One Social Ad Budget or Forty?

One campaign with location targeting, not forty campaigns. Meta's delivery system needs roughly 50 conversions a week per ad set before costs stabilize, and splitting a budget forty ways puts every piece below that line. Geography belongs in the targeting, not in the campaign structure.

This is the most expensive structural mistake in multi-location paid social, and it is made for a good reason. Splitting evenly feels fair, and every location manager can see their own number. It is fair and it does not work.

Why does splitting the budget fail?

Because the learning threshold is per ad set, not per dollar.

Meta documents this in its own Business Help Center. An ad set is described as "learning limited" when it is unlikely to receive around 50 optimization events in the week following your last significant edit. Below that, costs stay unstable and the system never optimizes properly.

Now do the arithmetic on a split budget. A chain with forty locations and $20,000 a month has $500 per location per month, or about $115 a week. If a conversion costs $30, that is under four conversions a week per location, against a threshold of 50.

Every single campaign sits permanently in learning. Not one of them ever reaches stable costs. The brand concludes paid social does not work for them, when what actually happened is that they ran forty underfunded experiments simultaneously.

The same money in one campaign generates roughly 165 conversions a week. That campaign optimizes. I set out the underlying arithmetic in the ad budget guide.

So how does geography get handled?

In the targeting, which is what targeting is for.

One campaign can target every location's catchment area at once. Meta will spend where it finds results, which means the budget flows toward the locations and audiences that are actually converting rather than sitting in equal piles regardless of performance.

That is the part people resist, and it is the point. Equal distribution is a decision to ignore performance. If location twelve converts at half the cost of location nine, sending them the same budget is choosing to buy fewer customers on purpose.

For most chains, the structure that works is one prospecting campaign covering all locations, and one retargeting campaign covering all locations. Two campaigns, not eighty.

When do separate campaigns actually make sense?

Three cases, and they are narrower than people assume.

Genuinely different offers. If location nine is running a promotion the others are not, that needs its own campaign because the creative and the landing page differ.

Genuinely different markets. A location in a different country, language or regulatory environment is a different business for advertising purposes.

A location with enough volume to stand alone. If a single site can generate 50 conversions a week on its own budget, it can have its own campaign and benefit from the focus. That is a real threshold you can calculate, not a matter of importance.

Everything else belongs in the shared campaign. Notice that "the franchisee wants their own campaign" is not on that list, which is a political problem rather than a media one, and worth handling directly.

That last row solves the argument that causes most of the fragmentation. You can report by location without structuring by location, and reporting is what people actually want.

What about local budget contributions?

Common in franchise systems, and the structure matters more than the amount.

Many franchise agreements require locations to contribute to a national fund and permit or require additional local spend. The trap is treating the local contribution as a separate campaign, which recreates the fragmentation problem one site at a time.

The better handling is to pool contributions into the shared campaign and weight the targeting, so a location contributing more gets more delivery in its catchment without needing its own ad set. That keeps the volume together where the system can use it.

Where genuinely local spend does earn its place is on things that cannot be centralized. A location sponsoring a local team, promoting an event at that store, or responding to something happening in that neighborhood this week. Those are small, occasional, and worth doing separately because they are actually different.

How should performance be judged?

Per location, on the numbers that survive comparison.

Cost per conversion by location is the useful report, and it is available from a shared campaign through breakdowns. What is not useful is ranking locations by spend or by raw conversion count, because those mostly measure catchment size.

The number to watch per location is cost per conversion against that location's own history, and against what a customer is worth there. A site with a higher cost per lead can still be the better investment if its average order value is higher, which is common across a chain with different formats or demographics.

The reporting question and the structure question get confused constantly. You can report at any level of detail you like from a consolidated campaign. Structure for the algorithm, report for the humans.

What happens when a new location opens?

Add it to the targeting, not to the campaign list.

An opening usually justifies a short, separate push, because the offer is genuinely different: a new store, an opening event, a launch promotion. Run that as its own campaign for a few weeks while it is genuinely a different message.

Then fold the location into the shared campaign and shut the launch campaign down. The common failure is leaving it running forever, so a chain accumulates one campaign per opening until the account has thirty campaigns nobody remembers creating.

Audit the campaign list quarterly and turn off anything whose reason for existing has passed.

What about creative across locations?

One set, refreshed, rather than one set per location.

The same fragmentation logic applies a level down. Every additional creative divides the same budget into smaller portions, and a portion below the learning threshold never reaches stable costs no matter how long it runs. Forty locations each with their own three assets is a hundred and twenty pieces of creative sharing one budget, which is a very efficient way to learn nothing.

What works is a small shared set, refreshed on a cycle. Two or three assets carrying the full budget reach stable costs. A dozen sharing it stay in permanent adolescence.

Where localization genuinely helps is in the copy rather than the imagery. Dynamic location insertion, or a small number of regional variants where the market really differs, gets you local relevance without multiplying the creative count. A store name in the headline costs nothing structurally. A separate video per site costs you the whole budget's momentum.

The exception is a location whose market is visually different in a way customers would notice: a coastal store versus an inland one, a city format versus a drive-through. That is a real difference and worth a variant. "The franchisee wants their own photos" is not.

Who should actually run it?

Centrally, with local input, and this is usually a political conversation rather than a media one.

The structure that performs best is a single account managed in one place. That is not a comment on anyone's competence. It is that the arithmetic only works when the budget is pooled, and a pooled budget cannot be run by forty people.

What locations should own is the input: what is happening at that store, what is selling, what the local competition is doing, when the market shifts. That information is genuinely local and central teams are bad at getting it.

The failure mode is a system where each location runs its own ads with its own agency or its own nephew. You get forty underfunded accounts, no shared learning, inconsistent tracking, and locations bidding against each other for the same people. Consolidating that is unpopular, because it reads as taking something away, and it is almost always the single highest-value change available.

What I see in multi-location ad accounts

The same shape, at every size of chain.

Dozens of campaigns, one per location, every one learning limited, none producing stable costs, and a monthly report showing wildly different cost per acquisition by location that is mostly noise rather than signal. With four conversions a week, the variance between locations tells you nothing.

The second pattern is launch campaigns that never ended. A chain that opened eleven locations over three years has eleven dormant or half-dormant campaigns still technically active, quietly consuming budget and attention.

The third is a national campaign and forty local ones running simultaneously against the same audiences, competing with each other in the auction. That is a brand bidding against itself with its own money.

Consolidation is most of what the paid team at Inertia Digital Marketing does in the first month of a multi-location engagement, and it almost always means turning things off.

What to do on Monday

Open your ads manager and count the campaigns. Then count how many are in learning limited. If most of them are, you do not have a targeting problem or a creative problem, you have a division problem.

Add up the total monthly spend and divide by your cost per conversion. That tells you how many conversions the whole account can generate in a week. If the answer is under 50, you have exactly one campaign's worth of budget no matter how many locations you run.

Then consolidate to one prospecting campaign with location targeting, and report by location rather than structuring by it.

FAQ

Should multi-location businesses run local or national ad campaigns? One consolidated campaign with location-based targeting, in nearly all cases. Meta needs around 50 conversions per ad set weekly to optimize, and splitting a budget across locations usually puts every campaign below that threshold.

How do you split social ad budget between locations? Ideally you do not. Pool the budget into a shared campaign and let the targeting cover every catchment, so delivery flows toward what converts. Report performance by location instead of structuring the account by location.

When should a location have its own campaign? When it has a genuinely different offer, sits in a different market or language, or generates enough conversions on its own budget to clear the learning threshold. Wanting separate reporting is not a reason, because reporting can be split without splitting the account.

Why are all my location campaigns in learning limited? Because each is receiving fewer than about 50 optimization events a week. That is a budget division problem rather than a targeting problem, and it is fixed by consolidating rather than by editing audiences.

Should franchisee local contributions run as separate campaigns? Usually not. Pool them into the shared campaign and weight the targeting toward contributing locations. Keep separate spend for things that genuinely cannot be centralized, like a local sponsorship or a store event.

About the author

Adam Palmer is President of Inertia Digital Marketing, a digital marketing agency working with clients from small business to multi-location enterprise and franchise operations. Across 15 years he has managed more than $50 million in paid digital campaigns, and he writes here about what actually moves numbers rather than what sounds good in a strategy deck. Inertia built and maintains Social Fonts, which is free and always will be.