← Back to 52-Week Low Lens

INTERFACTORY INC

Company Overview

Interfactory is a small Japanese software company that builds and runs enterprise e-commerce systems. Its core product, EBISUMART, is a cloud-based but heavily customizable commerce platform used by mid-sized to large Japanese merchants. The company has broadened into EC consulting and a product-information-management tool, but the economics still come overwhelmingly from the legacy platform business.

Growth has been real at the revenue line, but less impressive at the bottom line. Revenue rose from JPY 1.83 billion in FY2020 to JPY 2.86 billion in FY2025, a roughly 9.4% five-year CAGR. EPS, however, has not compounded cleanly through the cycle: FY2021 EPS was JPY 34.64 and FY2025 EPS was JPY 34.11, with a loss in FY2024 in between. The two concrete drivers have been higher recurring revenue per customer and new service revenue from EC business support. Monthly ARPU in system maintenance rose from JPY 261 thousand in FY2021 to JPY 384 thousand in FY2025, while GMV on the platform rose from JPY 127.7 billion to JPY 169.7 billion. Separately, EC business support revenue scaled from roughly JPY 0.10 billion in FY2024 to JPY 0.25 billion in FY2025.

Owner earnings sanity check. The cleanest reality check is cash flow, not accounting earnings, because software capitalization and amortization are material here.

That implies an owner earnings yield of roughly 10-14% on the current equity value. This is meaningfully better than the P/E suggests because FY2025 depreciation and amortization were JPY 202 million while total capex was JPY 175 million, and actual FY2025 free cash flow was about JPY 201 million. Put differently, the business generated more cash than net income in FY2025. I do not see evidence of material stock-based compensation expense in the reported numbers; disclosed option dilution was small at roughly 0.5% of shares.

Capital efficiency is acceptable but not elite. FY2025 ROE was about 12% and reported ROIC was about 9%. Through-cycle returns are volatile because FY2024 was loss-making. Incremental capital in the core platform appears to earn decent returns, shown by rising ARPU and higher segment profit. Incremental capital in the new segments has not yet proven itself.

Business quality is better than the group-level numbers first suggest. In FY2025, the cloud commerce platform segment produced JPY 868 million of segment profit on JPY 2.62 billion of segment sales. EC business growth support lost JPY 13 million and the data platform lost JPY 29 million; after roughly JPY 630 million of unallocated corporate costs, consolidated operating profit was JPY 196 million. In plain English, profits come from the installed base of customized enterprise EC systems and the monthly maintenance around them. The business has historically been decent because once a merchant has built deep integrations into payments, logistics, inventory, CRM and custom workflows, switching is disruptive and risky. That creates real, though not impregnable, switching costs.

How the Company Makes Money

Interfactory makes money in two layers. First, it charges upfront project revenue to build or customize commerce systems. Second, and more importantly, it charges monthly recurring fees to operate and maintain those systems. The monthly fee mix includes base maintenance, custom maintenance, options, and usage-based charges. There is also a small revenue-share model, but most recurring revenue is still fixed or quasi-fixed.

At the segment level, FY2025 sales were JPY 2.62 billion in cloud commerce, JPY 0.25 billion in EC business support, and effectively zero in the new data platform. The revenue engine matters because the development work is not just one-off income; it is also the feeder for future maintenance revenue. That is why the company tracks new order intake and maintenance ARPU so closely. FY2025 new order intake rose to JPY 1.276 billion from JPY 0.999 billion in FY2024.

The economic center of gravity is still the core platform. The company has over 800 cumulative site implementations, more than 1,000 published APIs, seven consecutive years of claimed share leadership in customizable cloud EC, and a customer base that includes well-known Japanese and international brands operating in Japan. Those facts do not create network effects, but they do create trust, references, and replacement friction.

Why the Stock Fell

The stock is near its 52-week low because the market now treats FY2025 as a peak earnings year, not as a new base. The shares traded as high as JPY 610 in April 2025 after the company raised FY2025 profit guidance. They later fell to around JPY 388-396 by March-April 2026, roughly 35-36% below that high and effectively at the year low.

The causal chain is straightforward. FY2025 looked strong: sales grew 10.4%, operating profit swung from a JPY 25 million loss to a JPY 196 million profit, and free cash flow turned strongly positive. Then, in July 2025, management guided FY2026 to higher revenue but sharply lower profit: sales up 12.5% to JPY 3.223 billion, but operating profit down 58% to JPY 82 million and net income down 61% to JPY 53 million. In January 2026, the first half broadly confirmed the pattern: sales rose only 2.4%, while operating profit fell 56.6% and net income fell 65.2%.

Investors appear worried about four things: first, that FY2025 benefited from temporary cost relief and delayed cancellations; second, that new segments are consuming capital and SG&A without proving they can earn returns; third, that customer count decline signals a maturing or narrowing core franchise; and fourth, that a small company with one real profit engine deserves a heavy discount if management chooses growth spending over margin.

What the Market Is Assuming

• (a) One-time / cyclical / sentiment-driven. FY2025 margins were temporarily inflated by lower server and contractor costs.
• (a) One-time / cyclical / sentiment-driven. The absence of expected customer cancellations in FY2025 pulled profit forward.
• (a) One-time / cyclical / sentiment-driven. FY2026 is the first proof that FY2025 was a peak, so the market is de-rating the stock accordingly.
• (b) Medium-term headwind. Customer-store count is falling, so recurring revenue breadth may be weakening even if ARPU rises.
• (b) Medium-term headwind. New businesses are diluting consolidated margins and may keep doing so for years.
• (b) Medium-term headwind. Revenue growth guidance leans heavily on project revenue and newer products, which have lower visibility than the installed maintenance base.
• (c) Potential long-term structural threat. The company still depends on one domestic enterprise EC platform for nearly all profits.
• (c) Potential long-term structural threat. Global platforms, domestic SIers, and AI-enabled development tools could compress replacement win rates over time.
• (c) Potential long-term structural threat. Higher cloud, security, and product-development requirements could keep the business more capital-intensive than a typical SaaS multiple implies.

Reality check vs market narrative. The company does not disclose ARR or churn, so the best retention and quality proxies are customer-store count, ARPU, GMV, segment profit, and cash flow.

Temporary or Structural?

Diagnosis: the current problem is mostly TIME, not ESSENCE, at the level of the core business. The core value-creation mechanism, winning complex Japanese enterprise EC projects and then monetizing them through sticky maintenance, does not yet look broken. The structural issue is narrower: the moat is real but moderate, and it is concentrated in one product family.

Time-as-a-moat test. Assume you had the current market capitalization, roughly JPY 1.6 billion, in cash and wanted to build a competitor.

This is the profile of a moderate moat: stronger than a commodity software vendor, weaker than a true platform monopoly.

Is the Market Wrong? By How Much?

The market is somewhat wrong, but not dramatically wrong. It is pricing the stock as if FY2026's depressed earnings are close to a new normal. I think that is too harsh on the core business, but not obviously harsh enough on the group if management keeps reinvesting without earning returns.

My base case is that normalized owner earnings are closer to JPY 160-180 million than to the JPY 53 million FY2026 net income guide. On a 10% required owner earnings yield for a small, concentrated Japanese software company, that supports an equity value of roughly JPY 1.6-1.8 billion. Against a current market cap of about JPY 1.6 billion, that is a gap of roughly JPY 0.1-0.2 billion in the base case. If one anchors instead to FY2025 free cash flow around JPY 201 million, the gap could be closer to JPY 0.4 billion. If FY2026 turns out to be the true earnings base, there is no mispricing at all.

Moat & Mispricing Score: 6/10. The market is getting one important thing wrong: it is treating group-level margin compression as if it proves core-franchise impairment. The data do not show that. But the market is getting two things right: the moat is only moderate, and management has not yet shown that the newer businesses can absorb capital without dragging returns. This is a modest TIME mispricing, not a deep ESSENCE misunderstanding.

Key Facts, Estimates, and Judgments

CoffeeAnd — 52-week low lens