Planning a two-city paid social campaign

Planning priorities

Why comparable city data matters

A CPM is an auction outcome, not a rate card. It reflects how many advertisers want the same impression at the same moment, how well your targeting matches available supply, and how the platform prices the placement. Two cities in the same country can produce very different numbers on identical creative.

London holds the largest concentration of UK paid-social agencies, in-house brand teams and platform representatives. That density means more advertisers bidding into the same audience pools, more agencies testing the same hooks, and more budget chasing finite attention.

Manchester is a major northern hub for ecommerce, retail and media agencies, with a distinct client base and lower staff costs. Fewer bidders, lower overheads and different client categories all push in the same direction.

The gap is not a discount you can bank in advance. It moves with the season, the sector and the objective. A retail brief in November behaves differently from a B2B software brief in February, in both cities. What stays stable is the direction: London costs more per thousand impressions for comparable inventory.

No public dataset publishes a London or Manchester paid-social CPM. Meta, TikTok and LinkedIn report it only inside your own ad accounts. For illustration only, a 2:1 London-to-Manchester CPM gap is enough to halve the impressions a fixed London budget buys, so the ratio matters more than the absolute figure.

That single fact reshapes the plan. If you hold the budget split fixed and simply swap London for Manchester, you buy more impressions in Manchester. If you hold impressions fixed, you spend less in Manchester. Neither is automatically right. The choice depends on where the revenue is and what the campaign needs to learn.

Start by writing down what the brief actually requires: reach, sales, leads, or evidence. Then price both cities separately before you commit a split. Most plans fail here because the split is decided in a spreadsheet before anyone has seen a live CPM from either market.

Using ONS regional data as cost framing

The Office for National Statistics publishes regional economic data that helps you frame a two-city plan. ONS regional accounts cover gross value added and household income by region, including Greater London and the North West. They tell you how much economic activity sits in each area, which is a reasonable proxy for advertiser concentration.

Use them for context, not for CPMs. Regional accounts do not measure advertising inventory, auction pressure or platform pricing. They tell you that London and the North West differ in output, earnings and industry mix. That is useful when a client asks why the same brief costs different amounts in two cities.

ONS economic output and productivity data adds the productivity angle. Higher output per head in the capital supports higher operating costs for agencies and brands, which in turn supports higher bids for attention. It is a chain of inference, not a measurement, and you should present it that way.

For cost direction over time, ONS inflation and price indices give you the general price trend. Media inflation is not the same as consumer price inflation, but the indices show whether the wider cost base is rising or easing. That helps when you forecast next quarter's budget.

ONS digital economy data covers the digital sector as market context. It shows how much of the economy is digital, which supports the case for paid social as a line item, but it says nothing about what a Manchester impression costs.

These sources explain why costs differ. They do not tell you what you will pay.

Audience size, competition and auction density in each city

Auction density is the number of advertisers competing for the same impression. It is highest in London, where brands, agencies and freelancers cluster, and that concentration lifts the clearing price on broad targeting.

Manchester has a smaller but real advertising market. The audience pool is smaller in absolute terms, but so is the number of bidders. For many targeting setups, that means less pressure per impression and a lower CPM. The effect is strongest in broad reach campaigns and weakest in narrow B2B targeting, where national or global competition sets the price.

Audience size matters in the other direction. A London-only geo-target may still be large enough to reach frequency goals quickly. A Manchester-only target may saturate sooner, which pushes frequency up and can raise effective cost per result even when CPM looks cheap. Watch frequency alongside CPM.

Platform mechanics add a layer. Meta, TikTok and LinkedIn all price through auctions, and each has different supply conditions.

When you compare the two cities, compare like with like. Same objective, same placement, same creative, same audience definition, same dates. A London CPM from a conversion campaign and a Manchester CPM from a reach campaign are not comparable, and presenting them side by side misleads the client.

Splitting budget between London and Manchester placements

The split should follow revenue and learning, not sentiment. If the client's customers are concentrated in one city, that city should carry more budget. If the client wants national coverage, the split should reflect where impressions are cheapest per relevant person, not where the agency has an office.

A practical method is to set the split on cost per outcome, not cost per impression. Run a calibration period in both cities with identical creative and targeting. Keep it running for at least 14 days per city and past the platform's learning phase.

A typical calibration floor is £1,000 per city, though the platform sets the real number. Read the cost per result from each, then allocate the next tranche in proportion to that cost, weighted by the revenue value of each market.

Building a paid social budget from evidence rather than benchmarks is the discipline that makes this work. The calibration period is the evidence. Without it, the split is a guess dressed as a plan.

For city-level CPMs, a UK planner has three routes. Meta Ads Manager, TikTok Ads Manager and LinkedIn Campaign Manager all report CPM inside your own account, split by the city targeting you set. Your agency's own historical account data is the second. Published UK media benchmarks are the third, but few of them report London and Manchester separately.

Take a £30,000 monthly budget and a 60:40 split. London receives £18,000 and Manchester £12,000.

At an illustrative London CPM of £18 and a Manchester CPM of £9, London buys 1,000,000 impressions and Manchester 1,333,000. With a minimum of 200,000 impressions per test, London supports 5 tests and Manchester 6.

The example below is arithmetic, not a market forecast. It shows the shape of a split, not the rates you will pay.

Rebalance by impressions, not pounds, if the goal is reach. Rebalance by cost per result if the goal is conversion. State the rule in the plan so the client can see why the split moved.

Channel mix consequences for Meta, TikTok and LinkedIn

Channel mix across Meta, TikTok and LinkedIn responds to CPM differences because each platform has its own cost profile and its own audience density in each city.

Meta usually offers the widest reach in both cities and the most stable CPMs. In Manchester, Meta can carry more of the reach load for the same money, which frees budget for testing or for LinkedIn. In London, Meta absorbs more budget for the same impressions, so the plan needs a clearer reason for every pound.

TikTok tends to reward creative volume. Lower CPMs in Manchester make it cheaper to test multiple creative routes there before exporting winners to London. That is a genuine planning advantage: you buy learning at a lower unit cost, then apply it where impressions are expensive.

LinkedIn is the least price-sensitive to city in relative terms, because its targeting is professional rather than geographic. A London job title and a Manchester job title often clear at similar CPMs, so LinkedIn budget should follow the client's decision-maker locations, not the CPM gap.

Choosing paid social channels on like-for-like criteria keeps the mix honest. Compare the same objective and the same audience definition across platforms before you move budget between them.

A sensible default for a two-city brief is to let Meta carry reach, TikTok carry creative testing, and LinkedIn carry precision targeting. Then adjust the weights by city based on measured cost per result, not on assumed platform strength.

How testing volume changes when CPMs differ

Testing volume is a function of impressions, not budget. A test needs enough delivery to produce a readable result. If Manchester delivers more impressions per pound, the same testing budget buys more tests there.

This has a direct planning consequence. Put the exploratory tests in the cheaper market. Run more variants, more hooks, more formats, and let the data pick winners. Then take the winners into London, where each impression costs more and wasted delivery is more expensive.

Set a minimum delivery threshold per test before you start. Decide how many impressions or results a variant needs before you can judge it. A workable floor is 200,000 impressions or 30 conversions per variant, whichever arrives first. Then calculate how many variants each city can support within the budget. That number will be higher in Manchester.

Giving every paid-social metric an event, a denominator and an owner makes the testing plan auditable. A test without a defined success metric is just spend.

Do not let the cheaper market become a dumping ground for weak creative. The tests still need a hypothesis, a control and a decision rule. Lower CPMs reduce the cost of being wrong. They do not remove the need to be deliberate.

Keep a shared test log across both cities. Record the variant, the city, the spend, the impressions and the outcome. When the client asks why London looks more expensive, the log shows what the extra cost bought.

Reporting a two-city plan without inventing benchmarks

Benchmarks are where two-city plans go wrong. A single national CPM benchmark hides the gap you are managing. A London benchmark applied to Manchester overstates cost. A Manchester benchmark applied to London understates it. Either way, the client gets a number that does not match the invoice.

Report actuals by city. State the CPM paid, the impressions delivered, the results and the cost per result for each. If you must reference an external figure, label it as context and say what it does not cover.

Why ONS, Ofcom and Google Ads figures make poor paid-social benchmarks explains why published figures rarely map onto a specific brief.

Use a simple reporting table with one row per city and one column per metric. Add a variance column against plan. If London came in above the planned CPM, say so and explain what changed: seasonality, competition, creative fatigue or targeting.

Keep the language plain. The client does not need a media theory lecture. They need to know what was spent, what it bought and what happens next.

Common questions