Audience research
Part of Match out of home advertising research methods to the buying decision
When to segment an out of home advertising target audience
How to segment an out of home advertising target audience in England, from journey data and employment figures to the compliance checks before activation.
What to take away
- The 2021 Census travel-to-work data and the Office for National Statistics release on people in work are the most recent official bases for audience sizing in England, so dated modelled figures are no longer good enough.
- Segment by journey purpose and time band before age or income, because place-based inventory is bought by daypart and location.
- Check every segment against the Information Commissioner's Office direct marketing checklist before any addressable or retargeted element goes live.
- A before and after table, built from your own campaign data, shows a client why a segment changed.
Start with the journey, not the person
Place-based advertising reaches people in transit, so the journey is the unit of analysis. A commuter walking to a station at 07:40 behaves differently from the same person driving to a retail park at 14:00.
Build your first cut around purpose: commute, shopping, leisure or service errand. Then layer the time band. A segment such as "weekly grocery trip, Saturday morning, 20 minute drive" is buyable. "Women aged 25 to 44" is not, because it does not tell you which sites to book.
If you are new to this method, the audience research guide for England sets out the definitions and data sources behind each step.
Check the employment base before you size anything
Employment patterns drive daytime footfall, so pull current figures rather than a number from a previous pitch. The Office for National Statistics publishes its people in work data as a recurring release, and the page date tells you how recent the numbers are.
Use it to sanity-check any claim about a working population. If your segment assumes 30,000 daytime workers in a district centre, the employment statistics should not contradict that by a wide margin.
Match any audio-adjacent claim, such as a drive-time radio audience used to justify a roadside format, to the latest quarterly radio listening data rather than a remembered figure.
Build the segment in four passes
- Geography: define the catchment by travel time, not postcode district.
- Daypart: split the day into bands that match your booking options.
- Purpose: assign each band a dominant journey purpose.
- Behaviour: add a modifier such as repeat visit or competitor proximity.
Four passes is usually enough. A fifth often produces a segment too narrow to buy. To understand who else competes for the same audience, the competitor research method maps rival advertisers against the same sites and dayparts.
Before and after: a worked comparison
The table below shows a typical rebuild. The figures are illustrative examples, not measured results.
| Stage | Segment definition | Sites in scope | Weekly impressions (illustrative) |
|---|---|---|---|
| Before | Adults 18 to 54, city centre | 14 | 410,000 |
| After | Grocery trip, Saturday 09:00 to 13:00, 20 minute drive | 6 | 165,000 |
| After | Station commute, Monday to Friday, 07:00 to 09:00 | 4 | 290,000 |
The narrower segments reach fewer people but map onto inventory you can book. That is the trade you are making.
Check compliance before you activate
Any segment that touches addressable screens, mobile retargeting or email follow-up moves into direct marketing territory. The Information Commissioner's Office publishes a direct marketing checklist covering consent, legitimate interests and the records you need to keep.
Read it before you brief a media owner, not after. Segments built on inferred location data are the ones most likely to attract a complaint.
Document what you rejected
Keep a short note on each segment you discarded and why. This matters when a client asks in month three why the audience shrank. It also stops the same weak segment being rebuilt next quarter by someone who was not in the room.
Store the note alongside the campaign plan, with the data source and its date. If a figure came from a census release or an employment bulletin, name it. Undated numbers are the main reason audience claims fall apart in a review.
Common questions
How many segments should one campaign use?
Two or three is usually workable. More than that and the budget spread across each segment becomes too thin to produce a readable result.
Can I use the same segments across England, Scotland and Wales?
No. Travel patterns, employment mix and available inventory differ, so rebuild the segment for each nation rather than copying an England set north or west.
Do I need new research for every campaign?
Not always. Reuse the segment logic, but refresh any figure that carries a date, particularly employment and listening data.
What if the client insists on a broad age and gender segment?
Run it alongside a journey-based segment in the same campaign. Comparing the two gives the client evidence for the next planning round.
