Restaurant site selection: Tokyo station areas, footfall by meal and the rivals next door
How to pick a restaurant site in Tokyo: which kind of station area suits your meal, how to read footfall by hour, and how to count rival restaurants by cuisine.
Part of the guide AI employee for site selection
For restaurant site selection, Tokyo asks one question before any other: which station area, and does its crowd eat when you serve? Shortlist station areas whose busy hours match your meal (lunch, dinner or late night), count the restaurants of your own cuisine within a short walk, and only then compare the units for rent. A good street near the wrong kind of station will be full of people who are not hungry when your kitchen is open.
This guide is for restaurant groups, franchise teams and chefs opening their own place in Tokyo. It sorts station areas into kinds, shows how to read footfall meal by meal, explains how to count rivals by cuisine rather than as one big number, works through a made-up comparison of three areas, and ends with an AI employee that scores candidate sites for you every month.
Four kinds of station area, four kinds of diner
Tokyo restaurants live off the people who come out of the ticket gates, and different stations release different people at different hours. Before looking at any listing, decide which kind of area your restaurant belongs in.
- Business district stations. Packed at weekday lunch with office workers who have under an hour, quieter in the evening and very quiet at weekends. Good for set lunches, bowls and noodles that turn tables fast.
- Large terminals and shopping districts. Busy from late morning to late evening, every day. Strong for almost any format, but rents and rivals are both at their highest.
- Nightlife areas. Slow by day, busiest from early evening until the last trains. The natural home of izakaya, yakitori and bars with food.
- Residential stations. A morning and an evening rush of commuters, families at weekends. Suits neighbourhood restaurants that people return to every week.
Reading footfall meal by meal
Footfall shows whether a spot is busy at the hours that matter to you. Mobility Footfall gives, for each candidate address, footfall per day, the busiest hour, the busiest and quietest days of the week, and how the day splits between morning, midday, evening and late night. It is measured in people-hours of presence, so a street where people stay counts more than one they hurry through, and each day arrives about ten days late. Read it as the shape of a recent week rather than a trend over seasons; the history only starts in mid-October 2026.
The split by part of the day is the number to look at first. A ramen counter in a business district wants a tall midday share and can live with a thin evening. An izakaya wants the evening and late-night shares to dominate, and a busiest day of Friday or Saturday. A neighbourhood bistro wants a balanced day and busy weekends. If the shape does not match your meal, a high daily total will not save the site.
Counting rivals by cuisine, not in one big number
Tokyo has restaurants everywhere, so a raw count of all of them near a unit tells you very little. What matters is how many serve your kind of food within the distance a diner will walk for it. xMap's places database has types for ramen restaurants, sushi and conveyor belt sushi restaurants, izakaya and many other cuisines, and Nearby places can count each type within a set distance of every candidate, with the nearest one, its distance, the average rating and how many belong to a chain.
- Same cuisine, close by. Several well-rated ramen shops within 300 metres can mean a street people already visit for ramen, which helps a strong newcomer, or a market that is full. Read the ratings and the number of reviews before deciding which.
- Chains against independents. A street of chain outlets competes on price and speed; a street of independents competes on character. Know which fight you are joining.
- Complementary food. Cafes, dessert shops and bars nearby keep people in the area after a meal and are rarely your direct rivals.
- New openings. Places with a recent opening date show where other operators are betting right now.
Three station areas, a made-up comparison
Say a team wants to open an izakaya and has three areas on its shortlist. The figures below are invented to show the method, not taken from any real station.
| Area | Busiest part of the day | Busiest day | Izakaya within 300 m | Their average rating |
|---|---|---|---|---|
| Business district station | Midday | Wednesday | 4 | 3.6 |
| Nightlife area station | Evening and late night | Saturday | 19 | 4.1 |
| Residential station on two lines | Evening | Friday | 3 | 3.9 |
A made-up example. Footfall shape from Mobility Footfall, rivals counted by Nearby places.
The business district has few izakaya for a reason: its crowd goes home after work. The nightlife area has the right crowd but nineteen established rivals rated above 4. The residential station has an evening peak, a Friday high and only three izakaya, none of them especially loved. The team chooses the residential station, then filters ground-floor units within five minutes on foot of its busier exit. A popular place in the nightlife area might be the second site, once the brand has regulars who will follow it.
Mistakes that cost Tokyo restaurants
- Chasing the daily total. The biggest number belongs to the stations whose crowd you may not be able to feed.
- Counting all restaurants as rivals. A soba shop and a steakhouse do not take each other's diners.
- Ignoring the floor. Many units by the busiest exits are upstairs or in basements, which suits a destination restaurant and hurts one that needs walk-ins.
- Forgetting the last train. Late-night trade depends on how late people stay before heading home, which the late-night share of footfall shows.
- Deciding from one visit. A Tuesday afternoon walk tells you nothing about Friday night.
An AI employee that scores your shortlist
Scoring three areas by hand is fine; scoring thirty, and doing it again each month as new restaurants open, is a job for an AI employee. The guide to AI employees for site selection explains the general approach. A Tokyo restaurant flow could be:
- Schedule: the first Monday of each month, Tokyo time.
- My data: your shortlist of station exits or candidate addresses.
- Mobility Footfall: footfall, the busiest hour and day, and the split by part of the day around each one.
- Nearby places: the restaurants of your cuisine within 300 metres, counted, with their average rating and chains.
- xMap LLM: a score out of 100 with a reason, told which meal you serve and how much you fear strong rivals.
- Sort and limit to the best ten, then Email with a map.
Each month the email lists your best candidates with their busiest hours, their rivals and a sentence on why each ranked where it did. When you are ready to look at units, the Shop space for rent near stations in Tokyo template brings new listings within a short walk of chosen stations, and New restaurants opening near my branches keeps an eye on rivals once you are open. For the listing side in more depth, read choosing a shop near a Tokyo station.
Questions
Which Tokyo station areas are best for a lunch restaurant?
Business district stations, where office workers fill the streets at weekday midday. Check the midday share of footfall and expect quiet weekends.
How do I know if a street has too many restaurants like mine?
Count only your own cuisine within the distance diners walk, then read their ratings and reviews. Many strong rivals means a full market; many weak ones can mean room for a better one.
Can I see how busy a street is at night?
Yes. Mobility Footfall splits each day into morning, midday, evening and late night and names the busiest hour, so you can see whether people stay out after dark.
Is population data available for Tokyo?
It may not be. Run Demographics on one Tokyo address first; in central Tokyo, footfall and the kind of station are usually more useful than residents anyway.
Templates that do this
Read next
More on site selection and expansion · Every article · Templates