If your Meta account has 10 campaigns, 30 ad sets, and performance that's been flat for months, the problem isn't the creative. It's the structure. The way most accounts were built before 2024 actively sabotages the algorithm that now controls your results.
In this post:
- Why fragmented campaign structures starve Meta's algorithm
- The math that shows exactly when you have too many ad sets
- A concrete before/after account structure
- Why budget moves from audience segmentation to creative variation
- How to restructure without triggering a cascade of learning resets
Why More Campaigns Break the Algorithm
Meta's delivery system runs on signal. It needs roughly 50 optimization events per ad set per week to exit the learning phase and stabilize delivery. In 2026, Meta lowered that threshold to 25 events per week for Advantage+ Sales Campaigns — but the principle is unchanged: each ad set needs a minimum volume of conversions to learn who buys and when.
Spread budget across 15 ad sets and you divide that signal pool. Each ad set gets a fraction of the events it needs. Every ad set sits permanently in "learning limited" status. Delivery is unstable. Meta is guessing on every impression.
The Andromeda algorithm update — fully rolled out in late 2025 — made this worse. Andromeda introduced a 10,000x increase in model complexity at the ad retrieval stage. It now uses computer vision and semantic analysis to read your creative assets and match ads to users. Creative content is a primary targeting signal. Fragmented ad sets split that signal across the system instead of concentrating it where it compounds.
The Math Behind Fragmentation
The learning phase threshold makes the problem concrete.
Take an account spending $6,000/month ($200/day) across 10 ad sets. That's $20/day per ad set. At a $40 target CPA, each ad set generates roughly 15 sales per week — well below the 50-event threshold. Every ad set is starved.
Consolidate to 2 ad sets at $100/day each: each generates roughly 75 sales per week. Both exit the learning phase. The algorithm stabilizes. Same budget. Entirely different outcome.
The formula for minimum daily spend per ad set:
(Target CPA × 50) ÷ 7 = minimum daily budget
At a $30 CPA, that's $214/day per ad set. At $50, it's $357/day. Most accounts run far below these thresholds per individual ad set — which is why they're learning-limited everywhere.
A controlled test from Five Nine Strategy makes this concrete. Same budget, same creatives, two structures: one campaign with one ad set and 25 creative assets versus one campaign with five ad sets and five creatives each. The consolidated structure delivered 17% more conversions at 16% lower cost per conversion. No new creative. No new audience. Just fewer ad sets.
The Fewer Campaigns Framework
The 2026 structure for most accounts is three campaigns, six to eight ad sets, and far fewer active ads than you're running now.
| Old Structure | 2026 Framework | |
|---|---|---|
| Campaigns | 10–15 | 3 |
| Ad sets | 30–40+ | 6–8 |
| Active ads | 100+ | 25 + 5 new tests/week |
| Audience approach | Interest/lookalike stacks | Broad with Advantage+ Audience controls |
| Budget method | Fragmented by segment | Concentrated, creative-led |
Campaign 1 — Advantage+ Sales Campaign (70% of budget, CBO)
One broad ad set. Advantage+ Audience with only hard controls set: geography, minimum age, and necessary exclusions. No interest stacks. No lookalike layers. 10–20 diverse creative assets — video, UGC, static, catalog, testimonial — give the algorithm material to work with. Let it find the buyers.
Campaign 2 — Manual Retargeting (20% of budget, ABO)
Two ad sets: 7-day website visitors and add-to-cart; 30-day visitors and checkout initiators. Tight audience definition makes sense here — you're targeting people with demonstrated intent, where first-party signal already exists.
Campaign 3 — Creative Testing (10% of budget, ABO)
One ad set, $50/day, broad targeting. New creative concepts only. Any ad that hits your target ROAS over 7 days at $50/day graduates to Campaign 1. The rest pause.
bulk can audit your current structure against this blueprint and flag exactly which ad sets to consolidate — before you touch anything live.
A DTC apparel brand documented by The Interconnections collapsed from 16 ad sets across 5 campaigns to 2 Advantage+ campaigns. CPA dropped 23% within six weeks. No creative changes.
Creative Variation Is the New Audience Targeting
The structural shift matters because Andromeda changed what the algorithm optimizes on. Budget previously spent on audience segmentation now funds creative variation. The algorithm does the audience work. Your job is to give it diverse assets to test.
Ten near-identical product shots count as one signal to Andromeda. Ten genuinely different concepts — different formats, different angles, different hooks — give it ten distinct signals to route. The accounts that adapted earliest now run roughly 30–40% more live creative assets than pre-2024, with far fewer ad sets.
This is a direct read-across to how Advantage+ automation is changing daily media buying decisions and why bidding strategy choices interact so tightly with account structure. Audience controls, creative volume, and budget concentration are now a single system — pulling one lever affects all three.
How to Restructure Without Triggering Learning Resets
Any structural change — new ad set, modified targeting, new campaign — triggers a new learning phase. Done wrong, a restructure can reset the signal you've spent weeks accumulating.
The sequence:
- Audit first. Pull 90-day performance by ad set. Flag every ad set generating fewer than 50 conversions per week at current spend. That's your consolidation list.
- Pause, don't delete. Pausing doesn't reset campaign-level learning or erase historical data. Deleting does.
- Build new structure in parallel. Launch the consolidated campaign alongside the old structure at low spend for 7 days before scaling.
- Graduate winning creatives. Take the 5–7 best-performing creatives from the old structure and seed them into the new consolidated campaign.
bulk connects to your Meta account, reads your live performance data, and walks through the entire restructuring plan before touching anything. It identifies which ad sets are learning-limited, which creatives qualify for graduation, and what the consolidated structure should look like — you approve the plan before any live ad changes.
Meta's Ads AI Connectors, launched in open beta in April 2026, expose 29 Marketing API functions to any MCP-compatible agent — including campaign creation and budget reallocation, all created in paused state pending human approval. Account restructuring now has a viable agent-driven workflow with a built-in safety layer that makes the process auditable at every step.
bulk handles the audit, consolidation plan, and execution for Meta ads teams — with your approval before anything goes live. Try bulk free →