Why national weight is rarely spread evenly
A national plan is a convenient fiction. Media bought nationally reaches every region in proportion to its audience, and the plan's reach goal is reported as a single national figure. Underneath it, the business usually sells far more in some regions than others, the category itself is larger in some places, the product is stocked or served more widely in some, and competitors spend more in some. Spreading weight evenly ignores all of it.
The opposite habit is just as common: putting most of the money where sales are already strongest because that is where the results show up. That defends today's business but does little to grow it in regions where the category is large and the brand is small. The media objectives and reach goals guide treats geography as one layer of the audience; this tutorial takes that layer and asks how much weight each part of it deserves.
Three neighbouring questions belong to other pages. Splitting a budget between countries by each market's role is in the guide to media and channels by market. Setting the radius and Location TargetingRestricting where ads can show to chosen countries, regions, cities, postcodes or a radius around a point. for digital ads is in the guide to geo-targeted campaigns. Deciding who controls local marketing in a multi-location brand is in the comparison of centralised and local marketing control. Here the plan is one national plan, run centrally, and the question is only how its weight is split.
Before you start
- A signed objective and a national reach goal, from the objectives guide, so the regional split serves a goal rather than replacing one.
- Sales by region for a period long enough to smooth out one-off events, from the business's own records: orders, deliveries, store sales or a stockist's sell-out data.
- A base for each region: population, households or the number of potential customers, from the same source for every region.
- Any category data the business already holds by region, such as stockist category reports or survey data, with its source and date.
- Distribution or service coverage by region: stockists, listings, branches, delivery areas.
- A map of the regions the media can actually buy. Regional media rarely follow the business's own sales regions, so the regions in this tutorial are the ones media can be bought against.
- The draft national plan from the building a media plan tutorial, with its budget, period and reach goal.
The steps
List the regions the media can actually buy
Start from the media, not from the sales map. Ask each candidate medium which regions it sells: regional broadcast areas, regional press editions, out-of-home by town or city, digital location targeting by postcode or area. Draw the regions the plan will use so that each can be bought by at least one medium, and so that the business's sales and distribution data can be re-cut to the same boundaries. Name them neutrally, Region A to Region E, so the debate stays on the data. Where a medium cannot be split by region, it will carry national weight and the regional split applies only to the rest.
Build a sales index from your own data
For each region, divide its share of the business's sales by its share of the base, whether population, households or potential customers. A region whose share of sales matches its share of the base sits at the national average; one above it sells more than its size suggests, one below it less. Use a long enough period and check the result against a second period. Where the counts are small, read them as the guide to reading small numbers advises: as a range rather than a single figure, and with suspicion when one region moves far from its own history.
Add a category index where you have one
Do the same for the category: each region's share of category buying divided by its share of the base. Use only data the business already holds and can name the source of. Read the two indices together. A region where both are high is established; one where the category index is high and the sales index is low is where the brand is under-represented in a large category; one where the sales index is high and the category index low is a stronghold in a small market. Where there is no category data, say so and work from the sales index alone rather than inventing one.
Check distribution and service coverage
Before adding weight anywhere, check that people in that region can act on the advertising. The mental and physical availability framework, drawn from the Ehrenberg-Bass Institute's work, describes advertising as building memory and distribution as making the brand easy to find and buy; demand created where the product cannot be bought is wasted. List stockists, listings, branches or delivery areas by region. A region with a large category and thin distribution may need its weight held back until stocking or service improves, with the reason written down.
Map competitor weight and share of voice by region
Estimate where competitors advertise and how heavily, using whatever sources the business has: monitoring of competitors' media, observed out-of-home sites, regional press, ad libraries for digital. The share of voice guide covers how to estimate category weight. Then compare the brand's Share of VoiceYour share of market visibility. with its share of market region by region. Work published through the IPA, as the objectives guide reports it, describes share of voice above share of market as tending to go with growth. Treat it as a tendency, not a law, and as a check on the split rather than a formula for it.
Choose a weighting rule and write it down
Turn the evidence into a rule the team can state in a sentence: weight follows sales; weight follows category opportunity; a blend of the two; or a base weight everywhere with extra weight in named regions for named reasons. The table after these steps compares the common rules. Whatever the rule, write it down with the reasons, the data it uses and the adjustments for distribution and competition. A written rule can be tested and revised; a split agreed in a meeting cannot. The marketing budget calculator helps check that the regional totals still add up to the national envelope.
Check each region still reaches its goal at a useful frequency
A region given too little weight may fall below the frequency the message needs, and a region given a great deal may pile frequency on the same people. For each region, estimate ReachUnique people who saw an ad. and frequency from its share of the weight and its audience size. As spend rises within an audience, each increase adds less, the pattern of Diminishing ReturnsEach extra unit of spend or effort buys less extra result than the one before it. described in the guide to marginal returns and forecasting. If a heavy region is already past the point where extra weight adds new people, move some to a region still building reach.
Match the weight to the regional media available
Translate each region's share of the budget into lines that can be bought there. Some regions will have a choice of regional media; others will have little beyond national media and digital location targeting. Where a region's weight cannot be delivered through media that reach its audience, either lower the weight or accept that national media carry it. Check overlap: national media already deliver weight to every region, so the regional lines sit on top, and a region's total is national weight plus its regional lines.
Set up a regional test of the split
The split is a hypothesis, and regions make it testable. Choose one region to receive heavier weight than the rule gives it, and a matched comparison region that keeps the rule's weight, chosen on similar sales history and size. The comparison region acts as the Holdout GroupA randomly chosen part of an audience kept from exposure to measure what exposure caused.. The guide to measuring offline media explains regional testsA test that runs a campaign in some regions and holds it back in comparable ones, then compares total outcomes, where people cannot be randomised one by one. for media that cannot be randomised at the person level, and the comparison of geo holdout tests and conversion lift studies explains when a regional design is the right one. Decide before launch what result would change the rule.
Review the split after the campaign
When the campaign ends, compare each region's delivery with its plan, read the test against the comparison region, and recalculate the sales index with the new period. Read small regional differences as ranges, not verdicts. Then revise the rule: keep it, shift the blend, or change the distribution adjustment. Record the result in the post-campaign analysis so the next plan starts from evidence about regional weight rather than from last year's split.
Weighting rules compared
Each rule embodies a view of where growth comes from. None is right in general, and most businesses end up with a blend and a test.
| Rule | How weight is set | Suits | Risks |
|---|---|---|---|
| Even by population | Each region's share of the base | A new brand with no regional history; national media only | Ignores where sales, category and distribution differ |
| Follow sales | Each region's share of the business's sales | Defending established regions; short-term response | Starves regions where the category is large and the brand small |
| Follow category opportunity | Each region's share of category buying | Growth in under-represented regions | Pours weight where distribution may not support it |
| Blend | A stated mix of sales and category shares, adjusted for distribution | Most established brands with category data | The blend becomes habit unless tested |
| Base plus named regions | Even base weight, extra weight in regions chosen for stated reasons | Launches into new regions; testing | Extra weight spread across too many regions to register |
Whichever rule is chosen, the adjustments matter as much as the rule. A region's weight is held back where distribution is thin and reconsidered where competitor weight is unusually heavy or light. The media planning fundamentals hub shows where the regional split sits among the other planning decisions.
A packaged coffee brand weights its regions (illustrative)
Nuvrello Coffee splits its autumn plan by region
Nuvrello Coffee, an invented packaged-coffee brand sold through grocers, has a £400,000 autumn plan. Online video and paid social are bought nationally for £220,000; the remaining £180,000 goes to regional out-of-home and regional radio, split across Region A to Region E.
From two years of its own sell-out data and household counts, the brand builds a sales index for each region: Region A 1.35, Region B 1.10, Region C 0.95, Region D 0.80 and Region E 0.70. A grocer's category report, which the brand receives as a supplier, gives a category index: Region A 1.05, Region B 1.00, Region C 0.90, Region D 1.25 and Region E 0.85. Region D stands out: the category is large there and the brand is small.
The brand's rule is a blend: half the regional budget by sales share and half by category share. On that rule Region D would receive £41,000. But the distribution check shows the brand is listed in only 40% of the grocers in Region D, against at least 75% elsewhere. The team holds Region D at £30,000, writes down that the extra £11,000 is released once listings reach 60%, and gives the held amount to Regions B and C.
To test the blend, Region B receives £10,000 more than the rule gives it, and Region C, with a similar sales history and size, keeps its rule weight as the comparison. After eight weeks, sell-out in Region B is 6% above its baseline against 2% in Region C. With only two regions compared, the team reads this as encouraging rather than proven and repeats the test the following spring.
Common mistakes
- Using the sales map instead of the media map. If the regions cannot be bought, the split cannot be delivered. Start from the regions the media sell.
- Following sales alone. It feels safe and measures well, but it defends today's regions and starves the places where growth is possible.
- Adding weight where the product cannot be bought. Demand created without distribution is wasted. Check stocking and service before the money moves.
- Inventing category data. A guessed category index looks like evidence. Use only data the business holds and can source, and say where there is none.
- Spreading extra weight across every region. Weight added everywhere is the same as no regional weighting. Name the regions and the reasons.
- Ignoring national weight. National media already deliver weight to every region; the regional lines sit on top.
- Reading one region's result as proof. Regional numbers are small and noisy. Read them as ranges and repeat the test before changing the rule for good.
