Fewer campaigns = more money. The new Facebook Ads structure

Last year at LK Media we took over the Facebook Ads account of a lovely fashion-accessories e-shop. Daily spend hovered around €80, and the results? Let's be honest: dismal. The moment we looked under the hood at the account structure, we knew exactly where the problem was buried.

The account was running 14 campaigns. Twenty-three ad sets. Six separate remarketing campaigns — 7-day visitors, 14-day, 30-day, 60-day, abandoned carts, people from Instagram. Acquisition sliced up by interest like an apple into strudel: one campaign for fashion, another for accessories, a third for gifts. And 3–4 ad sets inside each of them.

The budget? Shredded into scraps. The average ad set got roughly €3 a day. Peanuts. At a 2% conversion rate and an average order of €18, that worked out to about 3–4 conversions per week per ad set. Yet Meta needs 50 conversions a week for an ad set to exit the learning phase.

So you can guess the rest. Not one of those ad sets ever made it out of the learning phase. They all glowed permanently orange. The algorithm was hopelessly confused and never really learned who to show the ads to. And the budget went up in smoke.

What did we do? We deleted 11 campaigns. Kept 3. That's it. And the CPA (cost per acquisition) dropped by 34% within the first 14 days.

This is really not the exception. In my experience, most online stores run 3–5× more campaigns in their Meta account than they actually need. And that fragmentation is the number-one reason their ads underperform.

Let's cut to the chase. I'll show you how many campaigns you actually need, when CBO makes sense, and how to build the account so the algorithm can squeeze the most out of it. Ready?

The short version, for those who don't have time to read the whole thing:

Most accounts have too many campaigns. Meta needs 50 conversions per ad set per week to optimize. With 15+ ad sets and a budget of €80/day, you'll simply never reach that number. The fix? Consolidate into 2–3 campaigns.

Consolidation demonstrably lowers costs. The data from Meta Marketing Science doesn't lie: ad sets that exit the learning phase have a 19% lower CPA. Advertisers with less than 20% of spend in the learning phase achieve a 68% lower CPA. Advantage+ consolidation can cut CPA by up to 32%.

The ideal structure for most e-shops: 3 campaigns. (1) Acquisition — broad targeting, CBO, your best creatives. (2) Remarketing — cart abandonment, DPA catalog. (3) Creative testing — a sandbox with 15–25% of the budget. Done.

CBO for scaling, ABO for testing. ABO gives each ad set an equal budget — ideal for like-for-like tests. CBO lets the algorithm shift money to the winners — ideal for performance. The best approach: test in ABO, scale in CBO.

Andromeda needs data, not hand-holding. The more you fragment, the less data each ad set gets. In 2026, Andromeda targets primarily by creative, not by interest groups. Give it room and feed it quality creatives.

Why fewer campaigns = more money

Let's take an example. You have €2,400 a month for Meta. That's roughly €80 a day. A solid base. Now look at these two extremes:

What the data from Meta Marketing Science says:

• Ad sets that exit the learning phase have a 19% lower CPA.
• Advertisers with less than 20% of spend in the learning phase achieve a 68% lower CPA.
• Advantage+ consolidation: up to −32% CPA for e-commerce and lead-gen advertisers.

Why fragmentation hurts — three key reasons:

1. Your data gets scattered. Each ad set needs its own supply of conversion signals. The more ad sets you have, the less data each one gets — and the algorithm flounders.
2. You compete against yourself. Audience overlap between your ad sets is typically 40–60%. Your campaigns bid against one another in the auction and drive up your cost per impression.
3. You never exit the learning phase. Without 50 conversions per ad set per week, Meta can't optimize. Those permanently orange warnings aren't just cosmetic — they mean genuinely worse results.

The learning phase: what it actually is and why it matters

The learning phase is the period when Meta actively tests who, where, and how to show your ad. During it, results are unstable and the CPA tends to be higher. That's normal — the problem is when an ad set never climbs out of it.

Simple math: if your CPA is €12 and you want 50 conversions a week, you need at least €600 a week per ad set (about €86/day). With 10 ad sets, that's €6,000 a week. Most e-shops don't have that — so the only option is to consolidate.

💡 TIP: If you can't optimize for purchases because volume is too low, switch to optimizing for add-to-cart or checkout initiation. These events fire more often and help the algorithm learn who to target.

3 structure models by budget size

Model 1: Minimalist (under €1,200/month) — 2 campaigns

On a small budget, consolidation is absolutely critical. Every euro has to pull its weight — spreading it across many campaigns is suicide.

With this model, lean hard on Advantage+ Shopping campaigns — Meta designed them for exactly this situation: a limited budget and a desire to get the most out of the algorithm. They combine acquisition and remarketing in one and reallocate the budget automatically.

Model 2: Standard (€1,200–4,000/month) — 3 campaigns

This is the sweet spot. You have enough data to separate acquisition from remarketing, but still not so much that you can afford excessive granularity.

The key for this model: in acquisition, don't split by interest. Give Andromeda broad targeting and let it find the right people on its own. In 2026, the algorithm is far better at finding customers based on creative than on your interest segments.

Model 3: Scaling (€4,000+/month) — 3–4 campaigns

With a larger budget you can afford a more sophisticated structure. But be careful — "more sophisticated" doesn't mean "more campaigns." It means adding a strategic layer for testing and scaling.

💡 TIP: When scaling, don't raise the budget by more than 20% at a time. A bigger jump restarts the learning phase. Alternative: duplicate the campaign with a higher budget and run both in parallel — then turn off the one that performs worse.

CBO vs ABO: when to use which

CBO (Campaign Budget Optimization) means you set the budget at the campaign level and Meta decides how much to allocate to each ad set. The algorithm shifts money to whichever ad sets are delivering the best results at any given moment.

ABO (Ad Set Budget Optimization) means each ad set gets exactly what you assign it. No reallocation, full control.

The hybrid approach and workflow in practice:

1. Launch a new creative in an ABO sandbox — each variant gets the same budget (say €8/day) for 7–14 days.
2. Pick the winners by CPA, CTR, and ROAS.
3. Move the winning creatives into CBO acquisition and let the algorithm scale them.
4. Test the next batch of creatives in the sandbox.

This workflow ensures your CBO campaigns are always stocked with proven creatives — and testing never disrupts your performance campaigns.

A 15-minute account audit: 4 steps

Not sure whether your account needs consolidating? These are the questions to ask yourself:

Step 1: Count your campaigns and ad sets

Open Ads Manager. How many active campaigns do you have? How many ad sets? Divide your daily budget by the number of ad sets. If the average ad set gets less than €12–16/day, that's red flag number one.

Red flags:

• More than 5 active campaigns on a budget under €4,000/month
• An average daily budget per ad set below €8
• More than 3 remarketing campaigns

Step 2: Check your learning phase status

In Ads Manager, add the "Delivery" or "Learning Phase" column. What percentage of your spend goes to ad sets with a "Learning" or "Learning Limited" status? If it's more than 20%, you're wasting money.

Red flags:

• More than 30% of spend in the learning phase
• Ad sets older than 30 days that are still glowing orange
• Any ad set in the "Learning Limited" state

Step 3: Check your audience overlap

In Ads Manager, select several ad sets and use the "Audience Overlap" tool. If your ad sets overlap by more than 30–40%, you're competing against yourself in the auction — and paying a higher CPM for it.

Red flags:

• Audience overlap above 40% between acquisition ad sets
• Remarketing without excluding customers from the last 30 days
• Missing exclusion custom audiences

Step 4: Check your creative consistency

How many unique creatives do you have running? How many of them have statistically meaningful data (at least 500 impressions)? If you have 50 creatives and each got only a few hundred impressions, your results are noise — not data.

Red flags:

• More than 5 active creatives in a single ad set
• No creative with 1,000+ impressions in the last 7 days
• Different formats (video, image, carousel) mixed in one ad set

The 5 most common mistakes in Meta account structure

1. Overly granular interest targeting. In 2026, Andromeda doesn't need you to tell it who to target. Your interest groups are smaller than a broad audience and contain fewer conversion signals. The result: more expensive conversions and slower learning.
2. Six remarketing campaigns instead of one. 7-day visitors, 14-day, 30-day, 60-day, cart abandonment, Instagram followers — all in separate campaigns. The result: audience overlap, competing against yourself, and not one campaign reaching 50 conversions a week.
3. Daily tweaks to settings. Every budget change over 20%, every ad set you switch on or off, every edit to targeting — it all restarts the learning phase. Let the campaigns run. At least 7 days without changes after each adjustment.
4. A CPA goal (tCPA) that's too low. Setting a target CPA of €6 on a product whose historical CPA is €12 is counterproductive. The algorithm would rather not spend a cent than exceed your limit. The result: campaigns with minimal spend and zero results.
5. Ignoring Pixel health. A poorly configured Pixel, missing Standard Events, a broken Conversions API (CAPI) — all of it starves Andromeda of data. Without quality signals the algorithm flounders and CPA climbs. Check your Events Manager and CAPI setup before you even touch your campaign structure.

A real restructuring example: before and after

Here are the real numbers from the fashion-accessories e-shop account we mentioned at the start of the article. Results 30 days after consolidation:

The daily budget stayed the same — €80. The only change was the account structure. No new creatives, no bigger investment, no magic. Just fewer campaigns and more data for each of them.

The algorithm was always capable of these results. We just never gave it the chance to learn.

AI tip: let Claude or ChatGPT audit your account

If you want a quick outside perspective, try this prompt. Export your data from Ads Manager (CSV) and paste it into a conversation with an AI:

"You are a Meta Ads expert. Analyze this Facebook Ads account structure and identify: (1) campaigns that probably never left the learning phase, and why; (2) audience overlap problems; (3) consolidation recommendations — how many campaigns and ad sets the account should have at this budget. Here's the data: [paste CSV or enter manually: campaigns, ad sets, daily budgets, CPA, conversions per week, audience overlap %]"

AI won't give you a perfect audit — but it will help you spot the most glaring problems in 5 minutes. Then it's up to you to decide what to do about them.

Conclusion: less really is more

Consolidating your Meta account is one of the most effective moves you can make — and it requires no bigger investment, no new creatives, and no advanced technical know-how. All it takes is the courage to delete what isn't working.

Andromeda is an exceptionally capable algorithm. But it needs room and data. Every unnecessary campaign, every ad set on €3 a day, every interest micro-segment — all of it holds it back. Fewer obstacles = better results.

Start simple. Work out how many ad sets can reach 50 conversions a week at your current budget. Delete or merge the rest. Then let the account run for 4 weeks without changes and watch what happens. In most cases, the results will surprise you.

And if you don't know where to start — get in touch. We'll happily take a look at your account and tell you what to consolidate and how.

Frequently asked questions about Facebook Ads structure

How many campaigns should I have?

For most e-shops, 2–3 campaigns are enough: acquisition (broad targeting), remarketing (cart abandonment + DPA), and optionally a sandbox for creative testing. The smaller the budget, the fewer campaigns — otherwise each one gets too little data to optimize.

Does separate retargeting still work?

On smaller budgets (under €1,200/month), separate retargeting usually doesn't get enough data and stays stuck in the learning phase. Since 2023, Meta has recommended merging acquisition and remarketing into a single CBO campaign with Advantage+ Audience. On larger budgets above €4,000/month, a standalone remarketing campaign still makes sense.

Should I switch from ABO to CBO?

For scaling, yes — CBO lets the algorithm reallocate budget to the winning ad sets. Keep ABO for creative testing, where you need to guarantee each variant even delivery. The ideal workflow: test in ABO, scale the winners in CBO.

How many ad sets per campaign?

The rule: no more ad sets than can each reach 50 conversions a week. On a daily budget of €80 and a CPA of €8, that's roughly 2–3 ad sets. Better fewer ad sets with more creatives inside than many ad sets on tiny budgets.

How often should I refresh creatives?

Watch frequency and CTR. When frequency exceeds 3–4 and CTR starts to drop, it's time for new creatives. For smaller audiences under 500,000 people, that can be every 2–3 weeks. For broad targeting, 6–8 weeks is fine. Reactivate older creatives — after a while Andromeda treats them as new.