Most small businesses do not have a technology stack. They have an archaeological site.

There is the tool they bought because a vendor recommended it. The spreadsheet that existed before the tool. The LINE group that everyone actually checks. The Chatwork room with the client history. The freee account nobody fully understands. The Google Drive with three competing folder structures. The paper forms that still matter because someone in the workflow refuses to let them die. The old laptop with the original files. The shared password nobody wants to discuss.

Then there is the person.

Every Japan SME has some version of this person. The one who knows which customer always pays late, which supplier needs a phone call before Golden Week, which form should be ignored, which folder contains the real template, which vendor login uses the founder’s old email address, and which process breaks if you follow the official instructions too literally.

The business says it has systems.

In practice, it has a stack made of software, habit, memory, etiquette, and duct tape.

The official stack and the real stack

The official stack is what management thinks the company uses.

It might be Google Workspace, Microsoft 365, Chatwork, LINE Works, freee, Money Forward, Kintone, Salesforce, Notion, Dropbox, an industry-specific booking system, a payment processor, a website CMS, and a few vendor portals.

The real stack is what people use when the official stack gets in the way.

That is where things get interesting.

Customer notes live in personal inboxes. Quotes get copied from last year’s Excel file. Passwords sit in a browser profile. A staff member sends photos through LINE because it is faster than uploading them properly. A manager tracks renewals in a paper notebook. A vendor sends invoices as PDFs to one person instead of the accounting address. The company has a CRM, but the sales team still trusts the spreadsheet more.

Nobody planned this system. It accumulated.

That does not make it stupid. It makes it dangerous.

Why the duct tape survived

Most messy stacks survive because they solved real problems at some point.

LINE survived because people answered quickly. Excel survived because it was flexible. Paper survived because it was trusted by the Japanese-side workflow. Chatwork survived because a client insisted on it. A shared drive survived because nobody had time to migrate the files. A manual approval survived because one expensive mistake made everyone afraid of removing it.

This is why simplistic modernization fails.

If you walk into a business and tell people to stop using the messy workaround, you may be killing the part of the process that actually carries trust. The question is not “Why are you still using this?” The better question is “What job is this workaround doing that the official system failed to do?”

Sometimes the answer is embarrassing. Sometimes the answer is practical. Sometimes the answer exposes a serious gap in the company.

A spreadsheet may exist because the CRM was badly configured. A private LINE thread may exist because the official communication channel is too slow. A paper checklist may exist because the software does not handle Japanese customer edge cases well. A staff member’s memory may exist as infrastructure because nobody ever funded documentation.

The workaround is evidence.

The cost is hidden until something changes

A duct-taped stack can look fine for years.

Then someone leaves. A card expires. A vendor changes pricing. A customer asks for a record. A manager wants to automate reporting. The company tries to hire internationally. A founder wants to sell. A new employee needs onboarding. The business wants to use AI. Suddenly the old arrangement stops being charming and starts being expensive.

The issue is not only wasted time. It is fragility.

If one person knows where everything is, that person is part of the infrastructure. If a customer relationship exists only inside private messages, the company does not fully own the relationship. If invoices depend on one inbox, accounting is exposed. If the CRM is not trusted, reporting is performance art. If the website does not match the current offer, every inquiry starts with confusion.

The business may still be profitable. It may still have good people and good customers. But it is carrying a quiet systems debt that limits what it can do next.

Japan adds its own complications

Japan SME stacks often contain layers that foreign operators underestimate.

There are tools chosen for Japanese accounting norms. Vendor portals that only work cleanly in Japanese. Hanko-era approval habits that survived inside digital forms. Clients who prefer LINE or phone calls. Staff who are comfortable with paper because paper never changed its interface. Local vendors who are excellent at their craft but weak at documentation. Bilingual teams where the English-side and Japanese-side records diverge over time.

None of this means Japan is uniquely broken.

It means the stack has to be understood locally.

A foreign-owned business in Japan can easily make the opposite mistake. It imports a global software expectation, then gets frustrated when the Japanese-side workflow does not comply. The Japanese team keeps a parallel process because the imported system does not handle the local reality. Management sees resistance. Staff see a tool that ignores how the work actually happens.

Both sides are partly right.

The fix is not to choose one culture’s software habits over the other. The fix is to build an operating layer that makes the real workflow visible and maintainable.

What a stack audit looks for

When I audit a stack, I am not only counting subscriptions.

I want to know which tools are official, which tools are trusted, which tools are ignored, and which tools are quietly doing critical work. I want to know who owns each account, who pays for it, who has admin access, what happens if that person leaves, and whether the tool still matches the job it was purchased to do.

I also want to know where information changes form.

Inquiry to quote. Quote to invoice. Japanese to English. Email to CRM. Paper to PDF. Verbal agreement to task. Customer request to internal handoff. Vendor promise to actual delivery.

Those translation points are where systems usually break.

A healthy stack is not the one with the most modern tools. It is the one where the business can explain how work moves, who owns each layer, where records live, and what happens when something changes.

The stack nobody admits can be fixed

The good news is that most messy SME stacks are not hopeless.

They usually need a few practical moves.

Cancel the tools nobody uses. Consolidate duplicate functions. Move passwords into a proper manager. Document the real workflow, not the imaginary one. Decide which communication channels are allowed for which kinds of work. Make the CRM reflect the actual sales process. Create bilingual handoff notes for the parts where context keeps getting lost. Review vendor contracts. Name the owner for each system. Build a simple renewal calendar. Turn the one person’s memory into company infrastructure before that person burns out or leaves.

None of this is glamorous. That is why it works.

The businesses that win are not always the ones with the flashiest software. They are the ones that stop pretending the duct tape is invisible.

If your company runs on LINE, Excel, paper, Chatwork, freee, shared drives, vendor portals, and one person who knows where everything is, you are not unusual. You are normal.

But normal is not the same as safe.

For owner-led and foreign-owned SMEs in Japan, a Stack Audit is the first step toward making the real stack visible. We map what exists, what it costs, who owns it, where the workflow breaks, and what should be simplified before the business buys another tool.

The goal is not to become a software company.

The goal is to stop running the company on accidental infrastructure.

The SME stack is not just SaaS. It is the counter, the kitchen, the supplier call, the handwritten note, the shared password, and the person who knows which workaround still works. This external image and video point back to that operating reality.

Man cooking inside a kitchen in Japan — External visual reference: small business operations are physical before they become digital. Photo by Anton Nazaretian on Unsplash.

Source video: Build+ on Japan’s DX lag. It is useful context for the gap between official modernization language and the systems many SMEs actually run every day.