For years, a complicated Meta Ads account could look like a well-managed one.
There were separate campaigns for prospecting and remarketing. Different interests lived in different ad sets. Lookalikes were divided by percentage. New campaigns appeared for products, promotions, creative tests, and each theoretical stage of the funnel.
The structure looked organised.
But organisation inside Ads Manager is not the same thing as useful control.
Meta's delivery systems can now evaluate potential customers, creative, placements, and conversion signals at a scale advertisers cannot reproduce manually. Systems such as Andromeda are part of that wider shift.
This has reduced the need to divide accounts simply because we can.
It does not mean every business should run one campaign, one ad set, and one broad audience.
The better rule is:
Every campaign and ad set should earn its place.
If a separation gives the business better control, better learning, or better reporting, it may be useful.
If it only creates another box in Ads Manager, it may be making the account harder to manage without improving the result.
Start with the business, not an account template
There is no universal Meta Ads structure for eCommerce.
A store selling three products in one country has different requirements from a retailer operating across several markets, currencies, languages, product categories, and fulfilment models.
Before deciding how many campaigns or ad sets to create, we look at questions such as:
- What is the daily and monthly budget?
- How many products or meaningful product groups exist?
- Are margins similar across those products?
- Do different markets need different languages or offers?
- Is the main goal new-customer acquisition, repeat purchases, or both?
- How much reliable conversion data is available?
- How much new creative can the team produce?
- Are there important stock or fulfilment constraints?
- Does the business run large seasonal promotions?
Two stores can sell similar products and still need different structures.
One may want Meta to move budget freely toward whichever product can generate the best acquisition economics.
Another may need guaranteed investment in a new category even if an established bestseller converts more easily.
The second account needs more budget control.
The first may benefit from more automation.
What consolidation actually means
Consolidation is often reduced to “use fewer campaigns.”
That misses the point.
Consolidation means removing divisions that no longer represent a meaningful decision.
Imagine five ad sets that:
- Optimise for Purchase
- Promote the same products
- Use almost identical creative
- Reach heavily overlapping potential customers
- Have no separate commercial target
Their names may be different.
Their jobs are not.
Splitting them can divide budget and optimisation events across several ad sets without giving the business much additional information.
When we review that type of structure, we ask:
What decision becomes possible because this ad set exists separately?
If there is no clear answer, consolidation deserves consideration.
But consolidation can go too far
Putting everything together can also remove useful control.
That becomes a problem when the combined parts have genuinely different economics or objectives.
Examples include:
- Countries with very different acquisition costs
- Products with very different margins
- Different languages
- Different inventory constraints
- New-customer acquisition and retention goals
- A promotion with a fixed budget
- A product launch that needs guaranteed delivery
The goal is not the smallest account possible.
It is the simplest structure that still protects the decisions the business needs to make.
CBO and ABO are really budget-control decisions
CBO and ABO are still widely used terms, although only one is the current Meta product name.
CBO refers to what Meta now calls Advantage+ campaign budget. The advertiser sets a campaign-level budget, and Meta distributes spend across the ad sets inside that campaign based on where it predicts better opportunities.
ABO is industry shorthand for using individual ad set budgets. Each ad set receives its own budget instead of sharing one campaign-level pool.
Neither is automatically better.
| Advantage+ campaign budget (CBO) | Ad set budgets (ABO) |
|---|---|
| One budget at campaign level | Separate budget for each ad set |
| Meta decides how to distribute spend | The advertiser decides the allocation |
| Useful when ad sets share one overall commercial objective | Useful when individual ad sets need guaranteed spend |
| Can concentrate spend heavily on one opportunity | Can keep spending on weaker opportunities because the budget is protected |
| Requires less manual reallocation | Requires more deliberate budget management |
The useful question is not:
“Is CBO better than ABO?”
It is:
“Should Meta decide how this budget is distributed, or does the business need to decide?”
When Advantage+ campaign budget makes sense
A campaign-level budget makes the most sense when the ad sets are working toward the same overall result and the advertiser is comfortable allowing Meta to move spend toward the opportunities it predicts will perform best.
For example, a sales campaign may contain several product or creative themes that can legitimately compete for the same acquisition budget.
If one produces better results, letting Meta allocate more spend there can reduce unnecessary manual budget changes.
Meta also provides ad set spending controls in eligible campaign-budget setups when advertisers need some additional boundaries.
But those controls should have a reason.
If every ad set receives tightly defined minimums and maximums, we are gradually rebuilding an ad-set-budget structure while still calling it CBO.
Do not confuse campaign budget with shared learning
This distinction matters.
Advantage+ campaign budget can move money between ad sets.
The learning phase still happens at the ad-set level.
Three ad sets inside one campaign do not automatically combine their optimisation events into one learning pool.
If budget is spread across too many ad sets, some may still receive too little activity to produce stable delivery.
Campaign-level budgeting can help Meta direct more spend toward promising ad sets, but it does not remove the cost of unnecessary fragmentation.
When ad set budgets make sense
Ad set budgets give the advertiser more control over delivery.
That is useful when the business needs a particular segment to receive a defined amount of spend regardless of whether Meta would naturally prioritise it.
Examples can include:
- A new product that needs a controlled launch budget
- A market with its own commercial target
- A creative test where each concept needs enough delivery to evaluate
- A promotion with a fixed allocation
- A strategically important product group
- A customer segment that must be evaluated separately
The trade-off is simple.
With an ad set budget, the advertiser can force spend.
That means the advertiser can also force spend onto a weak opportunity.
ABO therefore requires a clear answer to:
Why are we protecting this budget?
“Because Meta was not spending enough on it” is not always sufficient.
Maybe Meta found a better opportunity.
Or maybe the new ad set genuinely needs more delivery before it can be evaluated.
The business context decides which explanation matters.
CBO is not for small budgets and ABO is not for big budgets
Budget size alone should not determine the choice.
A large advertiser may use Advantage+ campaign budget because it wants Meta to allocate a substantial budget dynamically across proven opportunities.
A smaller advertiser may deliberately use ad set budgets because two product groups need separate spending guarantees.
Likewise, a large creative-testing programme might use controlled ad set budgets for testing and campaign-level budgets elsewhere for scaling.
The structure should follow the decision that needs to be made.
The learning phase is useful context, not a profitability score
The Meta learning phase gets more attention than it deserves.
It is important, but it is often interpreted incorrectly.
After an ad set launches or receives a significant edit, Meta needs time and optimisation events to learn how to deliver it more consistently.
Meta's current guidance still uses approximately 50 optimisation events within the week after the last significant edit as the typical point where delivery can stabilise.
If Meta expects that an ad set is unlikely to generate enough events, Ads Manager may show Learning Limited.
That is useful operational information.
It is not a statement about whether the campaign makes money.
An ad set can generate 20 purchases per week and be highly profitable.
Another can generate 100 and lose money on every sale.
Meta's learning status describes signal volume and delivery stability. It does not understand your margin requirement.
Do not build the account around the number 50
The approximate 50-event guidance can help explain why splitting a small amount of conversion volume across ten ad sets is usually difficult.
It should not become a rule that says every ad set deserves unlimited budget until it reaches 50 events.
Suppose a product normally converts at 2% and the ad set has already generated hundreds of relevant visits with no sales.
The manager should not ignore that evidence simply because Meta has not completed its preferred learning volume.
Likewise, a high-price product with excellent economics may never generate 50 purchases every week.
That does not automatically mean advertising it is a mistake.
Use the learning phase as one diagnostic signal alongside:
- Spend
- CPA
- Conversion rate
- Revenue
- Margin
- Customer quality
- The amount of evidence collected
Do not protect learning from every change
Significant edits can move an ad set back into learning.
This is why constant reactive editing can make performance harder to interpret.
But the opposite behaviour can be just as damaging.
If the offer is wrong, tracking is broken, or the campaign is spending heavily on an obvious problem, keeping everything unchanged because the ad set is “still learning” is not good management.
The useful principle is:
avoid unnecessary changes, not necessary ones.
Before editing, know what problem the change is supposed to solve.
After editing, give the result enough time and data to evaluate where the budget allows it.
What broad targeting means in 2026
Broad targeting does not literally mean that Meta shows the ad to everybody.
Advantage+ audience separates audience inputs into two useful categories:
controls and suggestions.
Audience controls
Controls are the boundaries Meta should respect.
Depending on the campaign setup, these can include things such as:
- Location
- Minimum age
- Language
- Custom-audience exclusions
These are useful for real business restrictions.
If the store ships only to Sweden, that is a boundary.
If a product cannot legally be advertised below a certain age, that is a boundary.
If existing customers need to be excluded from a particular acquisition strategy, that can also be a meaningful control.
Audience suggestions
Other inputs can act more like guidance.
Advantage+ audience can use suggestions such as:
- Custom Audiences
- Lookalike Audiences
- Detailed targeting
- Age-range preferences
- Gender
Meta can then expand beyond those suggestions when it predicts that other people may produce the desired result.
This is very different from treating an interest group as a hard wall.
Broad does not mean targeting strategy disappeared
The role of targeting has changed.
More of the prediction can happen through Meta's delivery systems, conversion history, customer data, and the ads themselves.
But the advertiser still needs to decide:
- Which countries make commercial sense
- Which products can actually be delivered there
- Which customers should be included or excluded
- Which conversion event Meta should optimise towards
- Which markets need separate economics
- Which creative ideas Meta has available to match with potential customers
This is why “creative is the new targeting” is useful as a shorthand but inaccurate as a literal statement.
Creative has become more important as Meta's recommendation systems improve.
It has not replaced product, market, customer, and eligibility decisions.
Broad audiences need creative range
A broad audience is not automatically a strong campaign.
Meta still needs something useful to show the people it finds.
If every active ad communicates the same product benefit in slightly different colours, the delivery system has limited creative range.
A more useful library may contain different:
- Customer problems
- Benefits
- Use cases
- Objections
- Proof points
- Creative concepts
One ad may demonstrate the product.
Another may explain a customer problem.
Another may compare the product with an alternative.
A fourth might focus on an objection that prevents people from buying.
Broad delivery gives Meta room to identify different potential customers.
Creative diversity gives it different arguments to match with them.
Creative testing does not automatically require a new ad set
This is one place where account structures can grow too quickly.
Every new headline, visual, creator, or hook does not need its own ad set.
If the goal is simply to introduce another creative option and allow Meta to optimise delivery, the ad can often live inside an existing relevant ad set.
A separate testing structure makes more sense when the team needs controlled delivery to answer a specific question.
For example:
Does concept A outperform concept B when both receive enough independent spend to evaluate?
That is a different objective from:
Which of these ads should Meta use to generate the most purchases?
The first is an experiment.
The second is normal campaign optimisation.
Do not create extra account structure until you know which question you are trying to answer.
Product differences can justify separation
Consolidation becomes dangerous when it hides product economics.
Products can differ in:
- Margin
- Price
- Return rate
- Stock
- Seasonality
- Conversion rate
- Customer type
- Repeat-purchase behaviour
- Creative needs
If several product groups share one budget, Meta may naturally favour the products that generate the optimisation event most easily.
That can be exactly what the business wants.
Or it can mean most of the budget flows toward a low-margin bestseller while a strategically important category receives very little delivery.
This is where campaign structure becomes a commercial control.
Separation may happen through campaigns, ad sets, catalogue product sets, or other controls depending on the setup.
The important part is that the business can still answer:
Which products are Meta scaling, and do we want more of those sales?
Prospecting and remarketing do not need three funnel campaigns by default
The traditional Meta Ads structure often created separate TOF, MOF, and BOF campaigns for the top, middle, and bottom of the funnel.
That model looks tidy on a diagram.
Real customer journeys are less tidy.
Someone may see one ad and buy immediately.
Another customer may watch several videos, visit twice, search for the brand on Google, and return weeks later.
Meta's delivery systems already use behavioural and conversion signals when predicting who may respond to an ad.
That reduces the need to reproduce every theoretical stage of the funnel through a separate campaign.
When remarketing still deserves its own structure
Remarketing can still make sense when it has a genuinely different job.
For example:
- A different offer
- A different creative message
- A specific product reminder
- A defined retention strategy
- A controlled budget
What we would avoid is creating a remarketing campaign solely because every account template says one must exist.
Customers already close to purchasing can make remarketing performance look extremely efficient.
If too much budget is protected there, the account may report strong short-term ROAS while doing less work to acquire new customers.
How to recognise an overbuilt account
We become suspicious of fragmentation when:
- Several campaigns have almost no meaningful daily spend
- Many ad sets optimise for the same result and reach similar people
- The same creatives appear across several audience segments
- Most ad sets produce too little data to evaluate
- A new campaign is created for every creative batch
- Budgets are manually shifted every day
- The team cannot explain what a campaign is uniquely responsible for
A busy Ads Manager is not evidence of active strategy.
Sometimes it is evidence that the strategy has been divided into too many pieces.
How to recognise an account that is too simple
Consolidation can also create warning signs.
For example:
- One country consumes the budget while other important markets disappear
- One product group receives nearly all delivery
- High- and low-margin products are judged through one blended ROAS
- New ideas never receive enough delivery to evaluate
- A strategic product launch is starved by an established bestseller
- New- and returning-customer results are blended despite different targets
- The overall result looks strong while important categories lose money
At that point, the account has gained simplicity by sacrificing information or control the business actually needs.
A decision framework for account structure
When deciding whether something deserves its own campaign or ad set, we use three questions.
| Question | If the answer is yes |
|---|---|
| Does it need its own budget? | A separate ad set or campaign may be justified |
| Does it need different optimisation or business rules? | Separation may protect an important commercial objective |
| Do we need independent reporting or a controlled test? | A separate structure may make the result easier to interpret |
If the answer to all three is no, creating another campaign or ad set may not add much value.
How we would structure a smaller eCommerce account
We would normally start simpler than many older Meta Ads playbooks recommend.
A store with one market, a focused catalogue, and a moderate budget may not need separate campaigns for every audience stage.
A starting structure could use:
- One primary Sales campaign
- A broad or Advantage+ audience within the real business boundaries
- A small number of ad sets only where product or testing differences justify them
- Several meaningfully different creative concepts
- Reliable Purchase tracking
From there, complexity is added only when performance data or a business requirement creates a reason.
How the structure can change as the business grows
No account structure should be considered permanent.
A business may eventually need separate structures for:
- New markets
- High-priority product categories
- Customer acquisition
- Retention
- Creative testing
- Large seasonal promotions
- Products with different economics
Those additions make sense because the business changed.
They should not be added simply because a diagram of a “mature” account contains more boxes.
What we review before restructuring an account
When we audit a Meta Ads account, we do not begin by counting campaigns.
We look at the relationship between structure and actual business performance.
Budget distribution
Where is the money really going?
Does that match the business priorities?
Optimisation volume
Are several ad sets competing for too little conversion activity?
Creative overlap
Are supposedly different audiences seeing essentially the same ads?
Audience logic
Which settings are real controls, and which are only suggestions?
Product economics
Is Meta favouring products that look efficient in-platform but produce weaker margins?
Customer mix
How much performance comes from new versus existing customers?
Measurement
Can we trust the Purchase event and the values used for optimisation?
Only after those questions do we decide whether the solution is consolidation, more separation, a budget change, better creative, or no structural change at all.
Structure should make decisions easier
The best Meta Ads account is not the most consolidated account.
It is not the most complex account either.
It is the account where the structure helps the business make better decisions.
Advantage+ campaign budget can give Meta more freedom to allocate spend.
Ad set budgets can protect spend when the business needs more control.
The learning phase can tell us when an ad set has limited optimisation data, but it cannot tell us whether the sales are profitable.
Advantage+ audiences can reduce the need for excessive interest segmentation, but they still operate inside business boundaries.
And consolidation can improve delivery when it removes meaningless divisions, but it becomes harmful when it hides important differences in markets, products, customers, or economics.
At Creative Brackets, we use one simple rule:
Add complexity only when it improves learning, control, or reporting.
If an account has twenty campaigns but nobody can explain why each one needs to exist, the solution may not be another campaign.
It may be a structure with a clearer reason behind every part.
Not sure whether your Meta account is fragmented or over-consolidated?
FAQ
1. Is CBO still called CBO in Meta Ads?
The current Meta name is Advantage+ campaign budget. CBO, meaning Campaign Budget Optimisation, is still widely used by advertisers as the older and more familiar term.
2. What does ABO mean?
ABO is industry shorthand for setting the budget at the ad-set level. It is not a separate Meta product name. With ad set budgets, the advertiser controls how much each ad set can spend instead of letting Meta distribute one campaign budget across them.
3. Is CBO better than ABO?
Neither is automatically better. Advantage+ campaign budget makes sense when ad sets can compete for the same overall budget. Ad set budgets are useful when individual tests, products, markets, or customer groups need guaranteed spend. The decision is about budget control rather than account size.
4. Does CBO combine learning across all ad sets?
No. Advantage+ campaign budget distributes money across ad sets, but learning still occurs at the ad-set level. Too many ad sets can still divide optimisation volume and leave some with limited data.
5. Does a Meta ad set need exactly 50 purchases per week?
Meta generally uses around 50 optimisation events in the week after the last significant edit as the point where delivery can stabilise. That is not a profitability requirement. A profitable ad set can operate below that level, while an unprofitable one can generate far more than 50 events.
6. Is Learning Limited always bad?
No. It indicates that Meta expects the ad set to receive too few optimisation events to exit the learning phase normally. It should prompt investigation, but it does not automatically mean the ad set should be turned off. Business results still matter.
7. Does broad targeting mean Meta can advertise to anyone?
No. Advantage+ audience still has supported audience controls such as location, minimum age, languages, and custom-audience exclusions. Other inputs, such as interests, lookalikes, and some demographic preferences, can act as suggestions that Meta may expand beyond.
8. Should prospecting and remarketing always use separate campaigns?
No. Separate remarketing makes sense when it has a distinct offer, message, budget, or commercial role. It should not exist only because an old account template says every funnel needs a dedicated retargeting campaign.