The 2025 Digital Cliff Explained｜What the Problem Is and What to Do

The "2025 Digital Cliff" is a term that comes up in almost every conversation about DX in Japan. Many people know the phrase but have never quite pinned down what the actual problem is — or whether it still matters now that 2025 has come and gone. This article explains what the 2025 Digital Cliff is in plain terms, what caused it, what happens to companies that ignore it, where things stand as of 2026, and what to do from here.

What Is the 2025 Digital Cliff?

The 2025 Digital Cliff is a warning laid out by Japan's Ministry of Economy, Trade and Industry (METI) in its September 2018 "DX Report: Overcoming the 2025 Digital Cliff Involving IT Systems and Full-Fledged Development of Efforts for Digital Transformation." It describes a scenario in which companies that leave aging, over-complicated, black-boxed legacy systems in place could collectively cost Japan up to 12 trillion yen per year in economic losses from 2025 onward.

That 12 trillion yen figure is roughly three times the economic loss estimated at the time the report was published (around 4 trillion yen). The image of losses jumping sharply, like a step in the road, is what earned it the name "cliff."

In One Sentence

"Hold on to old systems, and you'll be dragged down by maintenance costs and talent shortages until DX stalls and you lose your competitive edge." That is the essence of the 2025 Digital Cliff. Crucially, it was framed not as an IT department problem but as a management problem.

Why 2025?｜Three Converging Factors

The year 2025 was singled out because several forces were expected to converge around that point.

1. Aging core systems: By 2025, roughly 60% of Japanese companies were projected to still be running core systems that had been in operation for 21 years or more. Years of layered modifications leave no one able to see the whole picture, and the system becomes a black box.
2. IT talent shortage and retirement: METI research projected a shortfall of up to roughly 430,000 IT professionals by 2025 and up to 790,000 by 2030. The retirement of engineers who can work with older technologies such as COBOL hits legacy system maintenance directly.
3. End of support for major systems: Support deadlines for major products, ERP in particular, were expected to cluster around 2025. The standard maintenance deadline for SAP ERP — the best-known example — was later extended, and the issue is now commonly discussed as the "2027 problem."

What Actually Goes Wrong for Companies That Do Nothing

The DX Report set out the consequences of leaving legacy systems in place as follows.

• Technical debt eats the IT budget: Maintenance costs climb until more than 90% of the IT budget goes to simply keeping existing systems running. Nothing is left to fund new initiatives.
• Data can't be used, so DX stalls: Systems siloed by department make company-wide data use impossible, and business models can't be adapted quickly to market shifts.
• Outage and security risk grows: As the people who maintain these systems disappear, so does the ability to respond to failures, data loss, and cyberattacks.
• Competitiveness erodes: Digital-native entrants move faster, and incumbents fall behind on price and on the value they can deliver. This is the core of the projected economic loss.

Now That 2025 Has Passed, What Happened to the Cliff?

The short answer: the cliff is not a closed chapter. Core system replacement and cloud migration have progressed to a degree, and the legacy migration market has expanded. Yet surveys report that as of 2026, around 60% of companies still have legacy systems in place, and rising maintenance costs and engineer shortages are, if anything, playing out as concrete day-to-day problems.

There is also a newer problem, one that affects the companies that did avoid the cliff. For many of them, replacing the system became the goal in itself, and the DX Report's actual message — redirect the freed-up resources into transforming the business with digital technology — got postponed. Modernization is the starting line, not the finish.

What to Do｜Five Steps to Start From Here

1. Make the current state visible: Start with an inventory of your systems. What runs where, who maintains it, and what does it cost? METI's "DX Promotion Index" gives you a way to self-assess your position objectively.
2. Sort, don't replace everything: Not every system needs modernizing. Sort each one into "retire," "leave as is," or "replace" based on both business value and risk, and narrow down where you invest.
3. Have leadership define the purpose: Not "make the system new" but "what are we changing this in order to achieve?" If leadership leaves this vague, DX will stall right after the migration is done.
4. Fit the business to standard functionality: Excessive customization simply breeds the next legacy system. A "Fit to Standard" approach — adapting processes to the standard features of cloud ERP and similar tools — shortens implementation and lightens future maintenance at the same time.
5. Build in talent and measurement: Alongside developing and hiring the people who will drive this, make sure the results are trackable with KPIs. Keep measuring the share of budget going to maintenance, and how decision-making speed changes as data becomes usable.

Frequently Asked Questions

How is this different from the 2027 problem?

The 2025 Digital Cliff refers to the broad problem of Japanese companies falling behind on DX. The 2027 problem refers to one specific support deadline: the end of standard maintenance for SAP ERP. The easiest way to hold both in your head is to treat the 2027 problem as one component of the cliff, made concrete.

Does this apply to small and mid-sized companies?

Yes. Even companies without a large ERP face the same underlying problem when data is scattered across spreadsheets and standalone tools that don't talk to each other. If modification costs spike every time a law changes, or if nobody can touch a system once one person leaves, you are already living on the cliff.

Is the 12 trillion yen figure credible?

It is an estimate built on assumptions, and some experts have criticized its precision. What matters is not the number itself but the fact that three structural problems are real: maintenance costs are rising, the talent pool is drying up, and data cannot be used.

Summary

The 2025 Digital Cliff is the scenario METI warned about in its 2018 DX Report: leaving legacy systems in place could cost up to 12 trillion yen a year from 2025 onward. Behind it sat three factors — aging core systems, a shortage of IT professionals, and a cluster of support deadlines. In 2026, legacy systems remain at many companies, and the problem is very much ongoing. Start by taking inventory of your own systems and sorting them into what to retire, what to keep, and what to replace. And remember that clearing the cliff for real means redirecting resources not into the migration itself, but into the business transformation that comes after it.