The conclusion first. The biggest cause of failure in modernizing legacy systems and analog operations is not tech selection but "trying to do a full migration all at once." Hand a field that has run on phone, fax, and Excel a brand-new system one day, and the field gets confused, says "the old way was faster," and the expensive system ends up unused. The right answer is to start from information sharing and visualization, and after the field feels the convenience, migrate operations one by one in stages. This article is a practical guide for buyers to succeed at legacy modernization, against the backdrop of the "2025 cliff."
1. The "2025 cliff" has passed but isn't over
The "2025 cliff" that METI warned of in its "DX Report" — the problem that, unless legacy systems that have aged, grown complex, and become black boxes are modernized, an economic loss of up to 12 trillion yen per year could occur from 2025 onward.
What matters is the fact that even now, past 2025, this problem isn't solved. Only a fraction of companies could complete legacy modernization as planned, and many companies are still in a state where:
- Maintainers have retired or left, and they keep using a system no one understands the inside of
- Analog operations by phone, fax, and Excel are person-dependent, with no records left
- Data is scattered in various places, not in a form usable for management decisions
The longer it's left, the more the cost and risk of modernization keep rising. The "cliff" should be thought of not as a single deadline but as a valley that deepens every year.
What buyers should first recognize: legacy modernization is not a "someday" issue but a "the more you postpone, the more expensive it gets" issue. Before the people who know the inside are gone, it's worth at least starting from an inventory of operations and data.
2. Why DX in legacy industries is hard
Manufacturing, construction, logistics, agriculture/forestry/fisheries — DX in these "legacy industries" is hard not because of technology. It's because it's a problem deeply rooted in the field's people and operations.
| The difficulty | Content |
|---|---|
| The wall of IT literacy | A field where "I can use Excel, but a web system is scary." If they can't use it, it's meaningless |
| Person-dependence of operations | Procedures aren't manualized and exist only in the veteran's head |
| An analog-premised commercial flow | Ordering by phone/fax is the industry's common language. The counterpart does the same |
| "It's running with the current way" | Inconvenient but running, so the motivation to change is weak |
Over the past two years, I worked on DX in the lumber-distribution industry, a "super-legacy industry," and developed a product that won the METI Minister's Award. Phone, fax, and Excel were mainstream; inventory was in Excel, ordering was by fax with no record left, and confirmation work took several hours every day — that was the field. What I learned there is that building "the correct system" and building "a system the field keeps using" are entirely separate problems.
3. The core of success: staged introduction (don't do a full migration all at once)
The biggest principle for succeeding at legacy modernization is staged introduction. While coexisting with existing operations, migrate while keeping risk low.
Phase 1（1〜3ヶ月）: 情報共有・可視化のみ
・既存のExcelをそのままアップロード→Web上で見られるようにする
・現場は今までのExcelを使い続けてOK
・「散らばっていた情報が一覧で見える」便利さを実感してもらう
↓
Phase 2（4〜6ヶ月）: 一部業務の移行
・発注など、効果が見えやすい業務から1つずつWebへ
・Excelとの併用を続け、戻れる安心感を残す
↓
Phase 3（7〜12ヶ月）: 全面移行
・現場が新しいやり方に慣れた業務から、旧フローを段階的に廃止
・産地証明・チャット・評価など、デジタルならではの価値を追加
This order is decisively important. Starting from Phase 3 all at once fails. In the lumber DX too, I first built a path that could import existing Excel as-is and turn it into a DB, so the field could migrate without discarding the tools they were used to. Build up the feeling of "this is convenient" before moving to the next — this divides success from failure in legacy industries.
4. Design so the field keeps using it
Alongside staged introduction, design usable even by a field unfamiliar with IT is important. Without weaving this in from the start, you'll lose them in Phase 1.
- Excel-like operability — minimize the learning cost with an operation feel close to the Excel the field is used to (table-format editing, range selection).
- Understandable error messages — make it Japanese that shows what to do, like "Enter the price as 0 yen or more (current input: -1000 yen)," not "ValidationError: field price must be positive."
- Reproduce existing forms — let them output quotes, delivery notes, and invoices with one click, in the layout the field is used to seeing. In the lumber DX, I made it possible to auto-generate these Excel/PDFs with one click.
Design philosophy: DX in legacy industries is not "replacing with the latest technology" but "gently adding digital convenience to the field's tools and operations." Design that respects the field's habits is what decides the adoption rate.
5. The cost of modernization: factor in the invisible costs
The cost of legacy modernization is also basically decided by person-months (person-month unit price × number of people × period). But unlike new development, modernization has three "invisible costs." Without factoring them in, the estimate is naive.
| Invisible cost | Content |
|---|---|
| Data migration | Format and move existing Excel / old-system data into the DB. The dirtier the data, the more expensive |
| Parallel operation | The load of double operation, running both old and new systems during the migration period |
| Training/adoption support | Field training, manual preparation, inquiry handling |
Budgeting these separately from the "development cost" is the realistic budget. Conversely, a modernization estimate that doesn't touch these at all is a danger sign.
Using public support (subsidies)
For SMB DX, public support like the IT-introduction subsidy can sometimes be used (the system and requirements change yearly, so confirming the latest application guidelines and a certified IT-introduction support provider is essential). Subsidies lower the hurdle of the modernization's initial investment, but "build it because we can get a subsidy" is backwards. Keep the order of backing up "a modernization worth building" with a subsidy.
6. You don't need to "rebuild everything" — modernize with a hybrid
Legacy modernization tends to make people brace for "rebuilding everything from zero," but that's often the most expensive choice. The realistic answer, as stated in the SaaS vs scratch article, is a hybrid.
- Use off-the-shelf foundations — auth, payments, email, storage, monitoring with off-the-shelf products (Cognito, Stripe, etc.).
- Build yourself only the industry-specific logic — the complex commercial flows and unique operations that become your business's core of differentiation.
The lumber DX too used off-the-shelf foundations of Cognito for auth and Stripe Connect for payments, and built itself only the industry-specific core of "industry-role-based multi-stage commercial flow and permissions." Modernization is not "full replacement" but "build only the core, and solidify the rest with off-the-shelf products" — this mindset greatly lowers cost and risk.
FAQ
Q. The "2025 cliff" has passed — do I still need countermeasures?
You do. Only a fraction of companies could complete legacy modernization as planned, and many still hold a depletion of maintainers, person-dependence, and scattered data. Since the longer it's left, the more modernization cost and risk rise, the "cliff" should be seen not as a single deadline but as a valley that deepens every year. It's worth at least starting from an inventory of operations and data.
Q. Can phone/fax/Excel operations really be systematized?
They can. The key is "don't do a full migration all at once." First import existing Excel and visualize it, and after the field feels the convenience, migrate operations like ordering one by one. Weaving Excel-like operability and understandable error messages into the design from the start, so even a field unfamiliar with IT can use it, is the condition for success.
Q. How much does legacy modernization cost?
It depends on scale, but modernizing a business system is from several million yen as a guide. Unlike new development, you need to factor in the "invisible costs" of data migration, parallel operation, and training. An estimate that doesn't touch these is a danger sign. SMBs can also use public support like the IT-introduction subsidy, but since the system/requirements change yearly, confirming the latest info is essential.
Q. Do I need to rebuild everything at once?
No. Full replacement tends to be the most expensive choice. The realistic answer is a hybrid that uses off-the-shelf foundations for auth/payments and builds yourself only the industry-specific core of differentiation. This greatly lowers cost and risk.
Q. I'm anxious about whether the field will use the new system.
That anxiety is correct and the most important point. That's exactly why staged introduction and "design so the field keeps using it" are the core. Start from information sharing so they feel the convenience, and migrate while coexisting with the familiar tool (Excel). Prepare a UI that doesn't confuse even people unfamiliar with IT, and Japanese error messages. Design that prioritizes adoption over technology is the trick to not wasting the investment.
Summary: modernization is decided by "the approach," not "the technology"
To succeed at legacy-system modernization, here's what buyers should grasp.
- The "2025 cliff" isn't over — the more you postpone, the more expensive. Start from an inventory.
- The true cause of failure isn't tech selection but "a full migration all at once" — staged introduction (information sharing → partial → full) divides success from failure.
- Design so the field keeps using it is the core — Excel-like operability, reproducing existing forms, understandable errors.
- Factor in person-months + invisible costs (data migration, parallel operation, training) — use subsidies as a backer.
- Don't rebuild everything — a hybrid that uses off-the-shelf foundations and builds only the core of differentiation from scratch.
"I don't know where to start" — legacy modernization usually starts there. I take on DX in any legacy industry — manufacturing, construction, logistics, agriculture/forestry/fisheries — one-stop, close to the field, from an inventory of operations through staged migration design, implementation, and operations.