Video summary

Facebook's NEW UPDATE just changed DTC forever

Main summary

Key takeaways

Business

What changed (Meta / Facebook)

  • Meta launched a Customer Lifecycle Strategy setting (shown as an alert in Ads Manager dashboards).
  • It’s positioned as a delivery strategy decision (not a creative tweak).
  • The intent is to help advertisers acquire incremental/new customers rather than harvest existing demand.

Where to use it (campaign setup)

  • At Ad Set level, select: Customer lifecycle strategy → Acquire new customers

  • This can change delivery and affect:

    • Audience mix
    • CPMs / CPAs
    • Meta’s previous “path of least resistance” tendency (often pushing to people who already know/visited/purchased)

Required configuration to avoid poor results

Exclusions (critical / “not optional”)

Build exclusions using both:

  • Customer list exclusions (your back-end “truth”)
  • Website purchase event exclusions via pixel (Meta “pixel truth”)

Common mistake:

  • Relying on only a single short signal (e.g., 1-week type logic).

Use multiple customer lists based on customer definitions, such as:

  • High-value vs low-value customers
  • One-time buyers vs subscribers / repeat purchasers
  • Refunded / chargeback / disqualified buyers vs “good” buyers

Goal:

  • Prevent Meta from training/executing new-customer campaigns using “terrible data.”

730-day retention window for purchase events

  • Meta allows purchase custom events with a retention window up to 730 days (~2 years) to define what counts as an “existing customer.”
  • Recommendation:
    • Match the window to the business’s real repurchase cycle
  • Examples:
    • Subscription/consumable: shorter window may be appropriate (recent buyers still count as existing)
    • Long repurchase cycle: longer window may be wrong; someone who bought 12–18 months ago might be effectively cold
  • Don’t assume “longer is safer.”

Measurement: the gating factor

The creator frames measurement as the “whole game” because:

  • Performance metrics can move after delivery changes.
  • You must answer: Did the campaign actually acquire a new customer?

Requirements:

  • A single source of truth / dashboard column indicating new customer purchase (incremental/new vs existing).
  • If you can’t measure new vs existing, don’t turn it on yet.

Caution:

  • Using “new vs engaged vs existing” breakdowns may not equal true CAC (it may measure something different).
  • If you optimize for new customers but evaluate with blended CPA, you may get a misleading conclusion.

Framework / playbook: Best practices checklist

  • Maximize prospecting accuracy with exclusions

    • Combine exclusions across wherever possible:
      • customer lists
      • website audiences (purchase events)
      • mobile app audiences
      • campaign/ad-set level exclusions where relevant
    • Use tiered exclusions (not one giant “exclude customers” bucket) to protect against edge cases and data contamination.

  • Signal quality & freshness

    • Keep match scores high: target ≥ 8.5
    • Use strong identifiers (not only email/phone—include mobile app IDs where applicable)
    • Keep customer lists fresh and dynamic (avoid stale lists that become “retroactive reports” rather than truth)

  • Structure customer lists for success

    • Don’t rely on one generic “customers” list.
    • Segment by labels such as:
      • customers vs leads
      • high-value vs low-value
      • one-time vs repeat/subscriber
      • refunded/chargeback/disqualified vs good buyers
    • Use these segments to optimize for different business objectives over time.

  • Measurement

    • Ensure new customer actions are tracked as a column in the ad account.
    • Optionally corroborate with browser/server signals.
    • Use a third-party measurement approach/tool if needed (examples mentioned: Zapier and other measurement partners).
    • Don’t trust “new/engaged/existing” reporting unless it truly maps to new-customer outcomes.

  • Optimize for your most valuable actions

    • Prefer custom conversion events over generic purchase when possible, such as:
      • new customer purchase
      • new customer purchase over $X
      • new customer purchase of a specific product
      • new customer purchase hitting a contribution margin threshold

  • Nuance on exclusions (avoid blunt permanent blocks)

    • Don’t permanently exclude low-value buyers just because they bought once.
    • Instead, exclude only high-value “already-existing” segments and let optimization/measurement promote desired future behaviors.

Strategic implications (business execution lens)

  • Core problem addressed:
    • Separate whether ads are growing the business vs harvesting demand from existing ecosystem users.
  • Intended benefit:
    • Tell Meta lifecycle intent directly (in the creator’s framing):
      • “Go find new people”
      • using your definitions of existing customers
      • optimize toward the most valuable net-new customers
  • Not a magic button:
    • Still depends on data quality, creative, and knowing unit economics/margins
    • “Cheapest customer” can be someone who was already going to buy

Win-back connection (using the same window logic)

  • Use the 730-day purchase window concept to define when someone transitions from “hot/existing” to winback eligible.
  • Purpose:
    • Prevent dilution of new-customer campaigns
    • Create a cleaner, intentional way to bring back past buyers at the right time

Concrete operating guidance by ad spend (tiered recommendation)

The creator provides decision guidance based on spend level, tied to learning/incrementality risk:

  • $100/day or less

    • Set up correctly, but do not optimize for new customers yet
    • Use the new customer conversion event as a dashboard column (manual optimization/creative decisions)

  • $100–$500/day

    • You can test optimizing for new customers, but it’s often still described as “luxury”
    • Adds cost and slower interpretation unless incrementality tracking is rock solid
    • Suggests waiting for maturity if not fully instrumented

  • $1,000–$2,500/day

    • Worthwhile test if:
      • you have robust incrementality measurement across channels
      • you can define the most valuable new customer
      • you can measure it (new-customer metrics)
    • Framed as potentially becoming a strategic lever

  • More than that

    • Worth trying, but creator still suggests not rushing

Key KPIs / metrics explicitly referenced

  • New customer CPA / CAC (primary measurement target)
  • Blended CPA (warned against using for evaluation if optimizing for new customers)
  • LTV (mentioned as impacted by correct exclusions/optimization; “contribution margin and LTV”)
  • CPM / CPA shifts (expected after delivery strategy change)
  • Contribution margin threshold (example of a “most valuable action” to optimize for)
  • Match score target: ≥ 8.5
  • Retention window: up to 730 days
  • Time to stabilize: expect improvement by ~1 month, not day one (learning delay)

Actionable “turn it on / don’t turn it on” logic (checklist summary)

Turn it on only if you can:

  • Configure accurate customer + website purchase exclusions
  • Choose a retention window aligned with your repurchase cycle
  • Track new customer purchase in a reliable ad-account column
  • Measure new customer CPA/CAC (not blended-only)

Otherwise start with:

  • Proper setup + measurement columns + manual optimization until incrementality is trustworthy.

Presenters / sources

  • Presenter: The unnamed YouTube creator (speaking throughout the transcript).
  • Referenced sources: “friends on the Meta engineering and product teams” and internal training docs (no specific individuals named).

Original video