Skip to content
whitead

PMax New Customer Acquisition (NCA)

glossary google ads updated 2026.05.20 7 min read
// contents · 9 sections
  1. What NCA is
  2. How it works
  3. What changed in 2025-26
  4. Common misconceptions
  5. "If I just turn on autodetection, do I need a Customer Match list?"
  6. "What should I set the new-customer value to?"
  7. "Can I use NCA for lead-gen, not just ecommerce?"
  8. Related concepts
  9. Sources

PMax New Customer Acquisition (NCA) is a campaign-level customer-lifecycle goal that lets Performance Max either bid higher for net-new customers ("Value" mode) or bid exclusively for net-new ones ("New customers only" mode).

What NCA is

NCA is one of the customer-lifecycle goals available on Performance Max, Search, Shopping, and Demand Gen campaigns. It tells Smart Bidding to treat a "new customer" conversion as worth more than (or instead of) a "returning customer" conversion. Google Ads Help frames it as:

"Customer lifecycle goals are a suite of goals that help you connect with customers at every stage of their purchase journey within Google's Performance Max, Search, Shopping, and Demand Gen campaigns."
— Google Ads Help: About customer lifecycle goals, retrieved 2026-05-20

NCA was originally announced for PMax on 2022-04-26 alongside the first wave of PMax features and has since expanded to Search, Shopping, and Demand Gen, plus a "high-value new customer" extension that rolled out beyond beta in 2025 (Google Blog: New features to grow your business with Performance Max, 2022-04-26).

Mechanically, NCA sits on top of tCPA / tROAS Smart Bidding — it does not replace the bid strategy, it modifies the conversion-value signal Smart Bidding optimises against.

How it works

NCA combines three identification signals to decide whether the user clicking on the ad is "new" or "existing":

  1. Customer Match lists — first-party email/phone lists uploaded into Google Ads. Anyone in the list is treated as existing; everyone else is a candidate for "new."
  2. The new_customer parameter on the Purchase conversion tag — a server- or client-side flag set to true for new customers, false for returning. Setup via Google tag, GTM, Shopify, Firebase, GA4, or offline conversion imports is documented in Google Ads Help: Set up new customer acquisition parameter, answer 12077475.
  3. Google's autodetection of logged-in users — fallback signal based on signed-in Google account history. Per Google Ads Help, this works best when supplemented by the first two signals.

Customer-acquisition reporting works only with conversions of type Purchase — not Sign-up, not Lead — per Google's own documentation. In Value mode, Smart Bidding adds a per-conversion premium (the value you set) to net-new conversions; in New customers only mode, returning-customer conversions are removed from the optimisation signal entirely.

To configure: Campaigns → select PMax campaign → Settings → Additional settings → Customer acquisition → choose mode + set new-customer value.

What changed in 2025-26

Methodology note (Whitead audit logic). Whitead's PMax audit flags NCA misconfiguration on three signals: (a) NCA is enabled but no Customer Match list is attached at the account or campaign level, (b) a Customer Match list is attached but match-rate is below 40% or the list is older than 180 days, or (c) NCA is in New customers only mode while the campaign has fewer than 30 conversions per month (Smart Bidding starvation risk on a halved signal). Severity rises from MED to HIGH when (a) or (b) coincides with a 14-day spend increase >20% on the NCA campaign — the bid premium is amplifying the cost of treating returning customers as new. The fix path always starts with refreshing the Customer Match list before touching the value setting; the value premium is a second-order lever.

Common misconceptions

"If I just turn on autodetection, do I need a Customer Match list?"

Yes. Autodetection covers logged-in Google users with prior purchase history that Google can attribute back to your conversion data, but coverage is partial and lags behind your own first-party data. Without a Customer Match list, the system over-identifies existing customers as new — and your NCA bid premium is wasted on people who would have converted anyway. Per Google Ads Help: Troubleshooting the new customer acquisition goal, answer 14005157, Customer Match is the strongest signal.

"What should I set the new-customer value to?"

There is no Google-published formula. Operator practice from Search Engine Land, ZATO, and Adchieve converges on: set the premium near the gross margin of an average order, so Smart Bidding trades acquisition cost against retention on a unit-economics basis. Setting it too low (e.g. €1 against a €100 AOV) gives NCA no bid leverage; setting it too high inflates reported ROAS with imaginary revenue and can blow out the budget on borderline-intent traffic. Always back the premium out of reported ROAS in your dashboard — Google adds it to conversion value but it is not actual revenue (ZATO PPC Marketing: Read before using NCA, 2021-08-31, updated 2026-04-21).

"Can I use NCA for lead-gen, not just ecommerce?"

Not directly. Per Google Ads Help: Set up new customer acquisition parameter, answer 12077475, customer-acquisition reporting works only with Purchase-type conversions. For lead-gen, the equivalent pattern is offline conversion import with deal stage as the conversion value — NCA's "new customer" framing does not map cleanly to MQL/SQL/closed-won.

  • Performance Max — the campaign type NCA modifies
  • Customer Match — the first-party list that defines "existing" for NCA
  • Smart Bidding — the tCPA / tROAS layer NCA modifies the value signal for
  • Asset Groups — NCA is set at campaign level, not per asset group
  • PMax Low ROAS Diagnostic — misconfigured NCA value premium is a common root cause of inflated reported ROAS without real lift
  • Performance Max Brand Cannibalization — NCA + branded queries is a classic trap: returning customers searching your brand get treated as new because they are not in the Customer Match list

Sources

// was this useful?
// anonymous · no personal data stored
← Back to Wiki