Consolidating Meta Ad Accounts: Why Fewer Campaigns Yield Better AI Learning

The Old Playbook Is Hurting Your Account

Five years ago, the conventional wisdom was: the more granular, the better. Build separate campaigns for each audience segment. Split ad sets by age bracket, gender, interest stack. Keep retargeting strictly away from prospecting. The logic made sense when Meta's algorithm was less sophisticated and advertisers needed to guide it carefully.

That logic is now actively harmful.

Meta's delivery system has shifted from "follow the advertiser's rules" to "optimise toward the conversion signal." The more data points the algorithm has — specifically, the more conversion events per ad set — the better it predicts who will convert next. When you fragment your account into dozens of small campaigns with thin budgets, you starve the algorithm of the data it needs to learn.

Why Fragmentation Kills Performance

The learning phase requires 50 optimisation events (usually purchases, or whatever conversion event you're optimising for) in a 7-day period before an ad set exits learning. If your budget is £30/day and your CPA is £25, you're generating about 1 purchase per day per ad set. At that rate, it takes 50 days to exit learning — by which time your creative is stale and you've likely made changes that reset the clock.

The second problem is auction self-competition. When two ad sets in the same account target overlapping audiences, they bid against each other. You're effectively competing with yourself in the auction, inflating your own CPM. This is audience overlap at the campaign structure level, and it's one of the quietest budget drains in fragmented accounts.

Signs Your Account Is Too Fragmented

Before consolidating, diagnose the problem. Look for:

• Multiple campaigns targeting the same objective with overlapping audiences. Open Ads Manager's audience overlap tool and run it across your active ad sets. Overlaps above 20–30% are a red flag.
• Ad sets with fewer than 5 conversion events per week. These ad sets are chronically in learning and will never optimise effectively at their current budget level.
• 5+ active campaigns with the same campaign objective. Unless you have distinct business reasons for each, this is almost always fragmentation from campaign sprawl rather than intentional structure.
• ROAS varying wildly between campaigns targeting similar audiences. This is often the algorithm competing with itself, not genuine audience quality differences.

Step-by-Step Consolidation Approach

Step 1: Audit What You Have

Map your current campaigns, their audiences, budgets, and 30-day conversion counts. Group campaigns by objective (conversions, traffic, etc.) and audience type (cold prospecting, warm retargeting, existing customer).

Step 2: Migrate to Campaign Budget Optimization (CBO)

Campaign budget optimization lets Meta distribute a single campaign-level budget across your ad sets dynamically, sending more spend to the ad sets that are performing. This is fundamentally more efficient than fixed ad set budgets when you have 2–6 ad sets within a campaign. Migrate each campaign to CBO before consolidating ad sets — it gives you a baseline for performance before you make structural changes.

Step 3: Merge Overlapping Ad Sets

Within each CBO campaign, merge ad sets that target overlapping audiences. Combine their budgets. If you had three interest-based ad sets each at £50/day, merging them into one broad ad set at £150/day gives the algorithm 3× the data signal and eliminates self-competition.

Step 4: Simplify Your Campaign Count

For most DTC accounts spending under £5k/day, the target structure is:

• 1 prospecting campaign (CBO, 3–5 creative variants in a single broad ad set)
• 1 retargeting campaign (CBO, segmented by funnel stage: viewed, add-to-cart, initiate checkout)

That's it. The algorithm handles placement, creative rotation, and audience refinement from there.

What to Watch After Consolidating

Consolidation triggers learning phases. When you merge ad sets or change campaign structure, expect a 1–2 week period of performance volatility as the algorithm recalibrates. During this window:

• Don't make further structural changes. Let the learning phase complete.
• Don't judge performance on day 1–3. Early data is noisy.
• Do monitor for anomalies. If spend is accelerating abnormally or ROAS drops below your floor for more than 3 days, something may have gone wrong beyond normal learning phase variance.

Automated monitoring is valuable here — consolidation followed by unmonitored learning phases is exactly when budget can leak unexpectedly. AdEvolver flags anomalies in real-time via Slack, so you're not flying blind during the transition.

The Long-Term Payoff

Accounts that consolidate consistently and resist the urge to fragment tend to reach CPA targets more reliably over time. The algorithm has more data, less self-competition, and cleaner signals to work with. The role of the media buyer shifts from constant structural tinkering to creative strategy and performance monitoring — which is where human judgment actually adds value.

Get a free ad account audit with AdEvolver.