The Day Copilot Became the ‘New OS for Work’—Do Small and Medium Enterprises Understand the Implications of Office Being Downgraded to ‘Apps That Operate Within’?

Conclusion Let’s get straight to the point: "Being able to use Office" is no longer a competitive advantage. In a nutsh

By Kai

|

Related Articles

Conclusion

Let’s get straight to the point: “Being able to use Office” is no longer a competitive advantage.

In a nutshell, what Microsoft declared at Build in May can be summarized as follows:

“Copilot is the OS for work. Word, Excel, and PowerPoint are merely applications that run on top of it.”

Do you understand what this means?

Just as Excel used to run on Windows, now Excel will run on Copilot. The control shifts from the application to AI. Users no longer need to “open Excel and create a pivot table”; they simply tell Copilot, “Compare last month’s sales by region.” Copilot will automatically invoke Excel, format the data, and even create graphs.

In other words, the market value of someone who can “use Excel” has begun to structurally decline.

For small and medium enterprises, this is not a threat. It’s an opportunity.

—

What Happened—Reading Three News Items in Context

There are three notable developments to focus on. They may seem unrelated, but they share a common root.

1. Microsoft Copilot: The ‘OS for Work’ Declaration

CEO Satya Nadella announced a “frontier ecosystem centered around Copilot.” In short, this represents a shift where Copilot becomes the command center, and the Office suite operates as its limbs.

What specifically will change?

  • Until now: Humans select applications, learn how to operate them, and work within those applications.
  • From now on: Humans communicate their intentions to Copilot, which selects the applications and executes the tasks.

Writing meeting minutes in Word, summarizing numbers in Excel, and creating slides in PowerPoint—Copilot will handle this entire workflow cross-functionally. The boundaries between applications will dissolve.

The license for Microsoft 365 Copilot is $30 per user per month (approximately 4,500 yen). For a company of five people, that amounts to 22,500 yen per month, or 270,000 yen annually. You might think this is “expensive.” However, if this 270,000 yen investment automates 30-40% of administrative tasks, what does that mean? If the annual labor cost for one administrative staff member is 4 million yen, then 30% of that is 1.2 million yen. Investing 270,000 yen would free up time worth 1.2 million yen.

Of course, the “30-40% automation” depends on the environment. However, for small and medium enterprises where routine document creation, data aggregation, and report generation make up the majority of work, this figure is within realistic reach.

2. Google Gemini Tasks: AI Agents Arriving for Free

Google has added a “Tasks” feature to Gemini. The noteworthy aspect is that it’s free.

Gemini Tasks can be thought of as a “self-operating secretary.” Here are some examples of what it can do:

  • Adjust meetings by checking available times on the calendar
  • Classify incoming emails, prioritize them, and summarize their content
  • Regularly compare prices online and notify users
  • Check inventory data and alert when to reorder

And this is free. Companies using Google Workspace can start using it at no additional cost.

Consider this in the context of a small business. In a company of five employees, if each spends 1.5 hours a day on email correspondence, schedule adjustments, and estimate comparisons, that totals 5 x 1.5 hours x 250 days (annual working days) = 1,875 hours. At an hourly wage of 1,500 yen, that amounts to approximately 2.8 million yen worth of time being consumed.

If Gemini Tasks can automate half of that, it would free up 1.4 million yen worth of time. The cost is zero.

How many small business owners recognize this structure of “creating 1.4 million yen worth of time at zero cost”? That is the question.

3. The Surge of Decision-Making Models—AI is Starting to Replace ‘Judgment’

The third point is somewhat technical, but the impact is significant. There has been a rapid increase in lightweight AI models specialized for decision-making, provided by companies like Sqwish Labs.

Until now, AI has primarily focused on generation tasks like “writing text” or “creating images.” However, decision-making models are different. They determine “whether to choose A or B” based on data.

Here’s how they can be used:

  • Determine which products to prioritize for procurement next month based on sales data from the past three years
  • Identify customers at high risk of churn based on their inquiries
  • Prioritize candidates for hiring based on their information

Decisions that previously relied on the “CEO’s intuition” or the “experience of veteran salespeople” can now be reproduced using data and models. The highly personalized nature of “decision-making” can be systematized.

Moreover, the cost of implementing such models is rapidly decreasing. What used to cost hundreds of thousands of yen to set up with a specialized data scientist can now be accessed for a few tens of thousands of yen per month using API-based solutions.

—

What is Changing Structurally—The ‘Seeds of Reversal’ for Small and Medium Enterprises

When you overlay these three movements, a structural change becomes visible.

The costs associated with “hiring, training, and retaining people” are dramatically decreasing in certain areas.

Large corporations have historically increased efficiency by hiring large numbers of people, dividing labor, and creating manuals. If AI can take over tasks previously handled by a team of 100, then those 100 people are no longer needed. However, large corporations cannot easily reduce their workforce by 100 due to organizational inertia, labor risks, and lengthy decision-making processes.

On the other hand, what about small and medium enterprises? A company of five can now handle the workload of ten using AI. There’s no need to hire an additional five people. Zero hiring costs, zero training costs, and zero turnover risks.

While large corporations struggle to reduce their workforce from “100 to 95,” small and medium enterprises have the potential to achieve “5 to 5 while doubling revenue.”

This is the structure of reversal.

—

So, What Should We Do?

Let’s move on from abstract discussions. Here are three actionable steps you can take starting tomorrow.

1. Try Google Gemini Tasks today (Cost: 0 yen)

If you use Google Workspace, open Gemini and experiment with Tasks. Start with something simple like “Summarize incoming emails and prioritize them every morning.” Even if it doesn’t work perfectly, there’s no cost.

2. Calculate the cost-effectiveness of Copilot using your own numbers (Cost: 30 minutes)

Calculate the annual cost by multiplying 4,500 yen by the number of users. Next, identify the time spent on administrative tasks in your company: document creation, data aggregation, report generation, email correspondence. Estimate what percentage of that could be replaced by AI. If it pays off in terms of labor costs, try it for just one month.

3. Write down one decision you currently make based on “CEO’s intuition” (Cost: 10 minutes)

This could be about procurement volumes, pricing strategies, hiring decisions, or customer prioritization. Just one is enough. Articulate “why you are making that decision.” If you can verbalize it, there’s a high likelihood that it can be entrusted to AI if data is available.

—

Is This Enough?

In the era when Office was the “OS for work,” those who memorized Excel functions were highly valued. In the era when Copilot becomes the “OS for work,” it will be those who can design what tasks to assign to AI that will be valued.

This change can be adapted to more quickly by small and medium enterprises than by large corporations, as the distance between decision-makers and the field is shorter. If the CEO says, “We’re implementing Copilot starting next week,” things can change by next week. In a large corporation, it might take three months for approval.

The question is whether they recognize this structural change.

Microsoft has effectively declared that “Office has been downgraded to applications.” Google is saying, “We’re giving away AI agents for free.” Decision-making models can be used for just a few tens of thousands of yen per month.

The tools are in place. It’s now up to you to use them.

While you say, “It’s still too early for us,” a competitor in the neighboring town may start handling the workload of ten with just five employees. Only then will you realize, “I was too late.”

Start with just one thing to try today. That’s all it takes.

POPULAR ARTICLES

Related Articles

POPULAR ARTICLES

JP JA US EN