Pricing did not fit small restaurants
Taiwan's reservation market was dominated by platforms with fixed monthly fees and minimum booking volumes that independent restaurants could not always justify.
A customizable reservation platform that lets independent restaurants take deposits, curb no-shows, and run reservations their own way.
Eatsy gives independent restaurants a flexible reservation system without forcing their operations into enterprise pricing or rigid booking rules.
I designed the core restaurant workflows, redefined the visual identity, and used interview feedback to reduce the manual work behind bookings, deposits, and guest communication.
Every control carried the same weight, so nothing on the screen told you where to look first. Status survived only as small tags. And the top navigation ate the height the floor plan needed, capping how many tables staff could see at once. Customers called it dull.
The product was built on MUI, so the redesign started there rather than reinventing a component library three engineers would then have to maintain. The new brand red came straight out of MUI's own palette — red 400, #EF5350 — which meant hover, disabled and focus states arrived with it instead of needing to be specified one by one.
Three providers controlled nearly 90% of Taiwan's online reservation market, but their models did not reflect how independent restaurants actually operate.
Taiwan's reservation market was dominated by platforms with fixed monthly fees and minimum booking volumes that independent restaurants could not always justify.
Lean teams handled reservations, modifications, cancellations, and deposit reminders through phone calls and messaging apps.
Private kitchens, high-value restaurants, and small neighborhood businesses needed different rules, schedules, and payment workflows.
Restaurants can control who may reserve, which tables are available online, when booking windows open, and when deposits are required. The system adapts to the business instead of asking the business to adapt to software defaults.
I ran interviews with owners alongside our CEO, some over video and some in their dining rooms, and sat with them during service rather than asking them to describe it afterwards. As the redesign took shape it went back to the same restaurants, so the people who named the problems were the ones judging the fix.
Owners named no-shows as their most painful problem and the one they could do least about. The table is held, the prep is done, the covers are gone, and nothing in the product so much as recorded that it had happened.
Direct bank transfers avoided third-party payment fees, but left owners chasing customers and reconciling incomplete information across Messenger, Instagram, and Line. We automated the reminders, standardized transfer confirmation, and routed payment details back into the dashboard.
For everyone else, restrictions instead of money. An owner can block a repeat no-show from booking online, permanently or for a while, without ever charging a regular in advance.
Two routes, because restaurants did not want the same one. Deposits for those willing to ask for money up front, and a visit history that makes a repeat no-show visible so an owner can stop them booking online.
To find out whether a booking carried a note, staff had to open it. Finding the one that mattered meant doing that to every card on the timeline, mid-service, on the one screen they watch all day — so in practice nobody checked, and the schedule told them very little.
The labels and notes are pulled straight onto the card rather than living a click away. Whether a guest is a VIP, wants a window seat, cannot eat crab, or has left a note that needs attention before the food goes out is legible from the timeline itself — nobody has to open a booking to find out whether this is the table that needs something.
Status, party size, guest tags and notes moved onto the face of each booking, so the timeline answers the common questions without being opened at all.
Unlike a chain, an owner remembers the regular who wants a window seat and the one who cannot eat crab. That closeness is what they compete on — and they were holding it in their heads, or on paper, because the product gave them nowhere to put it.
Labels, preferences, allergies and notes sit in the dashboard sidebar, next to the reservation they belong to. Nobody has to open the CRM while the guest is already standing in front of them — by the time it would be useful, the moment to use it has passed.
CRM stopped being a side feature. Guests carry tags, preferences, allergies and visit history, and those follow them onto the booking where staff are already looking.
The work does not disappear, it moves to the restaurant. Someone has to open their bank app, export or screenshot the transfer record, and upload it before anything reconciles. Owners told us plainly that this was annoying, and they were right — it is a small chore that never goes away.
They took it anyway, because the alternative was a processor taking a percentage of every deposit. Third-party processing priced the feature out of reach for exactly the restaurants that needed it most, which would have left it decorative. Our CEO and I worked out the cheaper route together and put it in front of owners while we were still building it, rather than shipping it and hoping.
Every rule a restaurant can change is a rule it can set wrong. The support that created did not land on a queue somewhere — it landed on our CEO, driving out to restaurants to answer setup questions in person.
A private kitchen and a neighbourhood restaurant genuinely do not run the same way, and defaults that suited one made the product useless to the other. Flexibility was the reason to choose us over the incumbents, so the configuration cost was the price of the pitch.
After a few of those visits we built an AI support layer to absorb the basic questions. The trips did not stop, but they stopped being about the same setup query every time.
Existing staff had to relearn where everything lived, mid-service, on a product they already knew — and we heard about it. The complaints during the transition were real, and they came from our longest-standing restaurants.
We went back and forth on this one longer than on anything else. What settled it was how the two are used: staff watch the timeline all day and touch the navigation a few times a day, and the top bar was spending vertical space the floor plan needed. Restaurants that arrived after the change reported the opposite of the complaints — that the schedule finally had room to work in.
The work showed up in who signed and in how often a booked table stayed empty. Owners also told us service ran faster once the timeline carried what they needed — a number we never captured, so it stays a sentence rather than a figure.
A reason to switch
restaurants that would not hand a processor a cut of every booking could finally take a deposit
13% → 8%
average no-show rate across restaurants taking deposits

Wakatake
“Eatsy offers flexible pricing that aligns with our operating rhythm. We have full control over when and how reservations open, and the booking process becomes part of the customer experience.”

Yang Er Lou
“I used to handle every message myself and sometimes missed bookings. After switching to Eatsy, guests can reserve on their own, so I can focus on prep work and truly rest on my days off.”