Set up NC & RC Purchase events in Meta

Custom events give ad platforms more specific signals to optimize against. The most impactful is NC Purchase — it tells Meta to find buyers who’ve never purchased from you before, instead of optimizing toward repeat customers. Without it, Meta’s algorithm treats a first-time buyer and a fifth-time buyer as equally valuable conversions.

​

New customer purchase (NC Purchase)

NC Purchase fires only when an order comes from a customer with no prior Shopify order history. This enables NC CPA and NC ROAS reporting directly in Meta Ads Manager — two metrics that separate prospecting performance from retargeting performance.

​

Set up in Facebook

For the complete end-to-end walkthrough, see Complete guide to setting up NC Purchase metrics in Facebook Ads Manager.

When to use NC Purchase vs. standard Purchase

Custom purchase events supplement the standard Purchase event — they don’t replace it. Keep standard Purchase active alongside the custom events so Meta’s baseline models still receive every order.

Scenario	Recommended event
General ROAS optimization	Standard Purchase
Scaling new customer acquisition	nc_purchase
Prospecting campaigns targeting cold audiences	nc_purchase
Retargeting campaigns for existing customers	rc_purchase or standard Purchase
Brand awareness / top-of-funnel	Standard Purchase (broader signal)
Measuring retention campaign performance	rc_purchase

Allow a learning phase. When you switch an ad set from standard Purchase to nc_purchase, it re-enters Meta’s learning phase. Meta needs roughly 50 optimization events per week to exit learning, so start with your highest-spend or highest-volume campaign first if new-customer volume is low.Compare before shifting budget. Run nc_purchase-optimized campaigns alongside standard Purchase-optimized campaigns for 2–4 weeks and compare new-customer rate, CAC, and ROAS before committing.

​

Returning customer purchase (RC Purchase)

RC Purchase is the inverse of NC Purchase: it fires only when an order comes from a customer with existing Shopify order history. Together, the two events partition every order into acquisition vs. retention signals, which you can use separately in audience building, retargeting, and retention measurement. Classification uses the same Shopify-order-history check as NC Purchase — if the customer has zero prior completed orders, Upstack fires nc_purchase; if they have one or more, it fires rc_purchase. The standard Purchase event still fires for every order regardless of customer status, so baseline reporting is unaffected.

​

Why it matters

• Build retargeting audiences that exclude first-time buyers. Feed rc_purchase into custom audiences of recent repeat customers for win-back campaigns, loyalty offers, or cross-sells.
• Measure retention, not just acquisition. Treat rc_purchase volume and value as a direct signal of how well retention campaigns (email, SMS, paid retargeting) are performing.
• Avoid over-indexing on repeat buyers in prospecting. Pairing rc_purchase with nc_purchase makes it obvious when a “Purchase-optimized” campaign is really just retargeting existing customers.

​

Set it up

RC Purchase uses the same four steps as NC Purchase above — wait for the first event, acknowledge it, create a custom conversion, and add it to your reporting. Only two things differ:

• Creating the custom conversion: select RC Purchase as the event (name it e.g. “Upstack RC Purchase”), set the category to Purchase, and leave the value at its default so Meta uses the order value sent with the event.
• Using it: make the RC Purchase custom conversion the optimization event for retargeting and retention ad sets, or add it as a measurement column alongside NC CPA / NC ROAS to see acquisition and retention side by side.

​

Subscription events

For subscription brands, Upstack handles subscription events automatically. No manual setup is required.

​

Advanced configuration

Category-specific events

If you sell across multiple product categories (e.g., apparel and skincare), you can fire purchase events scoped to each category — apparel_purchase, skincare_purchase, and so on — then create a category-specific custom conversion in Meta for each and optimize ad sets per product line.Set these up yourself with global event mappings: add a condition on the product category and map it to a new event name.

Order source exclusions

Not every order should feed back into Meta or Google’s optimization. Some sources would pollute your conversion data.To configure exclusions, go to Settings → Order Source Exclusions and select the sources you want to block:

• TikTok Shop orders
• Loop (returns/exchanges)
• Matrixify (bulk imports)
• Draft Orders
• POS (in-store transactions)

Recurring subscription orders are excluded automatically. Initial subscription orders are sent — these represent genuine new-customer acquisitions.Sources only appear in the exclusion list if you have orders from them. If you don’t see a source listed, you haven’t had any orders from it yet.

Send more data to Facebook, not less. Even TikTok Shop orders help Meta find similar customers — the algorithm benefits from volume. Only exclude a source if the orders don’t represent real customer acquisition (e.g., wholesale imports) or would actively mislead optimization (e.g., POS sales unrelated to your online ad spend).

​

Next step

​

Related