Loading page…
A Japanese restaurant reservation platform that wins deals by telling restaurants their current vendor is holding their customer data hostage — and lets anyone export everything, any time, with no contract lock
about $10M raised in total across 11 years, of which the Series A was only $2M · fixed monthly fee tiered by restaurant size, no cover fees and no transaction fees
"It's like going out to a country where nobody's wearing shoes — but once everybody starts wearing shoes, it's much easier to sell better shoes." His counterintuitive point: first-mover advantage was the hard version of this business, and competitors arriving made selling easier, not harder.
Fewer bars = easier, cheaper, or faster for an AI-assisted solo builder. Editorial judgments based on the case details.
Yu Taniguchi established the company that became TableCheck in 2011 and has been growing it steadily for eleven years, on roughly $10 million of total funding. The product is a table management system — reservations, CRM, billing and more, for restaurants. When he last appeared on this podcast four and a half years earlier, he had about 2,000 restaurants and his biggest competitor was paper and pencil. At the time of this interview he has about 7,000 restaurants in Japan and overseas, having more than tripled while two of those four years were spent inside the pandemic — for a product sold exclusively to restaurants, which is worth pausing on. His read on how the market changed is the opposite of what most founders would say. Japan now has two or three competitors and globally there are around ten major ones; restaurants increasingly treat a TMS as standard operating equipment. The competition is fiercer, and he says selling is easier: convincing a restaurant that your system is better than the one they already use is less work than convincing them to stop using paper. His analogy is the cleanest statement of the anti-first-mover case we have on file. Our read: this is a founder who has been in a category long enough to have watched the education burden move from his side of the table to his competitors', and who correctly identifies that as a gift rather than a threat.
TableCheck started in 2011 in a Japanese market where the competition was paper and pencil, so the first accounts had to be argued into existence rather than won away from a rival.
The quick-reference card — acquisition channels, replication playbook, risks & traps, and our verdict — is members-only.
Join the Unicorn Club — $5/moCancel anytime, in one clickPaste into Claude Code / Codex and get a working version of this product end-to-end.
Two production-grade prompts per idea — a full build spec and an SEO growth plan — plus 10 tailored build specs a week. Members only.
Join the Unicorn Club — $5/moCancel anytime, in one clickEvery claim traces back to a listed source — open any of them and check the number yourself.
Data credibility: Founder interview in English on the Disrupting Japan podcast, episode 194. Every figure is founder-stated in conversation; no dashboard, filing or third-party data is shown or cited. Founder-stated and unaudited: about 7,000 restaurants using the product versus about 2,000 four and a half years earlier, roughly $10M raised in total since 2011, a $2M Series A, pricing as a fixed monthly fee banded at up to 30 seats / 31–100 seats / 100+ seats with no cover or transaction fees, two or three competitors in Japan and around ten globally, eighteen supported languages, hotel-brand certification taking one to two years with certified lists of three or four vendors per region, and operations in Korea, Singapore, Thailand, Australia, mainland China, Dubai and Indonesia. ⚠️ Note what is absent: no revenue figure, no ARR, no growth rate, no churn, no ARPU, no margin and no headcount appear anywhere. The 7,000 figure is paying restaurants on a subscription, which is why this entry qualifies on traction rather than on revenue. The claim that TableCheck is the only vendor in the category with size-tiered fixed monthly pricing is his own assertion about competitors and is not independently verified. His characterisation of OTA contracts — that customer data belongs to the OTA and cannot be exported on exit — is a competitor claim made in a sales context and should be verified against actual terms before being republished as fact. The interview is dated: he refers to eleven years since 2011 and to the pandemic in the recent past, so treat it as a snapshot from roughly 2022 rather than a current statement. Auto-caption garbling: the company is rendered "table Che" and "Che" throughout; the founder's name appears as "UT tanaguchi" and "you"; "gurunavi" is Gurunavi, "chop" is Chope, "rest pack" and "seven rooms" are ResPak and SevenRooms, "raken" and "mercur" are Rakuten and Mercari, "piku media" is an unresolved former employer, and "nobuyuki id" is Nobuyuki Idei. Verify all names before publication.
The old price card tells you what the first invoice looked like. TableCheck used to charge an initial setup fee plus a menu of optional per-feature monthly fees, so the first money through the door was an installation charge paid up front by a single restaurant, with a smaller recurring line starting the month after.
Two engines started compounding, and the first is counterintuitive. Taniguchi says competitors arriving made selling easier, not harder: instead of convincing a restaurant that a table management system is worth having at all, the pitch became that TableCheck beats the system already installed.
The lean stretch here has two shapes. The first is the early years: VCs were sceptical of the business model, so the Series A came to only $2 million, and total funding across eleven years is about $10 million, a pace that forces a company to find out whether customers will pay before investors will.
Reservations, customer data and operations for restaurants — deliberately positioned as a tool, not an ecosystem, and that distinction is the whole strategy. Two things differentiate it, both unglamorous. First, localisation done seriously: eighteen languages, by far the most in the market, and localisation of functions and features rather than just interface strings. Second, integrations with the dominant local point-of-sale and payment vendors in each country, on the observation that his competitors enter new markets supporting only the globally usable POS and payment providers. That is a boring moat and a real one — every market has entrenched local vendors, and the competitor who only speaks Stripe and Square cannot serve them. Third, and structurally most interesting, he refuses lock-in on purpose. He states the goal directly: they do not want to bind merchants with contracts, they want to be sticky through the value they provide. Data export is available at any time. There are no strong contractual limitations on leaving. Of course they want a lower churn rate, he says, but they do not force restaurants to keep using them. The host names what this means in SaaS terms — they are selling tools rather than owning a marketplace and the customer relationship — and Taniguchi agrees, while adding that the roadmap intends to earn the ecosystem position rather than contract for it, by eventually connecting diners and restaurants with marketing automation built on the restaurant's own data.
A fixed monthly fee, tiered by restaurant size, and nothing else — and he claims to be the only player in the category doing it. No cover fees, meaning no per-reservation transaction charge when a restaurant takes a booking through their system. They used to charge an initial setup fee plus a menu of optional per-feature monthly fees, and have collapsed all of that into one monthly price banded by seat count: up to 30 seats, 31 to 100 seats, and 100-plus. His stated reason for believing this is the right structure is that every other TMS vendor charges a monthly fee regardless of how large the restaurant is, which means small restaurants overpay and large ones underpay relative to value received. The host's observation is the important one: this aligns the price with actual customer need rather than with the vendor's collection opportunity. Note what he is giving up. Per-reservation fees are the highest-leverage revenue line in this category, and the OTAs charge them — so TableCheck is deliberately declining the revenue model that the incumbents built their businesses on, in exchange for being the vendor that does not tax the restaurant's own traffic. No revenue figure appears anywhere in the interview; the only money numbers given are the funding total and the Series A size. Our read: 7,000 restaurants on a size-tiered monthly fee is a real, calculable business, and the absence of a stated revenue number is a choice on his part rather than a gap in the conversation.
The primary sales weapon is not a feature, it is a warning about somebody else's terms of service. In this market the incumbent TMS products are often supplied by OTAs — the online restaurant-listing platforms, which he names as Yelp in the US, Gurunavi in Japan and Chope in Singapore. Those contracts, he says, state that the customer data belongs to the OTA and do not allow the data to be exported when a restaurant decides to leave. So TableCheck's sales message to restaurants is that using an OTA's TMS is dangerous: the longer you use it, the more of your customer data accumulates inside it, and when you eventually want to leave, they have all of that data as hostage. The host adds the part that makes it work — restaurants signing up do not realise this, because it is buried on page 35 of a 75-page licence agreement. Taniguchi credits this message directly for the pace of market-share expansion. The second distribution engine is global hotel brands, and the mechanics are worth understanding because they are a real moat. A global hotel chain restricts which TMS each property may use, decided at headquarters from a certified vendor list. Getting certified means the HQ runs a research project comparing you against OpenTable, SevenRooms, ResDiary, ResPak and every other vendor in the world. That certification process takes one to two years per hotel brand, and the resulting certified list is only three or four vendors per region. TableCheck holds APAC certification for one brand and global certification for others. Once you are on the list, every new property in that chain is an addressable deal you do not have to fight for from cold.
This is the section to read twice. Around $10 million raised over eleven years, and the Series A was only $2 million because VCs were sceptical of the business model. Once the model proved out, the dynamic inverted: investors came to him asking why he was only raising 5 million, why not 20, why not 50. He had already tested the answer — they experimented with spending on advertising and marketing and it simply proved ineffective in this market, for reasons covered in the next section. He compares himself to competitors founded around the same time with similar headcount and a similar client base who raised five times more money, or more, and burned a lot of cash. His verdict on them is not moral, it is operational: it looks fancy, you can run a lot of advertising, and during COVID he realised that staying lean and close to profitability is very important, because if you are burning cash when the market turns you are at high risk of failing to raise and running out. "This is my third bubble," he says, "and you just never know." So the operating discipline is concrete and copyable: they always design two business plans. Plan one assumes they cannot raise any additional finance. Plan two assumes they can. Even after raising on plan two, they can always fall back to plan one, which keeps them profitable. On equity he is equally clear-eyed. Giving away control destroys the purpose of founding a company; he founded one because as an employee his proposals were never reflected. He cites an early employer in the Rocket Internet era that handed 51% to a VC and saw all its founders ousted in under six months.
Four and a half years ago they had one overseas office, in Korea. Now they operate in Singapore, Thailand, Australia, mainland China, Dubai and Indonesia. The market-selection logic is admirably mundane. Asia grows faster than the EU or US. Time zones matter operationally: their support team speaks English and can serve customers outside Japan, but a twelve-hour offset would force them to build a second support team — so staying in nearby time zones expands the market with the least investment. Total addressable market is measured concretely as how many Hyatts and Hiltons exist in each country. They prefer more mature markets where restaurants already use a TMS, so they can skip the education phase entirely, and they actively avoid markets where diners do not make reservations at all and simply show up and wait. The most interesting local surprise was China, where customers buy vouchers before making a reservation — paying the restaurant up front in exchange for a small discount, a behaviour he attributes to Groupon's arrival in that market. The US and Europe remain target markets with no date attached; Asia is large and fast-growing enough, a US-based competitor is currently struggling in the US and withdrawing from it, and OpenTable is losing share as that market fragments across Instagram, Google and review sites. He notes that fragmentation is good for a tool-not-ecosystem product: the more reservation channels exist, the more a restaurant needs something neutral sitting underneath them.
Copy three things. First, sell against the incumbent's contract rather than against its features — read the competitor's terms of service, find the clause that harms the customer, and make explaining that clause your sales pitch. Second, price on a dimension the incumbents ignore: size-banded fixed fees rather than flat fees or transaction cuts. Third, run two business plans permanently, one of which assumes you never raise again. Now the risks, and he does not raise most of them. Refusing lock-in means every renewal is genuinely voluntary, so your churn is a live referendum on the product forever — the strategy is only viable if you actually keep shipping. Refusing per-reservation fees caps revenue per customer, and a competitor with transaction revenue can outspend you on product for the same customer count. The hotel-certification moat cuts both ways: one to two years to get on a list is a barrier protecting you today and a wall that blocked you for two years, and losing a certification would remove a whole channel at once. The industry he chose has, in the host's phrasing, the worst profit margins of any on Earth, and Taniguchi agrees he would not have picked restaurants if he wanted to get rich. Localisation depth across eighteen languages and dozens of local POS and payment vendors is a permanent, compounding maintenance cost. And there is no revenue, growth-rate, churn or margin figure anywhere in this interview — the only hard numbers are restaurant counts and funding.
The four founder playbooks — first customers, first dollar, the flywheel, the silent stretch — are in the Unicorn Club, on this idea and on all 1,403 others. The story, the numbers, the sources and every later section stay open to everyone.
Join the Unicorn Club — $5/moCancel anytime, in one click