Meta's Acquire new customers setting is open to everyone. Who does it count as new?

KEY TAKEAWAYS

1. Meta's automatic exclusion defaults to 730 days of buyers. Guides that say 180 describe a limit that changed on 18 May 2026.
2. Meta, Google and Shopify each define a new customer differently. Pick one definition before you judge results.
3. In Haus's 640 tests, 30% to 35% of Meta's measured impact fell on people who were not new customers.
4. The switch pays when the new customers gained are worth more than the returning orders lost that the ads actually caused.

On 6 October 2026, at Advertising Week in New York, Meta opened Customer Lifecycle Strategy to all advertisers. A week later, Jon Loomer found that choosing its new-customer option now adds an exclusion you cannot remove. Much of the advice published around it rests on a number that is out of date. Before you switch it on, settle who the setting treats as new, and what you give up by excluding everyone else.

What does the setting actually do?

It is a choice at the top of an ad set in a manual Sales campaign. Common Thread Collective notes it is not available in Advantage+ shopping campaigns. There are two options. The default, now labelled Get conversions from all audiences, lets Meta find conversions wherever they are cheapest. The alternative is Acquire new customers.

Meta's own description of that option, in screenshots Loomer published in April, says it suits advertisers for whom first purchases are worth more, and warns that cost per result may go up. Foxwell Digital's screenshots show a second choice inside it: reach only engaged people who have not bought, or reach people with no contact with the brand as well.

Meta told MediaPost that marketers can mark a campaign as acquisition and get recommendations about which existing customers to exclude. It is also testing a way to exclude "an entire group of existing customers with a single Audience Label." Loomer calls the setting a successor to the Existing Customer Budget Cap that Advantage+ shopping campaigns used to offer. Meta removed that cap and told advertisers to rebuild it by hand.

Who does Meta count as new?

Anyone outside your exclusions. By 13 October those exclusions started with one Meta builds for you. Loomer's account, checked that day, showed a fixed line: customers from your conversion data, 730 days. It is built from purchase events sent to Meta, not from your custom audiences. The window can be set to 730, 180, 90, 60 or 30 days. The default is 730. The exclusion cannot be turned off, though you can add custom audiences to it.

Several guides say something else. Common Thread Collective's 11 October explainer defines a new customer as someone with no purchase in 180 days. Jason Burlin's September guide uses the same figure. The explanation is a platform change in May. Meta's notice, quoted by Loomer, said that from 18 May 2026 purchase-event audiences built from website and app activity could hold 730 days, and that existing 180-day audiences would update automatically unless the advertiser opted out. Common Thread Collective reported the same change on 28 May. So 180 days was the old ceiling. It is now one option among five.

David Tamachi points out that he could not find a public Meta help page confirming the migration wording. The evidence is Meta's in-account notice and what practitioners see in their accounts. Check your own ad set before you rely on any of these figures, including ours.

Why will your store disagree with Meta?

Because each system defines new in its own way.

Shopify's help centre says its customer reports use each customer's entire order history. Under that rule, a buyer from three years ago is returning. Meta will treat the same person as new if you choose a 730-day window. It will also miss buyers who purchased offline or whose events never reached Meta. Loomer makes that point and suggests adding a customer list. Google's documentation admits the same gap on its side: some ads in new-customer-only campaigns may still reach returning customers for privacy or technical reasons.

The result is predictable. A new-customer ad set will still show returning orders in your store. That does not mean the setting is broken. Decide in advance which definition settles the test, and use only that one.

Is excluding existing customers free?

The advice splits. Common Thread Collective recommends the setting in every active prospecting ad set. Loomer thinks the need is real but rare. In his view, most advertisers who exclude customers do so because they misread their conversion reports.

The best outside data comes from incrementality tests, not platform reports. Haus analysed 640 Meta experiments run on its platform since the start of 2024. The average advertiser spent about $14 million a year on Meta. In those tests, 70% of Advantage+ campaigns' measured impact fell on new customers, against 65% for manual setups. Retargeting took about 15% of spend. Measured studied more than 10,000 Meta campaigns from 2025 in the US and Canada, using geo tests and marketing mix models calibrated by experiments. It found 64% of incremental conversions came from new or reactivated customers.

Read together, these say that roughly a third of what Meta actually causes lands on people who are not new. That share is not zero, so excluding them has a price. Two limits apply. Both figures describe whole campaigns, not the marginal value of a prospecting ad shown to a recent buyer. And Haus does not define a new customer, while Measured counts lapsed buyers who return as new. Burlin's worked model assumes 80% of returning-customer purchases would have happened anyway. That is his assumption, not a measurement. No one we found has published a controlled test of this setting.

How do you know if the switch paid?

Compare what you gain with what you lose, both in store data. The sum needs four numbers:

• extra first-time customers after the switch
• the value of a new customer over a period you trust, such as 12 months of contribution
• returning orders lost from the same ad sets
• the contribution of one returning order

The break-even share is the extra new customers times their value, divided by the lost returning orders times their contribution. If fewer than that share of the lost returning orders were truly caused by the ads, the switch paid.

A hypothetical example, with invented numbers. A brand spends $50,000 a month on prospecting. On the default setting Meta reports 500 purchases, a $100 CPA. The store shows 300 first-time and 200 returning orders, so cost per new customer is $166.67. After the switch Meta reports 380 purchases, a $131.58 CPA. The store shows 340 first-time and 40 returning orders. Cost per new customer falls to $147.06.

The platform CPA rose 32%. The brand gained 40 new customers worth $120 each, or $4,800. It lost 160 returning orders worth $40 each, or $6,400. The break-even share is 4,800 divided by 6,400, which is 75%. Unless more than three in four of those lost repeat orders were caused by the ads, the switch paid. Against Burlin's assumption that only one in five is caused, it paid clearly. Against a measured share, you would know.

Which window fits which business?

A seven-step test

1. Write down your definition of new. Use the store's: no earlier order. If you prefer a window, write that down too, and keep it for the whole test.
2. Complete the customer data. Send purchases, including offline sales, to Meta. Add a customer list. Set the Existing Customers and Engaged Audience segments before you start, since Loomer notes the breakdown only applies from the moment they are saved.
3. Pick the window from your own repurchase curve. Pull second orders from the store and find what share arrive within 30, 90, 180 and 730 days of the first. Choose the window that matches the gap you want to keep for retention.
4. Record a baseline. Run two to four weeks on the default. Note store first-time orders, returning orders and spend for the ad sets in scope. Use the audience segments breakdown alongside.
5. Switch all prospecting at once, or by region. Burlin warns that switching some campaigns lets the rest absorb existing customers. If you cannot switch everything, use a geo test built for average lift.
6. Read the result in store data. Compare cost per new customer, returning orders lost, total orders and contribution. Expect the platform CPA to rise, as Meta's option text says it may. Cold audiences may need different creative, so size the test to your budget.
7. Run the break-even sum. If the share you need is high, measure it with a holdout before you decide. Report the result on the same basis you use for Meta attribution.

The setting does one thing well. It states your intent in the ad set instead of in a list you have to maintain. Whether that intent is right depends on how often your customers come back. People who already trust a brand also find it outside paid channels. If you want to see where yours stands there, the waitlist is open.

Questions

Does Meta's Acquire new customers setting use a 180-day window?

Not by default. In the account Jon Loomer checked on 13 October 2026, the automatic exclusion covered buyers from the last 730 days, with 180, 90, 60 and 30 days as alternatives. The 180 figure dates from before 18 May 2026, when that was the longest window a purchase audience could hold.

Can I turn off the automatic customer exclusion?

No. Once you choose Acquire new customers, the exclusion built from your purchase events stays in place. You can shorten its window and add custom audiences. To exclude only the buyers of one product, use Get conversions from all audiences and add your own exclusions instead.

Why does my store still show returning customers from a new-customer ad set?

Because Shopify counts anyone with an earlier order as returning, with no time limit. Meta only excludes people it can match, inside the window you chose. Buyers from before the window, offline buyers and unmatched buyers can still be reached and will show as returning in the store.

Should every prospecting ad set use Acquire new customers?

Not by default. Incrementality studies from Haus and Measured put roughly a third of Meta's measured impact on people who were not new. Excluding them has a price. Work out the break-even share for your account, then test it on the whole prospecting budget or by region.

Sources

1. MediaPost, Meta Unveils Enhanced AI Business Assistant, Automated Ad Tools mediapost.com
2. Social Media Today, Meta introduces new AI-powered tools at Advertising Week socialmediatoday.com
3. Jon Loomer Digital, Meta Adds Automatic Customer Exclusions jonloomer.com
4. Jon Loomer Digital, Customer Lifecycle Strategy, Attribution Updates, and More jonloomer.com
5. Jon Loomer Digital, No More Existing Customer Budget Cap jonloomer.com
6. Jon Loomer Digital, ChatGPT Ads Go Self-Serve, Purchase Retention Expands, and More jonloomer.com
7. Jon Loomer Digital, Breakdown by Audience Segments jonloomer.com
8. Common Thread Collective, Meta's Customer Lifecycle Strategy Is Now Live for All Advertisers commonthreadco.com
9. Common Thread Collective, Meta Extends Purchase Audience Retention to 730 Days commonthreadco.com
10. David Tamachi, Meta 730-Day Custom Audiences: The Retargeting Exclusion Trap davidtamachi.ca
11. Foxwell Digital, Meta Customer Lifecycle Strategy: What to Know 2026 foxwelldigital.com
12. Haus, The Meta Report: Lessons from 640 Haus Incrementality Experiments haus.io
13. Measured, Driving New Customer Acquisition: An Incrementality Analysis of Meta Technologies Platform Performance measured.com
14. Google Ads Help, How customer lifecycle goals detect customer segments support.google.com
15. Google Ads Help, Measure your lifecycle goals campaigns support.google.com
16. Shopify Help Center, Customers reports help.shopify.com
17. Jason Burlin, Acquire New Customers: The Meta Setting Most Brands Ignore jasonburlin.com

NEXT STEP

See where your brand stands.

Your Digital Authority Score, who AI recommends for your target questions, and the placements that close each gap. Your first report is free.

Get your free report →