Outdoor advertising has a reputation for being hard to measure. That is only partly deserved. A billboard does not record who saw it, so individual-level attribution from poster to purchase is not something to promise; but a campaign can be verified, its response signals can be tracked, and its effect can be estimated with a reasonable design set up before launch. This article explains what can and cannot be measured for out-of-home in Japan, which signals are worth tracking, and how to set up measurement so the result can be trusted.
What can and cannot be measured in outdoor advertising
Out-of-home is a broadcast medium in public space. There is no click, no cookie and no individual-level record of who saw a billboard, and in Japan, as elsewhere, outdoor and station media are not set up to track individuals. Some providers offer audience or visit estimates built from panels or aggregated location data; these are modelled at group level, not records of individuals. Any claim that a specific person saw a specific poster and then bought something should be treated with suspicion.
What can be measured falls into three layers. Delivery: did the campaign run where and when it was booked? Response: did the signals a campaign should move — searches for the brand, visits to the website, scans, mentions — move during the flight? Effect: compared with a baseline or with similar areas that did not see the campaign, how much of that movement can reasonably be attributed to it? Each layer answers a different question, and a good report keeps them separate.
Delivery evidence: proof that the campaign ran
For static media — posters, station advertising, billboards — media owners or their contractors typically provide posting photographs showing the creative in place. For digital screens, including digital billboards and DOOH networks, a play log or equivalent delivery report records when and where the spots ran.
Some media owners also provide audience estimates for their inventory, derived from footfall or ridership data, panels, on-site counting, anonymised camera-based counts or aggregated mobile location data. These are useful for comparing options and for planning, but they are modelled estimates of opportunity to see, not counts of people who looked. Report them as such, and do not add estimates from different owners together as if they were the same measure. The DOOH article covers what digital networks typically report.
Response signals worth tracking
Response signals are usually the most accessible evidence that a campaign is being noticed. Those most often useful for out-of-home are:
Branded search — searches for the brand or product name in Japan, tracked through your own search analytics and public search-trend tools. Outdoor advertising rarely produces clicks of its own, but it can show up as people looking the brand up. Watch branded queries by region and by language.
Direct and regional web traffic — visits that arrive without a referring link, and visits from the cities where the campaign ran, compared with the period before and with regions where it did not run.
QR codes and short URLs — QR codes are widely used and familiar in Japan, which makes them a practical bridge from a poster to a landing page. Give each campaign, and ideally each city or format, its own tagged code so that scans can be attributed.
Social mentions and photographs — especially for landmark placements, where people photograph and share the creative. Mentions and user-posted images are a signal of reach beyond footfall.
Enquiries and sales by channel — for B2B and higher-value products, the number and source of enquiries during and after the flight.
None of these is proof on its own. Together, and compared with a baseline, they show whether the campaign moved behaviour in the places and at the times it ran.
Brand lift studies
For brand-building campaigns, the outcome that matters is often awareness, consideration or perception rather than an immediate action. A brand lift study measures this with surveys: the same questions asked before and after the campaign, or asked at the same time of people in campaign areas and in comparable areas without the campaign.
The design matters more than the survey itself. Samples need to be large enough to detect a difference, comparison areas need to be genuinely comparable, and the questions need to be fixed before the campaign starts. A study commissioned after the fact, with no baseline, rarely produces a result that anyone can rely on.
Footfall, sales and test-and-control areas
Where the goal is visits or sales, the most convincing evidence comes from comparing areas. Run the campaign around one set of stores, stations or cities and not around a comparable set, and compare footfall or sales between the two over the same period. The same approach works with web traffic and enquiries split by region.
Japan’s rail-centred cities lend themselves to this, because station and line buys can be concentrated in clearly bounded areas. The difficulty is finding truly comparable controls: different districts differ in audience, and a national campaign running at the same time can affect both groups. Agree the test and control areas with whoever will analyse the data before booking the media.
Setting up outdoor advertising measurement before launch
Many measurement problems are set-up problems. A short checklist, completed before the campaign goes live:
Define one primary conversion — the single action that counts as success, such as an enquiry form submission, a sign-up or a store visit — and make sure it is recorded as a key event in analytics.
Record a baseline — branded search, direct and regional traffic, and the primary conversion for several weeks before launch.
Choose control areas if you want to estimate effect, and agree them before booking.
Build a dedicated landing page in the right language for the campaign, so that visitors from the creative land somewhere relevant and can be counted.
Tag every QR code and short URL with campaign parameters (UTM), separately for each city or format where the budget allows.
Collect delivery evidence — posting photographs and play logs — as part of the booking, not as an afterthought.
Fix the reporting dates and the questions the report should answer before the campaign starts.
Most of this costs little. It does need to happen at the brief stage, which is why it belongs in the plan alongside locations and dates — see the guide for international brands for what a complete brief contains.
What not to over-claim
A rise in branded search during a campaign is encouraging, but it is a correlation. Other activity — digital advertising, PR, seasonality, a competitor’s news — can move the same signals. Report the movement, the comparison it is measured against, and the other activity that ran at the same time.
Be especially careful with short flights and small numbers. A two-week campaign with a few dozen scans or enquiries can show large percentage changes that mean little. Where the result matters, longer flights, larger areas and a proper control give a more reliable answer than any single dashboard figure.
Frequently asked questions
Can we track who saw our billboard?
No. Outdoor advertising in Japan is not set up for individual-level tracking; where location-based estimates are offered, they describe groups, not people. Media owners may provide audience estimates for a placement, and campaigns can be measured through response signals and area comparisons, but not person by person.
Are QR codes a good idea on outdoor creative in Japan?
Often, yes. QR codes are widely used and familiar in Japan. Make the code large enough to scan from a realistic distance, give the viewer a reason to scan, and tag it so scans can be attributed. The code and destination are also checked in the media owner’s creative review.
What should we report to headquarters after a campaign?
Delivery evidence (posting photographs or play logs), the movement in the agreed response signals against baseline, and, if one was set up, the comparison with control areas — together with any other activity that ran at the same time.
Is a brand lift study worth it for a small campaign?
Usually not. Brand lift studies need samples large enough to detect a difference and a design fixed before launch. For a small or short campaign, response signals and a simple before-and-after comparison are often more proportionate.
Pricing depends on location, format, period and seasonality, and is quoted per campaign against confirmed availability.
Japan is one of the largest outdoor advertising markets in the world, and one of the least self-explanatory from the outside. In the large cities, billboard and screen inventory is organised around railways rather than roads, most of it is sold through media owners and agencies rather than self-serve platforms, and every creative passes a review before it goes up. This guide explains how the market is structured and how an international brand actually gets a billboard, station or transit campaign live.
The most common question we receive from brands outside Japan is a price: what does a billboard in Shibuya cost, what would a two-week station campaign in Tokyo cost. The honest answer is that it depends, and this article explains exactly what it depends on, so that you can set a budget band and brief a plan that fits it. It does not contain a rate card, because none would be true on the day you book.
The digital billboards around the Shibuya scramble crossing are, in our experience, among the most requested outdoor advertising placements in Japan, and the one international brands usually ask about first. This article explains what is actually for sale there, how a slot is booked, what determines the price, which creative works on the crossing, and what to plan if the dates you need are already gone.
Read the article
Contact
Plan your OOH campaign in Japan
Tell us the objective, the city and roughly what you have to spend. We will come back with media options that fit.
You do not need a finished brief to get in touch. A rough objective and a budget range is enough for us to start proposing formats.
