Intro: It's all free water, so why does it differ from place to place?
Walk into a few restaurants in Japan and you'll notice that water is handled very differently from one to the next.
At Saizeriya, a cheap Italian family-restaurant chain, water is self-service.
You grab a glass at the drink counter and press a lever to fill it. Saizeriya's official website says as much: "water is self-service" and "the glasses are the same ones used for the drink bar."
Other family restaurants, like Joyfull and Gusto (a big family-restaurant chain run by Skylark), often fold water into the drink bar, the all-you-can-drink corner you walk up to yourself.
Even at a conveyor-belt sushi chain like Hama-zushi, water and tea lean toward self-service.
There's a cup at your table, and it's normal to make your own tea and help yourself to the soy sauce and the pickled ginger (called "gari").
At Ohsho (Gyoza no Ohsho, a casual Chinese-style diner chain) and Sukiya (a beef-bowl chain), on the other hand, a staff member usually brings you water as soon as you sit down, or you pour it yourself from a pitcher on the table.
So here's the question:
If water is free, why not make everything self-service?
If the goal is to cut labor costs, isn't the Saizeriya approach the obvious winner?
And yet even among cheap chains, water is handled in completely different ways. Why?
Here's the really interesting part: Sukiya and Hama-zushi.
If you look at the brand list from Zensho Holdings, both Sukiya and Hama-zushi sit inside the same Zensho group's restaurant business in Japan.
Even so, how they serve water, and how they move customers around, is quite different.
In other words, the way water is served isn't decided just by the parent company or the price range.
The short answer: this isn't only about labor costs.
Who brings the water is a question of store operations design: where do you put the burden, on the customer, on the staff, or on the equipment?
In this article I'll use Saizeriya, Joyfull, Gusto, Hama-zushi, Sukiya, Ohsho and others as examples, and look at the self-service water question through five lenses: store flow, service design, the "servicescape," the service blueprint, and queueing theory.
The short answer: it isn't decided by labor costs alone
When people think about how water should be served, they tend to reason like this:
If staff bring the water, that costs labor.
So self-service is cheaper.
So making everything self-service is the rational choice.
Real store management isn't that simple.
Self-service does cut down the time staff spend carrying water.
But it creates other costs in return:
- Cleaning the water station
- Restocking glasses
- Restocking ice
- Dealing with spills on the floor
- Helping kids and older customers
- Answering "where's the water?"
- Congestion in the aisles
- Customers bumping into each other
- Crowding around the drink bar
- A bad impression when the water station gets messy
So self-service doesn't make staff work disappear. It just moves the work to a different place: the customer, the equipment, or another part of the staff's job.
That's why the number that matters isn't labor cost on its own but total cost.
How far staff walk
How far customers walk
Aisle congestion
Floor space
The type of customers
How long they stay
How much each customer spends
How fast tables turn over
Cleaning workload
Complaint rate
The feeling of being served
How well it fits with the drink bar and serving robots
You have to weigh all of these together.
Self-service water is less a labor-saving trick than one piece of how the whole service system is designed.
Theory 1: The servicescape. A store's physical setting changes how customers and staff behave
The first useful idea is the servicescape.
A servicescape means the physical environment of a business, such as the layout, equipment, lighting, sound, aisles, seating and signs, and the way it shapes what customers and employees do.
In Bitner's well-known paper "Servicescapes: The Impact of Physical Surroundings on Customers and Employees," the physical environment of a service is shown to affect how customers and employees react and behave.
Apply that to the water question and it looks like this.
At Saizeriya, the water station, the drink bar, the glasses and the customers' walking paths are all one integrated space.
Customers use the restaurant assuming they'll walk over to get drinks and water.
The water station was designed from the start as something customers touch themselves.
That's why self-service water feels natural there.
Try the same thing at a place like Ohsho or Sukiya, and it doesn't necessarily feel natural.
The shop is small.
The path where staff carry food crosses the customers' paths.
Many people eat quickly and leave.
The place wasn't built for people to walk over to a water station.
And at a Chinese-style diner or a beef-bowl shop, customers expect to sit down and start eating right away.
In that setting, self-service water isn't "streamlining." It can actually get in the way of both customers and staff.
In other words:
Whether self-service water works depends not on the water station itself but on the servicescape of the whole store.
Theory 2: The service blueprint. The work of serving water doesn't disappear, it just moves
Next up is the service blueprint.
A service blueprint is a way of laying out a service visually, split into layers like these:
- What the customer does
- What staff do where the customer can see
- What staff do out of the customer's sight
- Back-end support processes
- Equipment and systems
Shostack's "Designing Services That Deliver" is a classic paper that showed how a blueprint lets you test assumptions and weigh trade-offs between cost and benefit when designing a service.
Look at serving water through a blueprint, and here's how staff-served and self-service compare:
| Task | Staff-served | Self-service |
|---|---|---|
| Pouring the water | Staff | Customer |
| Carrying it to the seat | Staff | Customer |
| Managing the water station | Staff | Staff |
| Restocking glasses | Staff | Staff |
| Handling it when water runs out | Staff | Staff |
| Cleaning up spills | Staff | Staff |
| Telling customers where the water is | Staff | Signs + staff |
| Managing crowding near the water station | Rarely an issue | Staff |
The key point is that going self-service doesn't bring staff work down to zero.
The pouring moves to the customer.
But keeping up the water station, restocking glasses, giving directions, cleaning and handling problems all stay with the staff.
So going self-service really means this:
You aren't getting rid of the work. You're changing where the work sits.
At a place like Saizeriya, which already has a drink-bar flow and can manage the water station together with its other self-service equipment, the benefits of self-service show up easily.
At a restaurant that builds a water station just for self-service water, on the other hand, new management costs can pile up: directions, cleaning, restocking and crowding.
Theory 3: Store flow and facility layout. Get customers moving and the aisles jam up
The self-service water question is also a question about store flow, meaning how customers and staff move around inside the restaurant.
A restaurant has at least these kinds of movement:
- Customers going from the entrance to their seats
- Customers heading to the restroom
- Customers heading to the drink bar
- Staff carrying food from the kitchen to the tables
- Staff carrying used dishes away
- Customers going to the register
- Serving robots delivering food
With self-service water, customers have to get up and go from their seat to the water station.
When that happens, things like this follow:
- Customers walk through the aisles
- They carry glasses back to the table
- They might spill
- Kids run around
- Older customers walk slowly
- They collide with staff carrying food
- They get in the way of serving robots
- The area near the register and the drink bar gets jammed
So self-service water reduces how much staff have to move, but it increases how much customers move.
That's why in a store with poor flow, going self-service creates a traffic jam instead.
Saizeriya and other family restaurants can set up a space where customers getting up is expected fairly easily.
Drink bars, restrooms, cutlery and other self-service gear are built into the store's flow.
In a narrow beef-bowl shop or Chinese-style diner, though, aisles get blocked easily when customers get up for water.
So:
Whether self-service water makes sense depends not on whether you have a water station, but on whether the store can handle customers walking around without jamming up.
Theory 4: Queueing theory. Cutting the wait for staff may just add a wait at the water station
Queueing theory also applies to the water question.
Queueing theory is a way of analyzing situations where people or things wait for a service.
Waiting at the register, waiting to order, waiting for food and waiting to be seated are typical examples.
At first glance, going self-service seems to cut the wait for staff.
But other kinds of waiting can pop up in its place:
- Waiting for the water station
- Waiting for a glass
- Waiting for ice
- Waiting in the crowd around the drink bar
- Waiting for the aisle to clear
- Waiting on people with kids or older customers moving slowly
The waiting doesn't vanish. It just changes where it happens.
At places like Saizeriya and Joyfull, where there's a drink-bar culture and customers are used to getting up now and then, that kind of waiting is easier to put up with.
But at a place like Sukiya, where the pattern is "sit down, order, eat fast, leave," the extra step of going to get water can make the experience worse.
In a quick-stay restaurant, the cost to the customer of getting up from the seat is relatively large.
That's why it makes more sense for staff to put the water down first.
Theory 5: Self-service and customer participation. Is this a kind of restaurant where you can hand work to the customer?
Self-service water is, broadly speaking, a form of self-service.
Self-service means the customer takes on part of the service.
For example:
- Going to get your own water
- Using a drink bar
- Ordering on a touch screen
- Returning your own dishes
- Paying at a self-checkout
- Using mobile ordering
What matters here is that self-service isn't just "it's cheap, so let the customer do it."
Can customers accept that task naturally?
Does it fit the atmosphere of the place?
Does handing that task to the customer really make the whole experience better?
Those are the questions.
At Saizeriya, self-service water feels quite natural.
Low prices, a casual vibe, the drink bar and a self-service style of use already fit the experience there.
It's easy to buy into the idea that "I'll get my own water if it keeps prices low."
At Hama-zushi, self-service also feels natural.
You order on a touch screen.
The sushi arrives by a conveyor lane or a delivery system.
You make your own tea at the table.
You help yourself to soy sauce and pickled ginger.
In that setting, handling your own water and tea fits how the restaurant is used.
At Ohsho and Sukiya, though, self-service water doesn't always feel natural.
Ohsho has the feel of a neighborhood diner or Chinese restaurant, with staff serving you.
Sukiya has a fast-food feel, where everything is done right at your seat.
If someone says "the water's over there," some customers may find that cold.
In short, whether to go self-service is decided not just by work efficiency but by whether it fits that restaurant's idea of service.
Why Sukiya and Hama-zushi, under the same parent company, split on water
Sukiya and Hama-zushi are both brands of the Zensho group.
But they don't handle water the same way.
That's because, even under the same parent, each restaurant type wins in a different way.
How Sukiya wins
Sukiya is a place to eat beef bowls and set meals quickly and easily.
- Lots of solo diners
- Short stays
- Everything is easy to finish at the seat
- Fast turnover matters
- It feels more natural not to make customers walk around
- Putting water down first keeps things smooth
In this case, having staff put down water first makes for a more natural experience than making customers stand up for self-service.
How Hama-zushi wins
Hama-zushi is a conveyor-belt sushi place with a strong culture of touch screens, lanes and self-service at the table.
- Lots of families
- You make tea at the table
- You order on a touch screen
- You use the soy sauce and pickled ginger yourself
- The sushi arrives by lane or system
- Customers are expected to take part in how the restaurant runs
In this case, handling water and tea yourself feels natural.
So even with the same parent company:
Sukiya is designed so you finish at your seat and eat fast.
Hama-zushi is designed so customers operate things at the table.
That's why the way they handle water splits too.
Is the Saizeriya approach the best one?
The Saizeriya approach is quite rational.
Water is self-service.
Glasses are shared with the drink bar.
Hand towels are at the drink counter.
Cutlery leans toward self-service too.
Self-ordering is spreading as well.
As service design for keeping prices low, this is very strong.
That said, the Saizeriya approach isn't necessarily the best one for every restaurant.
Conditions where the Saizeriya approach fits are roughly these:
- A relatively large store
- A drink bar
- Customers are expected to get up from their seats
- You can build a flow to the water station
- Customers are used to self-service
- A culture where people accept self-service in exchange for low prices
- Relatively long stays
- Some customers drink water many times
Conditions where it doesn't fit are these:
- A small store
- Narrow aisles
- Busy paths for carrying food
- Short stays
- Lots of solo diners
- Lots of older customers
- Making someone stand up just for water feels off
- "Sit down and eat right away" is part of the value of the experience
In short, the Saizeriya approach makes sense as part of how Saizeriya builds its stores.
If you take just the water out and transplant it to another restaurant, it won't necessarily work.
What did the serving robots change?
Lately you see serving robots more and more at places like Gusto.
At the Skylark group (the company that runs Gusto), large-scale rollout of cat-shaped serving robots reportedly started around November 2021, and by December 2022 they had finished deploying them across 2,100 stores nationwide.
This is connected to the water question too.
Once robots come in, the division of work inside the store changes like this:
- Carrying food: robot
- Detailed explanations to customers: people
- Water and drinks: customer, or self-service equipment
- Paying: self-checkout, people or an app
- Clearing dishes: people plus robot
- Handling problems: people
So even with robots, not everything becomes automated.
If anything, the design gets finer about
where to put each task: with people, customers, robots or equipment.
At a family restaurant like Gusto, a hybrid setup comes together easily:
Robots carry the food
Customers get their own drinks
People handle the fine details
In small stores or places where customers stay briefly, though, the paths of robots, customers and staff may interfere with one another.
So whether it's robots or self-service water, in the end it comes down to store flow.
The grandma at Ohsho who kept apologizing: refilling water is a small task, but customers can see it
At an Ohsho, the water pitcher or jug on the table was empty.
So an older part-time staff member kept apologizing: "Sorry about that, just a moment, please."
Seeing a scene like that makes you feel a bit sorry for her.
But this isn't really that person's fault. It's a problem with how operations are designed.
Refilling water is a small task.
But the moment it runs out, customers can see it.
- No water
- No glasses
- No ice
- The pitcher is empty
- The water on the table has gone warm
- You have to call a staff member over
When that happens, customers notice right away.
At the same time, refilling water doesn't directly bring in sales.
Compared with taking orders, serving food, handling payment, clearing tables, dealing with complaints and answering the phone, it's easy to put off when things are busy.
As a result, a small hole called "refilling the water" ends up being plugged by the people on the floor apologizing.
That's pretty symbolic.
A tiny gap in operations is being handled with personal consideration and apologies.
Once you see this, it stops being just "there's no water" and starts looking like a store design problem.
A checklist for deciding whether to go self-service with water
Whether a restaurant should make water self-service can be thought through with the checks below.
Customer flow
- Can customers walk to the water station safely?
- Are the aisles too narrow?
- Does it collide with the path for carrying food?
- Is it easy for kids and older customers to use?
- Does it interfere with the paths to the drink bar and the restroom?
Staff flow
- Will staff walking distance really go down?
- Will restocking and cleaning the water station add too much work?
- Will questions like "where's the water?" increase?
- Can staff keep up with managing the water station when it's busy?
Store equipment
- Is there room to put a water station?
- Is there a place for glasses?
- Is it easy to drain and clean?
- Is the floor safe when something spills?
- Are the signs easy to understand?
Customer types
- Lots of solo diners?
- Lots of families?
- Lots of older customers?
- Lots of people who stay a long time?
- Are customers used to self-service?
Type of restaurant
- Is there a drink bar?
- Is there touch-screen ordering?
- Is there a conveyor lane or delivery system?
- Do customers accept self-service in exchange for low prices?
- Does the place put weight on a diner-style, staff-served feel?
Brand experience
- Does self-service water risk looking cold?
- Is staff-served water becoming over-service?
- Can customers think, "Of course it's self-service at a place like this"?
- Does it fit the restaurant's prices, atmosphere and customers?
For a restaurant that passes these checks, self-service water makes sense.
For a restaurant that doesn't, serving by staff or putting a pitcher on the table may make more sense.
Summary: self-service water isn't "self-service because it's free." It's the result of store design
How a restaurant serves water isn't simply a labor-cost question.
Self-service water isn't just "it's free, so let the customer do it."
What it really comes down to is this:
Where should the load go, on customers, staff, equipment or robots, so the whole store jams up the least?
Saizeriya has a servicescape built on the assumption that customers move around.
Drinks, water, hand towels, cutlery and the ordering flow are all gathered together, so self-service water is natural.
Sukiya is designed so you finish at your seat and eat fast.
Having staff put water down first fits the short-stay experience better.
Hama-zushi has a culture where customers operate things themselves: touch screens, lanes, tea at the table, soy sauce and pickled ginger.
That's why self-service water and tea fit well.
Ohsho is a hybrid of a diner-style, Chinese-restaurant feel with a small interior, busy paths and refilling at the table.
When a hole opens up in water refilling, it gets plugged by the apologies of the people working there.
In a single sentence:
The problem isn't who pours the water. It's where you put the whole store's work.
That's what's really behind the self-service water question.
The Saizeriya approach is rational.
But it can't be transplanted as-is into every restaurant.
Self-service water isn't just a way to save money. It's a philosophy of store design.
References
Saizeriya, "How to use Saizeriya" (official website)
https://www.saizeriya.co.jp/how-to-use-saizeriya/Zensho Holdings, "Brand list"
https://www.zensho.co.jp/jp/brands/Joyfull, "Drink bar Joy Cafe"
https://www.joyfull.co.jp/menu/134Bitner, M. J. (1992). Servicescapes: The Impact of Physical Surroundings on Customers and Employees.
https://eli.johogo.com/Class/jm-1992.pdfShostack, G. L. (1984). Designing Services That Deliver. Harvard Business Review.
https://hbr.org/1984/01/designing-services-that-deliverSkylark, article on its cat-shaped serving robots
https://cms-recruit.skylark.co.jp/company/post-20230526001/Katsutoshi Yada, "Customer flow analysis and string analysis in supermarkets"
https://www.ism.ac.jp/editsec/toukei/pdf/56-2-199.pdf
