Should we send a separate new-customer purchase event to Meta and optimize prospecting on it, instead of relying on customer exclusions?
Our answer
Test a new-customer event, don't switch to it. Where an ad set can reach about 50 new-customer purchases a week, optimize a duplicate of it on a separate new-customer purchase event. Keep your current customer exclusions on.
- Confidence
- Low
- Strongest proof
- 1 controlled test
- 2 practitioners, 5 opinions or claims
- Sources
- 7 sources
- 8 statements
DSB
New here? Take the 1-minute tour
Why this is the answer
3 reasons and our judgmentA new-customer event changes what Meta learns, while exclusions still reward it for every purchase (Mable). CustomerLabs, which sells the data platform that builds such an event, argues the same, and both pair the event with exclusions rather than replacing them.
The only results come from vendors. Northbeam measured 25.2% lower new-customer acquisition cost in Ka'Chava's ad-set A/B test, read in its own tool. CustomerLabs claims 19% lower at one consumer brand, but on Google Ads.
Purchase volume limits who can use a new-customer event. Mable keeps ad sets under about 50 conversions a week on Purchase plus exclusions. Jon Loomer warns that optimizing on a custom conversion narrows volume further.
Our judgment
Run the event on a duplicated ad set beside the Purchase-optimized one rather than switching prospecting outright. The only results are two vendor-run single-brand cases, one of them on Google, and the vendors' own setups run it on duplicated or added ad sets scaled gradually. Keep the exclusions you already use (Excluding existing customers from prospecting for which customers to exclude), since the event without them is untested. The advice applies to online stores generally, because the cases and guides are written for ecommerce brands, and Mable says the gain is small where buyers are almost all first-time buyers.
Our judgment, built from Northbeam, CustomerLabs and Mable: our reading of these sources, not a finding from any one of them.
What the sources say
8 statements from 7 sourcesDrew Sanocki reads rising frequency on cold campaigns as Meta running out of new people at that price, and says to check pixel-based new-customer CPA against your Shopify file. No data given.
“if ncCPA comes off the pixel, some of those new customers already bought from you. match it against your shopify file before you trust it.”
Jon Loomer says Meta's "Acquire new customers" setting now automatically excludes purchasers in your conversion data from the last 730 days, and this can't be turned off. Offline buyers still need a custom audience.
“Customers from your conversion data (730 days)”
Jon Loomer says that if you already lack conversion volume, optimizing for a custom conversion cuts it further; optimize for a standard event and add the custom conversion as a reporting column.
“If you're already struggling to get conversion volume, optimizing for a custom conversion will limit volume even more”
In a Northbeam-run test at one brand, Ka'Chava, Meta ad sets optimized to Northbeam's signal beat Meta's standard attribution by 43.2% on new-customer ROAS, read in Northbeam. Spend and duration are not given.
“ran a controlled A/B test at the Meta ad-set level with identical campaigns.”
Pixis, with no data, recommends Meta's Customer Lifecycle Strategy 'acquire new customers only' over manual exclusion lists, warns cost per result may rise, and says to judge on new-customer CAC and ROAS with holdouts.
“Meta also notes the tradeoff plainly, that choosing new-customers-only can raise your cost per result, since the algorithm is working from a smaller pool.”
Braincuber, a vendor of the integration it recommends, says to tag each purchase sent to Meta as new or returning from your own customer data, so Meta can optimize for new customers. No data.
“a returning customer buying again is a sale your email, SMS and loyalty flows would very likely have earned anyway”
CustomerLabs, which makes the data platform for it, says to send Meta a purchase event only for first orders and optimize a new ad set on it. Its one Google Ads case shows 19% lower new-customer CAC, no test design.
Mable, a tracking vendor, says excluding existing customers does not change Meta's optimization goal. It recommends optimizing for its new-customer purchase event and also excluding past buyers. No data.
“does not change the algorithm's underlying optimization objective”
In the full answer
For readers who sign in- What to do this week3 steps
- The numbers behind it2 figures from cited sources
- Where each source standsA chart placing each source on the question
- The strongest disagreementThe best case against the answer, from a cited source
- How sure we areWhat its confidence rests on, and what would change the answer