Meta Ad Account Structure for 2027: How to Consolidate After Andromeda

Meta ad account structure should start with fewer campaigns, broader eligible audiences, and creative tests that can earn meaningful delivery. For 2027 planning, use 1 to 3 campaigns per objective as a starting hypothesis, not a Meta requirement, and add divisions only when economics or operating constraints justify them. Andromeda improves ad retrieval, but it does not remove the need for reliable conversion signals, bidding decisions, or margin controls.

Quick answer: Start with one core campaign per objective and only add testing, retention, or market divisions when they serve different business decisions. Consolidate overlapping ad sets so budgets are not spread across too many learning units. Judge the restructure on new-customer acquisition cost and contribution, not campaign count or a promised algorithm lift.

What is the best Meta ad account structure in 2027?

The best starting structure is 1 to 3 campaigns per objective, with a small number of ad sets sharing compatible conversion events, markets, and economics. This is planning guidance based on current product capabilities, not a forecast of an unpublished 2027 platform rule.

For a single-market DTC shop, start with one sales campaign and one ad set. Add a controlled creative test or retention campaign only when you need an independent budget or audience boundary. Leads and purchases need separate success definitions, even when both contribute to the same business.

Record net revenue, acquisition cost, contribution, attribution settings, and returning-customer share before changing anything. A cleaner account that buys more repeat customers can appear efficient while producing less incremental demand. Our Meta Ads agency approach starts with that measurement question.

Key takeaway: Consolidate duplicate decisions, not objectives that require different outcomes.

Why did Andromeda change how Meta accounts should be built?

Andromeda strengthens personalized ad retrieval, making diverse creative and sensible delivery boundaries more useful than endless audience copies. Meta's engineering explanation describes a first-stage system that reduces tens of millions of candidate ads to a few thousand before later ranking stages.

Retrieval decides which ads become candidates; ranking and the auction still matter afterward. Read Meta Engineering's Andromeda retrieval explanation alongside our Andromeda and GEM overview. The engineering source supports the architecture, not a universal campaign recipe or guaranteed ROAS increase.

Fragmentation creates a separate problem: each ad set receives less budget and fewer optimization events. Example math: 140 weekly purchases divided evenly across 14 ad sets means 10 events each; two ad sets would average 70. Actual allocation is uneven, and neither arithmetic nor consolidation guarantees learning completion.

Key takeaway: Andromeda changes candidate retrieval; consolidation addresses how thinly you divide delivery and evidence.

How many campaigns and ad sets does a Meta account need?

A Meta account needs only enough campaigns and ad sets to express genuinely different business decisions at its available conversion volume. Use these monthly media tiers as proposed starting ranges, not industry benchmarks or platform limits.

Starting counts by monthly media spend

Under USD 10,000: 1 core campaign with 1 to 2 ad sets; test new concepts inside it unless a separate experiment is affordable. USD 10,000 to 50,000: 1 core plus 1 testing campaign, generally 1 to 3 ad sets in each. USD 50,000 to 150,000: 2 to 3 campaigns, with 1 to 4 ad sets per campaign where market or product boundaries justify them. Above USD 150,000: start with 3 to 5 across distinct business needs, not automatic duplication by interests.

These totals assume one main sales objective. More objectives or legally separate markets can require more campaigns. GBP and AUD accounts should use their own CPA and conversion volume, not pretend these USD brackets are fixed local equivalents.

Example math: GBP 6,000 monthly media at GBP 40 CPA supports 150 expected purchases; AUD 12,000 at AUD 80 supports the same count. Both may justify a simpler setup than a higher-spend account with expensive conversions. Our guide to scaling beyond USD 100,000 monthly covers the operating constraints at larger budgets.

Key takeaway: Count expected conversions per learning unit before counting campaigns.

How should you split prospecting and retargeting now?

Keep prospecting and retargeting separate only when you need measurable customer exclusions, different offers, or controlled retention spending. Otherwise, broader Advantage+ audiences can make parallel warm and cold copies redundant rather than automatically additive.

A proposed starting allocation is 80 to 90 percent acquisition and 10 to 20 percent dedicated retention, but zero dedicated retargeting is also valid. These shares are operating heuristics, not Meta recommendations. Check whether your actual campaign configuration permits the customer controls you need before promising an acquisition-only result.

A small remarketing pool can saturate quickly. Track customer status, purchase recency, frequency, and total account results. High reported retargeting ROAS does not prove incremental sales because some purchasers would have returned without ads.

For a subscription renewal or replacement part, a separate message can be justified. For one evergreen offer, test the combined approach first. Meta's Advantage+ automation overview explains the available automation rather than mandating a fixed funnel split.

Key takeaway: Separate customer groups for control and economics, not because every funnel diagram has two boxes.

How do you structure creative testing inside a consolidated account?

Structure creative testing around a written hypothesis and a delivery budget, using a separate campaign only when the core campaign cannot give the experiment a fair read. A proposed 80 percent core and 20 percent testing split keeps most media on established delivery while reserving learning capacity.

Example math: USD 30,000 monthly media means USD 24,000 core and USD 6,000 testing. That USD 6,000 is media, not production money. Four test batches would receive USD 1,500 each before any reallocation.

Test distinct concepts such as demonstration, objection handling, customer evidence, and product comparison. Five headline changes on the same footage are not five independent concepts. Where causal comparison matters, use an available randomized experiment rather than calling unequal delivery an A/B test.

Graduate creative after an agreed spend ceiling, conversion-lag window, and quality check, not after one purchase. Require a plausible acquisition cost plus acceptable contribution and customer mix. Reusing an existing post ID can preserve engagement where supported, but does not transfer an ad set's learning state. Our creative testing framework supplies the hypothesis and review loop.

Key takeaway: Give each experiment a reason, a budget, and a decision date before launching it.

When does a fragmented structure still beat consolidation?

A divided structure beats consolidation when separate budgets protect constraints that shared delivery cannot express reliably. Four common reasons are geography, margin tiers, compliance categories, and offer-specific funnels.

US, UK, and Australian campaigns may need different fulfillment promises, currencies, language, and legal wording. A USD 35 acquisition ceiling for one product and USD 90 for another are example thresholds that should not be averaged blindly. Separate financing or regulated offers where platform policies require different settings.

Business divisions deserve evidence. Two countries sharing identical economics do not automatically require two campaigns; one country with incompatible offers may require several. Document which decision each split enables and what evidence would let you remove it.

Our paid media services connect those boundaries to creative and outcome reporting. Consolidation is not a substitute for understanding the offer.

Key takeaway: Keep a split when it protects a real constraint; remove it when it only preserves an old naming convention.

What does a before and after account restructure look like?

A useful restructure replaces overlapping campaigns with fewer accountable budget pools while preserving real market and offer differences. The following 60-day DTC account is entirely illustrative, not a client case study or a promised performance improvement.

Assume one US store, one purchase objective, USD 30,000 monthly media, and identical inventory throughout. Example math distributes the same budget before and after:

Days 1 to 14 establish the baseline and validate tracking. Days 15 to 30 introduce the new pools gradually. Days 31 to 60 review mature purchase cohorts, refunds, contribution, and new-customer share. Twelve to two is a reduction of ten campaigns, not evidence of lower CPA.

Hold promotional conditions as comparable as possible and log stockouts, offer changes, and spend shifts. A simultaneous discount makes the restructure's impact harder to isolate.

Key takeaway: A credible before-and-after example reconciles spending and states what it cannot prove.

How do you migrate without resetting learning?

You cannot guarantee a migration without learning resets; instead, protect business continuity while allowing new or significantly edited ad sets to learn. Treat 20 percent as a cautious budgeting heuristic, not a universally safe Meta threshold.

Meta's learning phase guidance explains delivery exploration, while its significant edits guidance covers changes that can trigger renewed learning. Check current delivery status rather than assuming a particular percentage avoids it.

1. Export 14 to 28 days of results, covering conversion lag and the weekly sales cycle.
2. Verify purchase value, consent handling, deduplication, exclusions, and attribution settings.
3. Build or duplicate only the necessary destination campaigns; copied ad sets learn anew.
4. Introduce a limited budget while retaining a viable original pool, with overlap kept temporary.
5. Shift spending in documented stages, checking mature outcomes before the next change.
6. Phase out redundant units and compare account-level contribution, not just the new campaign's ROAS.

Example math: a 20 percent increase moves USD 500 daily to USD 600. It is not a guaranteed safe edit. Avoid stacking budget, creative, audience, and bid changes, and do not restructure immediately before a major promotion unless the existing setup is broken.

Key takeaway: Migration manages risk and attribution; it does not carry learning history into a new ad set.

What is the 7 point Meta account structure checklist?

The seven-point checklist tests whether every division has a business purpose and enough evidence to support its budget. Review it weekly during migration, then at an agreed monthly operating review.

1. Objective: Does each campaign optimize for one defined business outcome?
2. Economics: Are CPA ceilings tied to contribution, refunds, and customer status?
3. Volume: Can each ad set receive enough spend for a useful decision?
4. Audience: Are geography, customer controls, and compliance settings verified?
5. Creative: Does every test have a distinct hypothesis, budget, and stop condition?
6. Measurement: Are conversion events reliable and account totals reconciled with sales?
7. Change control: Is every significant edit logged with its date and review window?

A pass requires evidence, not seven checked boxes. If the answer to volume or measurement is unclear, fix that before adding another audience segment. Assign one owner to approve structural changes so several operators do not reset the baseline independently.

Key takeaway: The best account is the smallest one that expresses your real constraints and supports defensible decisions.

FAQ: what should you check before consolidating?

Check conversion volume, customer boundaries, and a full purchase cycle before consolidating; these three answers summarize the main decisions.

How many ad sets should a Meta campaign have?

Start with 1 to 3 ad sets when they share compatible economics and conversion goals. That is planning guidance, not a Meta limit. Add an ad set only for a meaningful budget, market, or offer distinction. A small account with expensive purchases may need just one. More spend can support more tests, but volume alone does not justify overlapping audience copies.

Should I still separate prospecting and retargeting?

Separate them when customer controls, retention offers, or different acquisition ceilings require independent spending. Otherwise, test a combined approach and report new versus returning customers. A 10 to 20 percent retention share is only a starting heuristic, not a mandatory allocation. Judge the total account's contribution and incremental demand rather than assuming high retargeting ROAS means additional sales.

How long should I wait before restructuring a Meta account?

Use 14 to 28 days as a proposed baseline window when it covers normal weekly variation and conversion lag. Longer buying cycles need longer observation. Do not wait if tracking or compliance is broken. After migration, give new delivery time to produce mature outcomes and avoid repeated edits. No calendar window guarantees stable learning or proves that a restructure caused better performance.