Looking back at Q1 2026
January through March 2026. How much have the three AI systems (Andromeda, GEM, REA) actually shifted what advertisers experience? Three months of observation.
1. Andromeda — the "creative diversity era" is settled
Late 2025 to early 2026: global Andromeda rollout complete.
What advertisers felt:
- Accounts with 1–2 creatives per ad set: CPM up, fatigue spiking
- Accounts with 3–5 creatives: CPA improved or held
- Accounts with 10+ creatives: scaling accelerated
The "load up on one winning creative" strategy is fully over.
2. GEM — AI quality improvements now felt
Q1–Q2 2026: GEM's Q3 architecture upgrade ("2x performance gain from additional data") showed up in the field.
What advertisers felt:
- Advantage+ Sales performance improved 5–15% overall
- Manual campaigns benefited too, but less
- Accounts with high event quality saw bigger gains
Accounts that invested in CAPI and EMQ management have a clear edge.
3. REA — algorithm volatility accelerates
Q1 2026: Ranking Engineer Agent goes public.
What advertisers felt:
- Daily CPA swings widened (±30–40% is normal)
- Weekly-average trend reading is now mandatory
- Accounts that adjusted budgets hastily broke learning and dropped performance
The "daily status check, daily adjustment" style of operator is struggling the most.
Other Q1 changes
Manus × Meta integration beta (March):
- Agencies are interested in report automation
- Direct impact on small advertisers is still limited
Friend Bubbles launch (March):
- Friend-based recommendations in Facebook Reels ads
- Lift for share-driven creative
Policy further tightened:
- More granular AI creative policies
- Better Special Ad Category detection accuracy
- Messenger and WhatsApp ads unified
Q1 performance gaps
Top accounts (top 20%):
- Monthly CPA improved 10–30%
- Advantage+ share 80%+
- 10+ creatives per week
Average accounts (middle 50%):
- CPA held or slightly improved
- Advantage+ share 50–70%
- 3–5 creatives per week
Struggling accounts (bottom 30%):
- CPA worsened (10%+ rise)
- Advantage+ share under 30%
- 1–2 creatives per week
The gap is widening. Adaptation to the AI era is determining performance.
Q1 2026 lessons
1. The cost of hesitating on automation
Accounts with low Advantage+ share struggle most. What people postponed through 2025 can no longer be postponed in 2026.
2. Event quality is decisive
Sequence Learning and GEM learn from event sequences. Poor quality means no AI upside.
3. No daily-level decisions
In the REA era, the minimum unit is a weekly average. Judging daily → daily budget tweaks → learning reset → vicious cycle.
4. Creative is everything
Targeting and bidding are automated, so the remaining lever for advertisers is creative. Without investment here, no differentiation.
Outlook for the remaining 9 months
Announced / expected:
- REA follow-on features (rollout and A/B automation)
- Expanded Advantage+ Creative video capabilities
- Manus integration opening to general advertisers
Advertiser checklist:
- Target Advantage+ 80%+
- Hold event quality EMQ at 8+
- Keep weekly CPA variance within ±20%
So where does that leave us?
Q1 self-assessment:
- [ ] Hit Advantage+ share target?
- [ ] CAPI + EMQ 8+?
- [ ] 5+ creatives per week?
- [ ] AI tooling embedded?
- [ ] Weekly reporting routine?
Q2 priorities:
- Shore up the gaps
- Prepare for Manus integration (agencies)
- Level up the AI creative pipeline
- Long-term brand assets (first-party data)
Operations, metrics, and scaling in the 2026 AI era are covered in Meta Ads Books 4–6.