: exclusions, lookalike seed sources, Advantage+ audience signals, and lifecycle segments all shifted underneath campaigns that nobody touched.

If you have not opened your Audiences tab since April, assume your windows changed. Verify rather than trust.

Why 730 days hits differently in Malta

The strategic value of a longer retention window is inversely proportional to your addressable market size. Malta is roughly 560,000 residents. Strip out under-18s, non-Facebook users, and people outside your category, and a typical local retailer or service business is bidding into an addressable pool of tens of thousands, not millions.

Three consequences follow directly.

1. Retargeting pools were previously too thin to be useful

A Maltese ecommerce brand doing 40 orders a month builds a 180-day purchaser audience of roughly 240 people. Meta will not reliably deliver against an audience that small; it either under-delivers or bleeds into broad. At 730 days, that same brand has a pool closer to 960, thin, but now workable as a lookalike seed and genuinely useful as a win-back segment.

2. Lookalike quality improves where it was previously unbuildable

Lookalike audiences need a meaningful seed. Many Malta accounts could never assemble one from purchasers alone and defaulted to page engagers or video viewers, far weaker intent signals. Two years of purchase data changes what is buildable, which matters for anyone running structured lead generation in Malta where the qualified pool is inherently small.

3. Frequency and audience burn become the binding constraint

Small markets saturate fast. A wider exclusion window is a blunt instrument for fatigue management, but a blunt instrument applied to the wrong list does real damage, which brings us to the problem most accounts have right now.

The exclusion trap: the mistake costing Malta accounts money

Here is the failure mode. Standard prospecting hygiene says: exclude purchasers from cold campaigns so you do not pay to acquire someone you already own. Sensible at 180 days. Actively harmful at 730 days for most categories.

If your cold campaign excludes everyone who bought in the last two years, you have removed from reach:

• The customer who bought a mattress 22 months ago and is now furnishing a second bedroom
• The client who used your service last summer and is due to renew
• The buyer whose annual replenishment cycle is 12–18 months

These are your highest-converting available prospects. They have transacted, they trust the brand, and their acquisition cost is a fraction of a genuine cold user. Excluding them is not hygiene. It is deleting your best segment from the auction.

The damage compounds in Malta because the remaining unexcluded pool is small to begin with. Remove two years of buyers from a market of 560,000 and delivery costs rise sharply as Meta hunts for eligible impressions in a shrinking pool. Rising CPMs with flat creative and flat budget is the tell.

The correction

Replace one blunt exclusion with a tiered structure:

• Exclude 0–90 days from cold prospecting. Recent buyers genuinely should not see acquisition creative.
• Do not exclude 90–730 days from prospecting, or better, target them deliberately in a separate win-back campaign with different creative and offer.
• Match the window to your actual repurchase cycle. A café should think in weeks. A furniture retailer in years. Pull the real number from your own order data rather than accepting a platform default.

This is the kind of structural decision that belongs in an account's growth strategy rather than being left to whichever setting the platform migrated you into.

Five plays worth running with 730 days of purchaser data

Play 1: Build the win-back campaign you could not build before

Target purchasers from 270–730 days ago, exclude the last 90. Creative should acknowledge the relationship. "It's been a while", a returning-customer offer, a new-range announcement. This audience typically returns the lowest cost per purchase in the account and previously did not exist as an addressable segment for most Malta advertisers.

Play 2: Reseed your lookalikes

Rebuild lookalike audiences from the 730-day purchaser list rather than the 180-day one. More seed data generally produces a more stable model. Test the new lookalike against the incumbent rather than swapping blind, with small seeds, more data is usually better but not guaranteed.

Play 3: Feed Advantage+ a better signal

Advantage+ campaigns accept audience suggestions as directional input. A two-year purchaser list is a materially richer signal than six months. Given how much control Meta's Andromeda and GEM delivery systems have absorbed in 2026, the quality of the signal you hand the system is one of the few levers still fully under advertiser control.

Play 4: Build a value-tiered structure

Two years of purchase history is enough to segment by order value and frequency. Separate one-time low-value buyers from repeat high-value buyers and treat them differently: different offers, different budgets, different creative. Most Malta accounts have never had the data density to justify this until now.

Play 5: Re-baseline your CPA targets after the change

Changing exclusion windows changes the composition of who your campaigns reach, which changes blended CPA. Do not compare post-change performance to pre-change benchmarks without accounting for it. This is a recurring theme across paid media in Malta in 2026, several platform changes this year have shifted reported numbers without any change in real-world performance.

How this interacts with the rest of the 2026 Meta stack

The retention change did not land in isolation. Meta spent 2026 moving automation from an option to a default, rebuilding delivery around its Andromeda retrieval system and the Generative Engagement Model, and narrowing manual control over placements, audiences and creative rotation across the Sales, App and Leads objectives.

The practical implication: as manual targeting levers disappear, the inputs you still control, first-party data quality, Conversions API coverage, exclusion logic, and creative, carry proportionally more weight. A 730-day purchase audience is one of the last genuinely high-signal inputs an advertiser can hand the system. Getting it wrong is more expensive in 2026 than the same mistake would have been in 2023.

One related note for Malta advertisers evaluating channel mix: TikTok Ads are still not available in Malta as an ad delivery market, though a rollout is expected during 2026. Until that lands, Meta and Google remain the only two platforms with meaningful paid reach against a Maltese audience, which raises the stakes on getting Meta's audience architecture right.

A Malta implementation checklist

1. Open every ad account you manage and check the retention window on all Purchase-based custom audiences. Assume they were migrated.
2. List every campaign using a purchase audience as an exclusion. These are the ones at risk.
3. Pull your real repurchase interval from your ecommerce or CRM data, median days between first and second order. This number, not a platform default, should set your exclusion window.
4. Rebuild exclusions as tiers (0–90 exclude, 90–730 target separately).
5. Verify Conversions API coverage before relying on any of this. A 730-day audience built on incomplete purchase data is a 730-day audience with holes in it.
6. Launch one win-back campaign against the 270–730 day segment and measure it separately from prospecting.
7. Re-baseline benchmarks and note the change date in your reporting so future-you knows why the line moved.

Frequently asked questions

What is the Meta 730-day purchase audience update?

On 18 May 2026, Meta raised the maximum retention window for custom audiences built on Purchase events from website and app activity from 180 days to 730 days. Existing 180-day purchase audiences were automatically migrated to 730 days unless the advertiser manually opted out in each ad account before the switchover date.

Does the 730-day window apply to all Meta custom audiences?

No. The extension applies specifically to Purchase events within website and app activity custom audiences. Other events, ViewContent, AddToCart, Lead, CompleteRegistration, keep their existing shorter retention windows. Customer list audiences uploaded manually are governed by separate rules.

Should I exclude 730 days of purchasers from my prospecting campaigns?

For most businesses, no. Excluding two years of buyers removes your highest-intent, lowest-cost available audience from reach, and in a small market like Malta it shrinks the eligible pool enough to push CPMs up. A better structure is to exclude only recent buyers (roughly 0–90 days) from cold campaigns and target the 90–730 day segment in a dedicated win-back campaign with its own creative and offer.

How does this change affect small Maltese businesses with low order volume?

It generally helps. A business doing 30–50 orders a month could never build a purchase-based retargeting or lookalike audience large enough for Meta to deliver against under a 180-day window. Extending to 730 days multiplies the available pool roughly fourfold, which can move a previously unbuildable audience into workable territory. The caveat is that exclusion logic must be corrected at the same time, or the benefit is cancelled out.

Can I still opt out and go back to a 180-day window?

You can set a shorter retention window manually on any custom audience at any time, the 730 days is a maximum, not a mandate. What you cannot do is recover data that was never retained. Set the window that matches your actual repurchase cycle, and set it deliberately rather than accepting whichever value the migration left behind.

The bottom line

Meta handed advertisers four times more purchaser data to work with, then migrated most accounts into a configuration that quietly weaponises it against them. In a market the size of Malta, where every campaign is already fighting audience-size constraints, that is not a setting worth leaving on default.

Audit the windows. Fix the exclusions. Build the win-back campaign. It is an afternoon of work against one of the better returns available in a Meta account this quarter.