An ecommerce team marks a campaign as “acquisition.” Meta recommends excluding existing customers. The media setup becomes easier in minutes, but the harder question moves upstream: what should the excluded customers see next? A first-time buyer who needs reassurance, an owner ready for an accessory, and a lapsed customer considering a return do not have the same creative job.
On October 6, 2026, Meta said Customer Lifecycle Strategy was available to all advertisers. The feature lets marketers identify an acquisition campaign and receive recommendations for excluding existing customers. Meta also said it was still testing a simpler Audience Label that could exclude an entire existing-customer group. The announcement did not publish a country, language or account-level availability list, so teams should verify the control inside their own Ads Manager before building a launch plan around it.
This is more than an interface update. It exposes a procurement problem that performance teams have often hidden inside targeting: “new customer” and “existing customer” are not just audience buckets. They are different commercial assignments. A platform can help you avoid buying some of the wrong impressions. It cannot decide which proof your next customer needs.
What changed—and what did not
The confirmed change is narrow and useful. Meta now lets advertisers express an acquisition intent and recommends ways to exclude existing customers. In the same announcement, Meta described image-to-video in Advantage+ creative as generally available. Ads Creative Studio was beginning a wider rollout, while natural-language audience discovery, the one-label exclusion shortcut and several agentic campaign actions were still tests, betas or future releases.
That distinction matters. Customer Lifecycle Strategy does not mean Meta now operates a complete customer journey for a brand. It removes some audience-setup friction; it does not repair incomplete customer data, define lifecycle economics, choose a message or prove incrementality.
The platform still leaves four decisions with the advertiser:
- Whether the “existing customer” record includes refunds, retail purchases, distributor orders and cross-device buying;
- What a first order, repeat order, cross-sell and reactivation are worth to the business;
- Which claim, demonstration and call to action belong to each state;
- Whether the content created incremental demand or merely collected credit for an order that would have happened anyway.
Google’s September 10, 2026 measurement update described three requirements for AI-era measurement: a strong data foundation, multiple signals and causal proof. That order applies here too. Without a reliable customer state, automation simply executes an incomplete instruction faster.
Why “top-performing creative” is becoming a misleading label
An ad may convert well among existing customers because they already trust the brand. Move it into cold prospecting and the result may not travel. Conversely, a video that explains a category problem and earns a stranger’s trust can be strategically valuable even when its day-one ROAS trails a remarketing ad.
Ranking both in one creative leaderboard creates three common errors:
- Familiarity is mistaken for persuasion. Lower resistance among owners does not prove that a concept can acquire someone new.
- Discount response is mistaken for durable value. A win-back offer may generate one order without restoring normal purchase frequency.
- Attributed conversion is mistaken for incremental demand. A platform receiving credit does not prove the sale depended on that exposure.
If the team has not yet connected product cost, commissions, fulfillment, refunds and repeat behavior, start with a contribution-based creative model. Lifecycle creative cannot be funded sensibly when every stage is judged by the same surface-level ROAS.
The ONCE LIFE card: define the decision before producing the asset
ONCE recommends writing one LIFE card for each lifecycle assignment instead of asking a master brief to “support acquisition and retention.”
L — Label the business state
Define who counts as a prospect, first-time buyer, repeat buyer and lapsed customer—and specify the time window. The definition should reconcile orders, returns, subscriptions, loyalty activity and relevant offline transactions rather than relying only on the purchases visible to one platform.
Data responsibility stays with the advertiser. Meta’s Customer List Custom Audiences Terms require advertisers to have the necessary rights, permissions and lawful basis for customer data and to honor opt-outs. A list being technically uploadable does not make it commercially accurate or legally appropriate.
I — Identify one next decision
Do not begin with the number of deliverables. Begin with the decision a person must make next.
| Lifecycle assignment | Next customer decision | Useful creative evidence | Usually a distraction |
|---|---|---|---|
| Discover | Is this problem or category worth my attention? | Situation, differentiated claim, fast product causality | Loyalty points and complex accessory bundles |
| First purchase | Is this product credible, compatible and low-risk for me? | Demonstration, comparison, sizing, compatibility, delivery and returns | Upgrade language only owners understand |
| Expand | Should I replenish, add, upgrade or use it in another way? | Usage depth, replenishment timing, accessory relationships, new use cases | Repeating the entire brand origin story |
| Return | What has changed enough to justify coming back now? | Product improvements, new use cases, service changes, honest time limits | Re-running the original acquisition discount unchanged |
These assignments can share a production day and a source library. They should not share one definition of success. A single shoot might capture category education for prospects, an operating demonstration for first-time buyers and an upgrade comparison for owners. The edit, CTA, landing page and measurement plan should separate the jobs.
F — Fund the stage by its economics
Acquisition content is normally governed by an affordable new-customer CAC and payback window. Expansion content may be judged by contribution margin, repeat interval or accessory penetration. Reactivation needs to account for discount cost and the possibility that a customer would have returned without paid media. One ROAS threshold cannot value all four assignments.
A practical procurement denominator is fully loaded cost per independently testable lifecycle hypothesis:
(customer-state definition and maintenance + stage brief + modular production and post + rights and localization + measurement readout) ÷ lifecycle hypotheses that can be deployed and read independently.
The denominator is not the number of exported files. A hypothesis counts only when its audience, claim, action and result can be read separately.
E — Equip the assignment with proof
Prospects need a reason to believe; customers need a reason to buy again now. Every LIFE card should name one customer state, one barrier, one core claim, one filmable proof, one CTA, one destination and one stop-or-scale condition.
If creative tools or ad agents help generate, deploy or optimize the work, preserve the source of each claim, usage rights, approval owner and rollback version. ONCE’s creative operating contract for ad agents explains how to keep human authorization attached to faster execution.
When two creative systems are enough—and when four are justified
Not every brand needs a four-stage content factory.
Use a simple prospect / owner split when repurchase cycles are long, accessories and services contribute little revenue, the customer list is small or unstable, or media spend cannot produce readable results for four stages. The first win is to stop paying to reacquire known owners and give those owners useful onboarding, support and referral content.
Use the full four-assignment portfolio when consumables, subscriptions, accessories, upgrades or cross-sell materially affect value; the brand can identify first-order date and recency; and each stage has a different margin, timing or product question. Not every assignment needs paid social. Lifecycle creative can serve email, SMS, the storefront, customer service, retail media and paid advertising.
Delay complex lifecycle execution when consent, deduplication or list-refresh rules are unclear; retail and direct orders cannot be reconciled; or the team has no agreed definition of new-customer rate, contribution margin or payback. Fixing the data and commercial language will create more value than producing dozens of additional variants.
Five actions for the next content commission
- Define prospect, first-order, repeat and lapsed windows from the last 12 months of orders, including refund and cancellation rules.
- Verify that Customer Lifecycle Strategy is visible in the actual ad account; do not treat a global announcement as an account-level guarantee.
- Sample known customers to test the lists, pixel or server events behind exclusion coverage.
- Write one LIFE card per stage. If two cards have the same barrier, proof and action, combine them until the economics justify a split.
- Read results against stage economics, not blended ROAS, and retain a control that does not use the new creative or exclusion logic.
Platforms are making media operations easier. That makes the brand’s responsibility more visible: customer state, commercial intent and creative proof. The next production brief should not begin with “How many videos do we need?” It should begin with “Which customer decision are we trying to change?” Different answers deserve different creative systems.
Sources and evidence dates
- Meta, “Advertising Week New York 2026: Nuevas capacidades de IA para guiar campañas y llegar a los clientes,” October 6, 2026.
- Meta, Customer List Custom Audiences Terms, effective April 25, 2023.
- Google, “Drive profitable growth with new data and measurement tools,” September 10, 2026.
- Mockingbird Digital, “What Meta announced at Advertising Week 2026, and what is actually available,” updated October 9, 2026. Used to cross-check what is live, rolling out, in beta or still a test.
Platform announcements are treated here as confirmed facts about what those companies said or released. The LIFE card, the four-assignment portfolio, the cost denominator and procurement recommendations are ONCE analysis—not promises made by Meta or Google.
