What regional advertising means for a national brand

Regional advertising is media bought for particular parts of the country rather than for the whole. For a national brand that can mean a television region, the radio stations of a few cities, outdoor sites in selected towns, or local digital targeting. It is not a separate or lesser campaign. It is a decision about where the same message is shown, and how the creative and the offer can reflect local life.

The argument for it is simple. Most people do not live in London, and media planned from London can drift away from how people in other places live, shop and talk. A column by Emma Armstrong, an AV executive at Republic of Media, made the case from the agency side. It argued that regional media is treated as harder to plan and buy than it needs to be, and that tailoring national campaigns to regional life can reduce wastage. This post builds on that argument and tests it against public data.

Regional is not the same as local. A regional approach can cover a wide area, such as the North West or Scotland, and still be planned as a set of areas that can be compared with one another. That is the way this post treats it, because a comparison is what turns an idea into a result you can act on.

Why the audience outside London matters

The Office for National Statistics put the UK population at 69,281,400 in mid-2024 (ONS, 2025). The Greater London Authority, drawing on the ONS mid-2023 estimate, puts London at 8.95 million people (Greater London Authority, 2024). The two figures come from different years, so the comparison is approximate. It still makes the point: roughly one person in eight in the UK lives in London, and the rest live elsewhere.

That proportion matters because much media planning starts from London. A planner who works in the capital and reads the same papers and listens to the same stations as colleagues can hold a picture of the country that does not match the people in other regions. The Radiocentre study discussed below found a gap between what advertisers and agencies believe about media and what the evidence shows. The same logic applies to place. The people a brand most needs to reach are shaped by local shops, local work and local news.

Media habits point the same way. Ofcom's Media Nations 2026 report says that just under nine in ten UK adults still tune in to live radio each week (Ofcom, 2026). Radio is a medium built around local stations and local content, which makes it a natural tool for regional planning.

Why regional media is under-invested

Regional media has a reputation for being hard to plan and hard to buy. Buying across several local outlets takes more time than one national deal, and the audience data is less standardised. Many brands therefore default to national channels, even when their customers are concentrated in particular areas. The habit is understandable, but it is not the same as a judgement about what works.

The evidence does not always match how planners rate media. A 2018 study for Radiocentre, carried out by Ebiquity, reviewed 75 research reports and Ebiquity's own data sets, and surveyed 116 senior advertisers and agency executives. Radio ranked second to TV on the evidence but sixth in advertiser and agency perceptions (Radiocentre, 2018). The study covered the UK as a whole. It did not look at London planners in particular, so the gap it measures is about how radio is perceived, not about where the perception is strongest.

Market size is a third gap, and it is hard to pin down. The Grove Media estimates the regional advertising market at £1.5bn to £2bn a year. That figure appeared in an opinion piece by its managing director in The Medialeader in July 2023, and the same piece says there are no definitive sources for total regional adspend (The Grove Media, reported by The Medialeader, 2023). Treat the range as an industry estimate, not a measured total, but it is a useful sign that the category is large enough to deserve proper attention.

How a regional test works, and what the evidence shows

The most reliable way to judge regional advertising is a geo experiment. Areas are assigned to receive the advertising or to act as controls, and the sales or visits in the two groups are compared. A 2011 Google research paper by Jon Vaver and Jim Koehler describes the approach. Separate, non-overlapping areas are randomly placed into control and treatment groups, and geo-targeted advertising carries out the assignment (Vaver and Koehler, Google, 2011). The paper was written about search advertising, but the logic applies to any media that can be switched on by area.

Regional media suits this design, because the areas are large enough to show a difference and the media can be bought by area. An older UK example is a Radiocentre-published study that compared radio regions with matched control regions. It reported that brands averaged a 9% sales uplift in radio regions. The analysis used Tesco Clubcard data and covered 17 brands in campaigns run between summer 2000 and summer 2002 (Radiocentre, undated). Because the data is that old, it shows the method rather than the size of the effect today.

The method has limits. Meta's GeoLift documentation notes that people-level experiments are preferred when they are possible, because they have greater statistical power, and that geo experiments fill the gap where people-based measurement is not feasible (Meta, undated). A test across two regions is fragile. A test with more areas, a clear baseline and a fixed plan is far more useful.

A method you can follow

Here is a practical sequence. For example, a consumer brand with national television and social spend might want to know whether a burst of regional radio lifts sales in one part of the country. The steps below are a simple version of a geo test, and they work for outdoor and television as well.

The output is not a single number to celebrate. It is an estimate of how much extra activity the media created, with a range around it. If the range is wide, the honest answer is that the test was not big enough, and the next test should be larger.

  • Choose the areas before you buy. Pick regions that match on population, store coverage and prior sales, so that the control areas are fair comparisons.
  • Set a baseline. Record sales, store visits or branded search for each area for several weeks before the campaign, so you know the normal pattern and its variation.
  • Assign the areas in advance. Decide which areas receive the media and which stay as controls, and keep the plan fixed for the length of the test.
  • Hold other activity steady. Keep national media, promotions and prices the same across test and control areas where you can, and write down anything that differs.
  • Run long enough to cover a normal purchase cycle. A product bought every week needs a shorter read than one bought once a year.
  • Read the result as a range. Compare the estimated uplift with its uncertainty before you decide whether to scale the spend.

Common mistakes, and how to measure regional media

The most common mistake is choosing areas for convenience, such as where a salesperson lives or where stock is easy to move. That choice can bias the result before the test starts. The second is judging regional media by clicks. Radio, outdoor and television work largely through memory and recall, so click data will understate their effect. The third is testing too small. Two areas can produce a large swing by chance, which makes a single result hard to trust.

The fourth mistake is changing the plan mid-test. Adding a promotion to the test areas halfway through makes the result impossible to read. If something unplanned happens, such as a local store closure, record it and treat the affected area with care.

Measures should match the job. For store-based brands, footfall and store sales by area are the closest measures, and the earlier post on outdoor, TV and radio store visits explains how to set those up (outdoor, TV and radio: measuring store visits). For online brands, branded search and direct visits by area can show whether awareness is moving. Whatever you choose, agree it before launch.

What good looks like, and what to ask a supplier

A good regional plan starts with the customers rather than the rate card. It names the areas, explains why they were chosen, sets out the test design and says how the result will be read before any money is spent. It also treats regional spend as a test-and-learn budget that can move between areas as the evidence builds.

Regional advertising will not suit every brand. A national brand with a thin sales base outside its key cities may see little signal from a small test. The question worth asking is not whether regional media is good in general, but whether a well-designed test will show whether it adds sales for your business.

  • Which regions do you recommend, and why those rather than others?
  • How will you choose control areas, and how will you check that they match the test areas?
  • What baseline data do we need, and how far back does it go?
  • What will we measure, and how will you handle small differences?
  • How will the result change our next budget decision?

Put this into practice

Book a discovery call and we will review your channels and results first, then agree what to test.

Plan a regional test for your brand

Questions and answers

Is regional advertising only for brands with regional customers?

No. It is a way to test where media adds sales. A national brand can use regional areas to learn where its message works best, even if most of its customers are in other places.

How big does a regional test need to be?

There is no single answer. A test needs enough areas and enough sales to show a difference above normal variation. Set the size before launch, using your past sales data by area.

What is the difference between a geo test and an A/B test?

An A/B test splits individual people or visits between two versions. A geo test splits areas, so it suits media that cannot target individuals, such as outdoor, television and radio.

Do I need specialist tools for a regional test?

Not always. Store sales and footfall by area, compared across matched areas, can answer many questions. Specialist tools help with larger tests and with the statistical checks.

Can I start with a small budget?

Yes, if the test is designed to read a clear result. A very small test across two areas can mislead, so it is often better to run fewer, larger tests than many tiny ones.

References

  1. Office for National Statistics, "Population estimates for the UK, England, Wales, Scotland and Northern Ireland: mid-2024", 2025
  2. Greater London Authority, "London's Population", 2024
  3. Radiocentre, "Re-evaluating Media" (Ebiquity study), 2018
  4. The Medialeader, opinion piece by David Price on the regional advertising market, reporting an estimate from The Grove Media, 2023
  5. Vaver, J. and Koehler, J., Google, "Measuring Ad Effectiveness Using Geo Experiments", 2011
  6. Ofcom, "Media Nations 2026", 2026
  7. Radiocentre, "Radio: the sales multiplier" (study of campaigns from 2000 to 2002, undated)
  8. Meta, "About GeoLift", undated

← Back to all articles