Why Foreign SaaS Stalls in Japan After Ten Customers

A foreign SaaS company's first year in Japan usually goes fine. Ten or twelve customers sign: subsidiaries of companies from back home, a few startups, one or two IT managers who like new things. Year two targets a hundred and ends at twenty. The sales team didn't get worse. The first ten customers and the next hundred are different animals, and in Japan the gap between them has a name and a procedure.

Where is the chasm?

Between the early adopters and the mainstream. Geoffrey Moore's Crossing the Chasm (1991, third edition 2014) makes one argument: a new technology product can sell to the small group that likes to try things first, and the same approach will fail with the much larger group that buys on practical grounds. The book is still in print after thirty years. The gap hasn't closed.

Moore built on Everett Rogers' diffusion of innovations. Rogers split adopters into five groups: innovators 2.5%, early adopters 13.5%, early majority 34%, late majority 34%, laggards 16%. Moore drew his line between 13.5% and 34%. People on the left buy possibility and will gamble on an incomplete product. People on the right buy the fact that someone like them already uses it and nothing went wrong. They want a whole product, references from their own industry, and someone to call when it breaks.

The theory is contested, and I should say so. Rogers himself rejected it: past research, he wrote, shows no support for a chasm between adopter categories, and innovativeness is continuous. I use the chasm as a sketch, not a law. In Japan the sketch happens to fit unusually well.

Who sits on the far side in Japan?

Mainstream Japanese companies, meaning the ninety you were counting on in year two. They are spending. The Japan Users Association of Information Systems (JUAS) surveyed 957 listed and comparable companies for its 2026 IT trends report. 52.6% raised their IT budget for fiscal 2025, and the budget diffusion index of 43.3 was the highest planned figure in ten years.

Look at where the money goes. The top reason for raising the budget was refreshing or expanding existing systems, cited by 66.3%. Growth in cloud services came next at 45.0%, then AI-related spending at 36.3%. Asked what DX is for, the most common answer was cutting costs in existing businesses, 36.1%. Entering new businesses or rebuilding the business model: 13.5%. On development, roughly seven in ten said they aim to combine in-house work with outsourcing. The incumbent system integrator isn't going anywhere.

So the far side looks like this. Budgets up. Money to old systems first. New tools welcome if they cut cost. Vendors unchanged. That is Moore's pragmatist. Over 80% of Japanese companies already use the cloud, and Fuji Chimera Research puts Japan's public cloud market at 3.44 trillion yen in fiscal 2024, heading for 6.25 trillion by fiscal 2030. The same release notes that large companies are adopting "Fit to Standard," reshaping their processes around SaaS defaults rather than customizing. The far side is buying. It just buys differently.

How does a mainstream Japanese company buy one SaaS product?

Through four gates, over nine to eighteen months. First, someone in a user department has to sponsor it and be able to explain internally why. Second, the IT department runs a security review: a questionnaire of dozens to a few hundred items on where data lives, who can access it, and how incidents are reported. Third, procurement and legal read the contract, in Japanese, with an SLA, exit terms and data ownership. Fourth, ringi, the circulating approval document that collects seals one level at a time. I wrote in the first-year budget piece that mid-sized companies and up take nine to eighteen months from first visit to signature. SaaS gets no exemption.

From overseas this reads as "Japanese companies are cautious." I'd put it differently. They are pragmatists in Moore's sense, and they only buy in a form where someone on their side can carry the responsibility.

There's a gate that isn't on the flowchart: the incumbent system integrator. Most Japanese companies' systems were built and are maintained by an SIer. A new SaaS product gets sold by the SIer, accepted by the SIer, or blocked by the SIer. In the country with the most century-old companies, supplier relationships are hard to pry loose, and software is no different.

Two public cases show the crossing. Slack launched its Japanese version in November 2017, and by January 2019 it was calling Japan its second-biggest market. The Japanese customers it names include Mercari and Sansan, classic left-of-chasm companies, and also Musashi Seimitsu, a manufacturer based in Toyohashi with more than 16,000 employees that rolled Slack out company-wide in 2018. Manufacturing, 16,000 people: the far side. Slack then announced a Japan data region planned for the first quarter of 2020. ByteDance's Lark is on the same road. Lark Japan keeps all data on servers inside Japan, and in March 2025 it signed the listed company NEXYZ.Group as a "platinum partner," so a Japanese company sells to Japanese companies. Both did what Moore calls the whole product: the Japanese language, local data, partners and rollout support that sit outside the software itself.

The two sides need two different sales organizations

On the left, the founder or someone sent from HQ sells directly. Show the product, find agreement, sign. The right side needs a different cast. An enterprise account executive who has sold inside large Japanese companies, knows the rhythm of ringi, and can draft the paper the customer's champion has to submit internally. A partner manager whose whole job is SIers and resellers. A presales engineer who answers the questionnaire and writes the technical documentation. A customer success manager who owns rollout, training and the first renewal. The left side needs none of these four roles. The right side can't skip any.

One condition is harder than hiring: the Japan entity must be able to set service levels and partner terms itself. Far-side customers ask whether Japan can decide when something goes wrong. "We'll check with headquarters" ends the ringi. I laid out which decisions belong in Japan and which at HQ in Product, sales, organization: why this order.

The failure patterns repeat. Promote the left side's best seller to country manager (I wrote about that one separately); the way they sell doesn't exist on the right. Set right-side quarterly targets at left-side closing speed; the team breaks up within six months. Attack large, mid-sized and small companies at once with one salesperson each; attention spreads thin and no tier gets crossed. Pick one tier for the first two years is the rule that matters most in SaaS.

How to cross

Moore's prescription is narrow on purpose. Don't spread. Pick one vertical, solve its most painful problem completely with a whole product, win one or two mainstream reference customers, and use their case to move into the next vertical. He calls it bowling: knock down the first pin.

Applied to Japan, my view is this. One tier, one vertical, for two years. Treat customers tied to your home country as revenue and never as evidence; they don't prove that a mainstream Japanese company will buy. Discount the first reference customer if you must, in exchange for a public case study with their name on it. And the first KPI for your Japanese sales hire should be a drawn map of one mainstream prospect's ringi path: who initiates, who reviews, who approves, and how many weeks each step takes. Without that map there is no road to the next hundred.

Two notes to myself. Rogers' objection stays with me; the Japanese chasm may be a slope rather than a line, and I can't say at which step you count as across. And one thing has nothing to do with the product. A Chinese vendor gets a few extra questions on the security sheet: which country the data sits in, which law governs, who the parent company is. Concern or pretext depends on the customer. Inside the ringi, they are questions you need an answer to before you walk in.

How the price band decides who buys and what sales team you need is in Your Price Decides What Kind of Sales Team You Need.