Zendesk Japan: From SMB to Enterprise — and a Warning for AI Entrants

Written By Murray Clarke

Zendesk entered Japan in 2013, won its first customers in the SMB and commercial market, and only reached real enterprise scale after a patient five-year build and a deliberate addition of senior leadership. The short version for anyone planning a Japan entry in 2026: that patience was not a bug in the plan. It was the plan.

Most foreign SaaS companies that "win Japan" don't win it the way the pitch deck promised. They arrive expecting the logos to fall in the first eighteen months, discover that Japanese enterprise buyers move on consensus and proof rather than momentum, and either re-scope or retreat. Zendesk is one of the cleaner case studies to learn from, because the climb is unusually well documented and most of the people who made it are still findable.

The interesting part isn't that Zendesk succeeded. It's the sequence. A near-founder figure who held a Country Manager title but operated more like a chief of staff, and then an enterprise closer brought in to do the thing the first phase structurally could not. Map those five years against what AI companies are doing in Japan right now, and the contrast is the whole point of this post.

A short caveat before the data. The TalentHub counts referenced below are directional, not a perfect census. I queried our pipeline and network data, then filtered for current and former roles with Zendesk in the position or company fields. Some profiles are pipeline records rather than confirmed current employees. The pattern is clear; the exact headcount is not.

The Context: Zendesk Entered A Market That Wasn't Ready For SaaS

Zendesk was founded in a Copenhagen loft in 2007, moved to San Francisco, and went public on the NYSE under the ticker ZEN in 2014. Japan came before the IPO, not after it — the company set up its Japan subsidiary in 2013, its fifth regional branch after London, Melbourne, Copenhagen and Dublin, in an office of barely a hundred square metres with three people in it.

The timing mattered. In 2013, "SaaS" was still a phrase you had to explain in a Japanese sales meeting. Cloud-delivered software was treated as a security question before it was treated as a productivity one, and adoption ran years behind the United States — a measured, risk-averse posture that the market has only recently grown out of. Kan Kunimura, Zendesk's first employee in Japan, said it plainly years later: "When Zendesk entered Japan, SaaS was a relatively new concept and we needed to convince people."

That is the part most entry plans underweight. Zendesk wasn't selling against a competitor in 2013. It was selling against the status quo of on-premise software and a buyer who had never run a mission-critical workflow in someone else's cloud.

The Leadership Sequence: A Builder With A Country Manager Title, Then An Enterprise Closer

Here is where the org chart and the mandate diverge — and where the real lesson lives.

On paper, Kunimura-san was "Japan Country Manager" from March 2013. In practice his mandate was a founder's: be the first person on the ground, localise the product and the message, and prove that a Japanese company would trust a foreign cloud help-desk at all. He grew the customer base to roughly 700 accounts, then moved on to lead SMB across the wider APAC region. The title said Country Manager; the job was closer to a chief of staff for the market itself — single-handedly standing up demand, evangelism and the first reference customers.

That is an interesting way to enter a market, and it works, but it takes time. A chief-of-staff entry buys you legitimacy and a base. It does not, by itself, buy you seven-figure enterprise deals, because the person doing it is busy doing everything.

The enterprise mandate arrived as a separate, deliberate hire. In November 2016, Zendesk brought in Yutaka Fujimoto as Country Manager and Senior Director for Japan — and Fujimoto-san was the first executive recruited specifically to take the business upmarket. His record is why. Before Zendesk he had built ServiceNow's Japan business from scratch as its first Country Manager, taking it from zero to multi-million-dollar revenue and landing names like Toyota, Panasonic and Nomura. Before that, senior enterprise roles at Microsoft Japan and Oracle Japan. At Zendesk his brief was explicit: move the company "from SMB to enterprise segment."

This is the Mandate Audit in one company. Same title — "Japan Country Manager" — attached to two completely different jobs. The first was land: presence, proof, the first hundreds of SMB logos. The second was expand: bigger deal sizes, recognisable enterprise names, and a full functional team underneath. Read either appointment by the title alone and you would have mis-scoped the search.

Our Read on the Sequence

When I met Kunimura-san, he was candid about the nature of his role — the Country Manager title was real, but the job was closer to a chief of staff for the market itself. First person in, building everything from nothing, proving the concept. He wasn't trying to be something he wasn't. That honesty is actually rare, and it's exactly the kind of self-awareness that makes a land phase work.

The Fujimoto placement came through our team. What made him the right call at that moment wasn't just the ServiceNow pedigree — it was the timing. He had spent three years building ServiceNow Japan's initial foundations, had stepped away, and was on a considered break when we engaged him. That pause mattered. It meant he wasn't running from something; he was choosing deliberately. And what he was choosing was a market he already understood at a structural level — not just the enterprise sales motion, but where the SMB-to-enterprise transition breaks down in Japan specifically. He had lived that inflection point once already. Zendesk was the second time.

That pattern — a builder who knows what they are, followed by a closer who has done the turn before — is not something you can manufacture on a two-year timeline. It has to be sequenced. And it has to be resourced with enough patience to let the first phase finish before the second one starts.

What Our Data Shows About How Zendesk Staffed The First Five Years

The hiring sequence in our own records tracks that two-phase story closely. Kunimura-san joins as an account executive and business-development hire in 2012, takes the Country Manager title in 2013, and runs the SMB build through to late 2016. Fujimoto-san arrives in November 2016 and the org changes shape: the brief stops being "find customers" and becomes "build a company" — marketing, inside sales, partner sales, customer success and support as distinct functions rather than one person wearing five hats.

You can see the second wave in the data. The Zendesk Japan names we hold cluster in the 2019–2022 scale years — a head of Japan sales and business development, a head of professional services carrying an 執行役員 (corporate officer) title, a marketing director who became a corporate officer in 2022, and a run of enterprise account executives and solutions consultants. Almost every Japan-based Zendesk profile we track is bilingual, fluent in Japanese first. That is the texture of an expand organisation being layered on top of a land beachhead, not a single big-bang launch.

The honest read: a founding builder, a deliberate enterprise hire three to four years in, and only then a middle layer of repeatable-motion talent. Nobody compressed it.

The Public Numbers: A Slow Curve That Compounded

The growth curve looks unremarkable year to year and impressive end to end. Roughly 700 customers when Kunimura-san handed over in 2016. By September 2018, more than 2,000 Japanese customers — including JapanTaxi, Pioneer DJ, CyberAgent and LIXIL — close to 50% year-on-year revenue growth, and Japan ranking inside Zendesk's top ten markets globally, prompting a Tokyo office expansion. By 2020, around 3,000 customers.

The mechanism behind the upmarket move is worth copying. Fujimoto-san treated going enterprise in Japan as a credibility problem before a sales problem, prioritising concise, practical case studies that let a cautious buyer build internal consensus. In a consensus market, the reference customer is the product demo. You don't argue a Japanese enterprise into a deal; you remove its reasons to say no, one proof point at a time.

The Partner Strategy Was Earned, Not Bought

There is a tempting shortcut in every Japan plan: sign a big distributor, announce it, and call the channel built. As our Box analysis put it, a distributor deal is not a commitment — a logo on a press release is not a go-to-market.

Zendesk's most durable partner came the slow way. Independent consultants who had implemented Zendesk became such committed advocates that they eventually spun up a company built almost entirely around Zendesk implementations. That firm, Eclect (エクレクト), went on to win Zendesk's GTM Partner of the Year for Japan four years running and built a packaged cloud contact-centre offering on AWS around the product. A partner that grows out of genuine product conviction outlasts one that was bought with a margin.

The Last Five Years: Taken Private, And Pulled Into The AI Era

Zoom out to the recent half-decade and the story shifts from market entry to market maturity. In November 2022 Zendesk was taken private by an investor group led by Hellman & Friedman and Permira in an all-cash deal valuing it at roughly $10.2 billion. Japan kept getting investment through the transition: by its tenth anniversary in 2023 Zendesk had committed to a second Japanese data centre in Osaka, making Japan the only country outside the United States where it runs two.

The company that spent 2013–2018 explaining what SaaS was now sells AI-era customer experience into a Japanese market racing the other way. IDC expects Japan's AI infrastructure spending to grow roughly seven-fold across three years and clear $5.5 billion in 2026. The buyer who needed convincing about the cloud in 2013 now has an AI budget and a deadline. Which is exactly where the comparison gets uncomfortable.

Why This Matters For AI Entrants: Box Did It In Ten Years, Glean Wants It In Two

Box is the patient-model peer. It established its Japan entity in 2013, the same year as Zendesk, hired a founder-style Country Manager, and ground its way up. The result today: Box Japan supports around 21,000 Japanese companies and roughly 84% of the Nikkei 225, Japan accounts for close to a quarter of Box's global revenue, and the original Country Manager has been elevated to chairman with a successor installed as president — a textbook lead-phase transition. That 0-to-10-and-beyond took the better part of a decade. We wrote the full version of that story in Dropbox vs. Box in Japan.

Now look at how the AI generation is entering. Glean — the enterprise "Work AI" company — raised $150 million at a $7.2 billion valuation in June 2025, explicitly to fund international growth. Glean had been selling in Japan since May 2023 through its first local partner, Ashisuto, and added a distribution deal with Tokyo Electron Device in September 2025, and Glean has just brought in its first Japan Country Manager — a hire we've watched closely from inside the search market. The framing in the category is open about the urgency: as TechCrunch put it, "the enterprise AI land grab is on."

So the contemporary entrant looks at Zendesk's five-year SMB-to-enterprise climb and sees a timeline it cannot afford. The thesis is that AI compresses everything: faster proof, faster budgets, faster logos, so the patient sequencing that Zendesk and Box relied on is a relic of slower times. My view is that this is half right and dangerously half wrong.

The Bigger Lesson: Patience Was The Strategy — AI Is Testing Whether It Still Works

What AI genuinely compresses is the demand side. Japanese enterprises that needed five years of convincing about the cloud now have boards asking why they don't have an AI plan yet. That tailwind is real, and an AI entrant should absolutely move faster than Zendesk could in 2013.

What AI does not compress is the trust side. Consensus buying, the reference-customer-as-proof, the preference for a vendor that has obviously committed to staying — none of that changed because the product got smarter. Zendesk's win came from sequencing those two jobs and not confusing them: a builder to earn the right to be in the room, then an enterprise leader with a real P&L mandate to convert that into scale. The mistake the land-grab framing invites is hiring one person and handing them both jobs at once, on a two-year clock, and calling it a Country Manager search.

The patient model isn't sacred. But the discipline underneath it is. Separate land from expand. Audit the mandate before you write the title. Resource the builder phase even when the AI tailwind tempts you to skip it. The companies that win Japan in the AI cycle won't be the ones that were most patient — they'll be the ones that moved fast on demand and stayed honest about trust.

If you're scoping a Japan Country Manager search right now and trying to decide whether you're hiring a builder or a closer, that is precisely the conversation we have every week — see how we approach it in our executive search practice and our note on why Japan, why now. Get the mandate straight first; the title is the easy part.

Sources

Murray Clarke
Next

Dropbox vs. Box in Japan: Why The Bigger Brand Lost The Enterprise