New vs Returning Delivery on Meta: Who Is Really Seeing Your Ads?

Meta's Audience segments breakdown shows how your spend divides between a new audience, an engaged audience and existing customers. In a healthy prospecting setup most delivery goes to new people, so a split that leans heavily toward engaged or existing customers usually points to a setup problem. Pull the split for both spend and revenue, turn any warm share into dollars, and confirm what you see with a net new reach read before you change anything.

A note on where this comes from: we did not invent these reach reports. The methods and formulas here are pulled together from Meta's own tools and documentation, reports we have seen across the industry, and other practitioners' work. We are sharing how we use them, not claiming to be the originator.

Definition

New vs returning delivery is how your spend and results split between people new to your brand and people already engaged or buying, shown in Meta's Audience segments breakdown.

What is new vs returning delivery?

New vs returning delivery is the split of your ad spend between people who are new to your brand and people who already know it. Meta does not let you set this split. Its budget system decides it, auction by auction.

Here is why it matters. Warm people (site visitors, social engagers, past buyers) are easy to convert, so reported results on them look great. But many of them would have bought anyway. If too much of your budget drifts toward them, your ROAS (return on ad spend) can look healthy while the business stops growing. Knowing the split is how you catch that.

What are Meta's Audience segments?

Audience segments is a reporting breakdown that labels delivery as new audience, engaged audience or existing customers, based on audience definitions you set in your ad account. It is the most direct view Meta gives you of who your money is actually reaching.

  • New audience: people who do not fall into your engaged or existing customer definitions.
  • Engaged audience: people who have interacted with your brand, such as site visitors or people who engaged with your page or ads, depending on how the account defines it.
  • Existing customers: people in the customer audience you have told Meta about, usually from your purchase data or customer lists.

The labels are only as good as the definitions behind them. If your existing customer list is old or missing, a lot of past buyers will show up as "new." Check those settings before you trust the split.

Where to find it

  1. In Ads Manager, open the campaign, Ad Set or ad you want to check.
  2. Open the Breakdown menu and choose Audience segments.
  3. Add the columns you care about: Amount spent, Purchases, Purchase conversion value and Reach.
  4. Set the date range you want to review.

The breakdown is native to Advantage+ sales campaigns. Many manual campaigns do not support it. If you cannot find it, skip to the section on what to do when it is missing.

What split should you expect?

In an account built for growth, most delivery should go to the new audience, with a smaller slice to engaged and existing customers. An even split, or delivery concentrated on existing customers, is a red flag and usually a setup problem.

Remember that nobody controls this split directly. There is no setting that forces it, so there is no exact right answer. What we look for is whether the split fits the campaign's job:

  • Creative Testing (strict exclusions). What you hope to see: Almost all new audience. Anything else suggests exclusions are not working.
  • Scaling Campaign (past customers excluded only). What you hope to see: Mostly new, with some engaged delivery by design. Little existing customer spend.
  • Remarketing. What you hope to see: Mostly engaged and existing, by definition. Judge it on incrementality, not ROAS.

The right mix also varies by account, so treat any rule of thumb as a starting point. The trend inside the account over time is often more telling than a single month's split.

How do you do the math on the split?

Calculate each segment's share of spend and share of revenue, then cost per purchase for each, and put them side by side. The gap between spend share and revenue share is where the story is.

Here is an invented example for one Scaling Campaign over one month:

  • New audience. Spend: $70,000. Purchases: 1,000. Revenue: $150,000. Share of spend: 70%. Share of revenue: 50%. Cost per purchase: $70
  • Engaged audience. Spend: $20,000. Purchases: 500. Revenue: $80,000. Share of spend: 20%. Share of revenue: 27%. Cost per purchase: $40
  • Existing customers. Spend: $10,000. Purchases: 500. Revenue: $70,000. Share of spend: 10%. Share of revenue: 23%. Cost per purchase: $20
  • Total. Spend: $100,000. Purchases: 2,000. Revenue: $300,000. Share of spend: 100%. Share of revenue: 100%. Cost per purchase: $50

The calculation lines:

  • Share of spend for existing customers: $10,000 divided by $100,000 = 10%
  • Share of revenue for existing customers: $70,000 divided by $300,000 = 23%
  • Cost per purchase for existing customers: $10,000 divided by 500 = $20

How to read it

At first glance, existing customers look like the best deal in the account: $20 per purchase against $70 for new people. This is the trap. Those are people who already bought from you. Many of them would have bought again with no ad at all. Their cheap purchases are the least incremental in the account, meaning the least likely to have been caused by the ad.

The tell is the gap. Existing customers took 10% of spend but 23% of reported revenue. When warm segments take a small share of spend and a much larger share of revenue, the account is often paying to reach buyers it already has, and the reported ROAS is flattered by it. Our post on whether your remarketing is incremental goes deeper on this.

Turn the warm share into dollars

Percentages hide size. In the example, $10,000 a month on existing customers is $120,000 a year. The same 10% at a tenth of the spend would be a rounding error. Always state the dollar figure next to the percentage, because that is what tells you whether it is worth fixing.

Why does warm delivery leak into campaigns that exclude it?

Because exclusions do not catch everyone. Privacy changes, opted out tracking and people Meta cannot match mean some of the people you exclude are invisible at auction time, so they get served anyway.

Two things follow from that. First, some existing customer delivery will always show up, even with every exclusion set correctly, so a small residual is not a setup failure. Second, never promise anyone that exclusions guarantee zero warm spend. In Advantage+ campaigns, the existing customer setting acts more like a soft budget cap than a hard wall, so a small amount of existing customer spend there is expected once you have verified the settings are on.

What is not normal is a large or growing share. That usually means the customer list is stale, an exclusion is missing, or a campaign that should prospect is set up like remarketing.

What if your split looks wrong?

Confirm it before you act. A strange split can come from bad audience definitions as easily as from bad delivery.

  1. Check the definitions. Is the existing customer audience current and complete? Is the engaged audience defined the way you think?
  2. Check the exclusions. Is the past customer exclusion on in every prospecting and scaling Ad Set?
  3. Run a campaign scoped net new reach read. If Audience segments says mostly new but the campaign's net new reach is flat, the "new" group is probably being served over and over. See net new reach for a Scaling Campaign.
  4. Look at the attribution windows. Warm delivery inflates 7 day click and view numbers more than 1 day click. Comparing them helps size the effect; see Meta attribution windows.
  5. Run a Conversion Lift test. It is the only way to measure how many of those warm conversions the ads actually caused. See Conversion Lift tests.

What do you do when Audience segments is not available?

Use daily frequency and campaign level net new reach instead. Together they tell you roughly the same thing: whether an ad or campaign is meeting new people or serving known ones again.

Daily frequency as a stand in

Daily frequency is impressions divided by reach for one day. An ad that reaches 100,000 people with 105,000 impressions in a day has a daily frequency of 1.05, which means almost everyone saw it once. The closer an ad stays to 1.0 day after day, the more it behaves like a prospecting ad. The higher it sits, the more it behaves like a warm audience ad, whatever campaign it lives in.

To get one number per ad over a period without blurring it, weight each day by spend:

  • Spend weighted frequency = the sum of (daily frequency times daily spend) divided by total spend.

Invented example for one ad over three days:

  • 1. Daily frequency: 1.05. Spend: $100. Frequency times spend: 105
  • 2. Daily frequency: 1.10. Spend: $200. Frequency times spend: 220
  • 3. Daily frequency: 1.30. Spend: $100. Frequency times spend: 130
  • Total. Daily frequency: . Spend: $400. Frequency times spend: 455

Spend weighted frequency = 455 divided by 400 = about 1.14. This is better than pulling one blended frequency for the three days, which drifts up because the same people pile up across days.

Sort ads into rough top, middle and bottom of funnel groups by this number and add up spend in each group, and you have a funnel split built from behavior instead of labels. Our guide to daily frequency as a funnel signal shows how to build it.

Campaign net new reach as the cross check

Frequency tells you about each ad. Campaign net new reach tells you whether the campaign as a whole keeps finding new people. Where Audience segments is missing, that campaign level read is the closest thing to a direct measure.

Key takeaways

  • Audience segments splits delivery into new audience, engaged audience and existing customers; Meta decides the split, not you.
  • Most delivery should go to new people in prospecting and scaling; an even split or heavy existing customer delivery is a red flag.
  • Compare share of spend to share of revenue; warm segments taking a small share of spend and a big share of revenue often means paying for buyers you already had.
  • Always state warm spend in dollars, not just percent.
  • Exclusions leak because of privacy signal loss, so a small residual is normal but a large or growing one is not.
  • When the breakdown is missing, use spend weighted daily frequency and campaign net new reach.

FAQ

Can I force Meta to spend more on new people? Not directly. You can shape it with exclusions, creative built for strangers and a clean campaign structure, but Meta's budget system makes the final call. That is why you measure the split rather than assume it.

Is engaged audience delivery bad? No. Engaged people are often close to buying, and a Scaling Campaign is meant to reach some of them. It becomes a problem when engaged delivery crowds out new people and net new reach falls.

Why does existing customer delivery show cheap purchases? Because past buyers are the easiest people to convert. Many would have bought again without the ad, so their low cost per purchase overstates what the ad caused.

How often should I check the split? Monthly in reports and audits, and any time ROAS is flat while spend grows. Read it next to net new reach so one number can confirm the other.

See who your ads are really reaching

If you want to know how much of your Meta budget is finding new people, we offer a free Meta account audit. It includes a full account audit plus net new reach and cumulative reach reports. Request yours here.

Keep reading

More on reach reports:

See every reach report article

Ready for your free Meta audit?

We will review your account and run net new reach and cumulative reach reports, free.

Get your free audit

Check what you learned

Tap an answer to see if you got it right.

1. Who decides the split in Meta's Audience segments?

Audience segments shows how Meta delivered your ads: new audience, engaged audience and existing customers.

2. In prospecting and scaling, where should most delivery go?

An even split or heavy existing customer delivery is a red flag.

3. Warm segments take a small share of spend but a big share of revenue. What does it often mean?

Those people would likely buy anyway. Always state warm spend in dollars.

Ready for the full test? Take the Reach Reports exam and get certified. Want this run on your account? Get my free audit.

About the author

Jake Mehani, founder of Creative Analytics

Jake has spent 15 years running Meta and paid social ads for consumer brands. Meta published his RBX Active Advantage+ case study, he taught the DMAnc Meta and TikTok ads masterclass, and he spoke at Motion's 2024 Creative Strategy Summit. His senior team runs media buying, creative and measurement like an in house team.

Jake on LinkedIn  |  Case studies  |  Free Meta audit

Previous
Previous

Daily Frequency on Meta: Split Your Spend Into TOF, MOF and BOF

Next
Next

Meta Ad Set Audience Overlap: How to Detect It in Ads Manager, With the Math