Campaign Setup Handoff — Which Audience Goes In Which Campaign

For: Lauren

From: Mo / GTW

Date: June 13, 2026 (rev. 2 — corrected to Advantage+ best practice)

Account: LG Military Ads (1034037341617540)

What's new

I built a fresh audience layer to replace the 10-month-old seed that was feeding the campaigns. Everything is live, prefixed GTW_. This doc says exactly which audience to add to which ad set and the right settings.

Match rates came back healthy (~45-51%) — strong, clean seeds.

⚠️ Important method note (read first)

Per current Meta best practice for lead gen under the Financial Services category: we are NOT hard-locking lookalikes. We **add the audience as a suggestion inside Advantage+ Audience (leave Advantage+ ON)** and let Meta expand from it. This outperforms the old "lock the LAL" approach — and the account already proves it (the Advantage+ broad ad set is the top performer). Audiences are a guide, not a wall.

Budgets: don't change any daily budgets off this doc — separate conversation with Anthony. Targeting only.

The audiences (all live, prefixed GTW_)

Seeds + Lookalikes (use as Advantage+ suggestions):

GTW_HighValue_Seed_06-13 + Lookalike (1/3/5%) - GTW_HighValue_Seed ← top-payout customers
GTW_Auto_Seed_06-13 + Lookalike (1/3/5%) - GTW_Auto_Seed
GTW_Home_Seed_06-13 + Lookalike (1/3/5%) - GTW_Home_Seed
GTW_Combined_Seed_06-13 + Lookalike (1/3/5%) - GTW_Combined_Seed

Exclusion (every prospecting ad set):

GTW_Mil_AllCustomers_7mo_06-13 (suppression — existing customers)

Warm pool:

GTW_Warm_PageEngagement_365D (~1.6M page/IG engagers)

Per-campaign setup

Age 18-65 + US are locked by Special Ad Category on all. Advantage+ Audience stays ON everywhere — we add the GTW audience in the suggestion/Include box, and always set the suppression exclusion.

🚗 Auto Max Val — `120241310920250498`

Why: this was the weakest campaign on ROAS (its value signal is degraded by the tracking bug). Feeding it our top-payout customers as a signal helps Meta find payers, not just form-fillers.

Ad set	Add as Advantage+ suggestion	Adv+ Audience	Exclude
Broad ( `120241310920230498` )	`GTW_HighValue_Seed` (+ optionally HighValue 1% LAL)	ON	`GTW_Mil_AllCustomers_7mo`

🚗 Auto ALC — `120240808866140498`

Why: this campaign throttled to ~$0/day on stale targeting. A fresh Auto-vertical signal revives it — it has ~$8.5K/day of budget waiting.

Ad set	Add as Advantage+ suggestion	Adv+ Audience	Exclude
Broad 2.O ( `120243776055450498` )	`GTW_Auto_Seed` (+ optionally Auto 1/3% LAL)	ON	`GTW_Mil_AllCustomers_7mo`

🚗 Auto BAU — `120240808864310498`

Why: this is the efficiency engine (1.6-1.7x ROAS) but only spends ~20-55% of its $19.8K budget — it runs out of efficient delivery. Suppression keeps spend on net-new; the warm layer opens a fresh, cheap pool.

Ad set	Setting	Adv+ Audience	Exclude
Broad ( `120240808864580498` )	Keep broad (no suggestion needed)	ON	`GTW_Mil_AllCustomers_7mo`
New: Warm Layer	Include `GTW_Warm_PageEngagement_365D`	ON (suggestion)	`GTW_Mil_AllCustomers_7mo`

🏠 Home Max Val — `120241432784690498`

Ad set	Add as Advantage+ suggestion	Adv+ Audience	Exclude
Broad ( `120241432784700498` )	`GTW_HighValue_Seed`	ON	`GTW_Mil_AllCustomers_7mo`

🏠 Home ALC — `120238737024600498`

Ad set	Add as Advantage+ suggestion	Adv+ Audience	Exclude
Broad ( `120238737024450498` )	`GTW_Home_Seed`	ON	`GTW_Mil_AllCustomers_7mo`

🏠 Home BAU — `120227142278050498`

Ad set	Setting	Adv+ Audience	Exclude
Broad ( `120227144637480498` )	Keep broad	ON	`GTW_Mil_AllCustomers_7mo`
New: Warm Layer	Include `GTW_Warm_PageEngagement_365D`	ON	`GTW_Mil_AllCustomers_7mo`

Account-level segments (already set by Mo — no action)

✅ Existing customers → GTW_Mil_AllCustomers_7mo
✅ Engaged audience → 4 page-engagement audiences

How to add an audience as a suggestion (steps)

Ad set → Audience section → keep Advantage+ Audience ON

In "Audience suggestions / Include these custom audiences" → search the GTW audience name → add it

In the Exclude box → add GTW_Mil_AllCustomers_7mo

Leave everything else default (Meta expands from the suggestion)

Publish

Rollout plan (phased — do it in this order)

Key context: at this account's volume the big Auto ad sets clear Meta's learning phase in hours, not days (Auto BAU ~307 leads/day → ~4 hrs to 50 events; Auto Max Val ~182/day → ~7 hrs). So we don't need a slow week-long stagger — but we DO change things in waves so we can attribute results and protect lead quality.

Phase 0 — Pre-flight (before any change)

Snapshot today's per-campaign CPL + net (baseline to compare against)
Note which audience is currently in each ad set you'll edit (so nothing in-use gets removed by accident)
Do NOT delete any audience yet (the old stale-seed LALs may be feeding live ad sets — deleting an in-use audience pauses that ad set)
Hold all budgets flat (we isolate the audience effect first)

Phase 1 — Day 1 AM: the rescue + revive (most upside, least to lose)

Auto Max Val "Broad" → add GTW_HighValue_Seed as Advantage+ suggestion + suppression exclude
Auto ALC "Broad 2.O" → add GTW_Auto_Seed as suggestion + suppression exclude
Leave the working Auto BAU untouched for now (so we can read Phase 1 cleanly)
These re-stabilize same day given the volume

Phase 2 — Day 2: read, then extend

Next morning, check Phase 1 on CPL AND accept rate / conversion value (not CPL alone)
If it held: extend the same pattern to Home Max Val (GTW_HighValue_Seed) + Home ALC (GTW_Home_Seed)
Add the suppression exclude to Auto BAU + Home BAU (tiny, low-risk change)

Phase 3 — Day 3-4: warm layer (additive, zero disruption)

Build the new Warm Layer ad sets in Auto BAU + Home BAU using GTW_Warm_PageEngagement_365D
These are net-new ad sets — they don't disturb the existing broad ad sets that are working

Phase 4 — Day 5-7: validate, then the budget conversation

After ~3-5 days: assess accept rate / conversion value across all changed campaigns
IF quality held AND efficiency improved → that's when we talk budget with Anthony (Auto BAU + ALC have idle budget + now-fresh audiences = room to scale)
IF accept rate dropped → tighten back toward the seed suggestions, ease off broad

Phase 5 — only after the above is validated: cleanup

Retire the old stale-seed LALs + delete the duplicate suppression uploads (2gtw_..., "Qualified leads")
Verify zero active usage before deleting anything

🚦 The #1 rule throughout — judge on lead QUALITY, not CPL

Advantage+ lowers CPL but can raise the share of unqualified leads. QuoteWizard pays on accepted leads — a cheap CPL with a low accept rate loses money. Track accept rate / conversion value as the primary read at every phase.

Short-term risk (for Anthony)

Learning-phase wobble: minor here — Auto clears in hours, not days. Only tiny Home ALC (~$800/wk) is slow.
Lead-quality drift: the real watch item — monitored via accept rate, with the value-optimized campaigns as guardrail.
Deletion risk: avoided by holding all deletions to Phase 5 and verifying usage first.
Everything is reversible; budgets stay flat through Phase 3 so each change is isolated. Net: short-term transition noise, no irreversible moves.

🛑 Rollback triggers (undo a change if any of these)

Accept rate / conversion value drops materially vs the Phase 0 baseline (3-day rolling)
A changed ad set's CPL stays elevated >24h after it's exited learning
Account net revenue trends down vs baseline over a 3-day rolling window

Questions → Mo. Full audience build + match rates: "Audience Architecture — AS BUILT Jun 13" on the portal.

← Back