Conclusion First: The Best Approach to AI Tools is “Don’t Implement”

156 new tools. This is the number of new LLM-related tools observed weekly on Hacker News. When calculated annually, it amounts to about 8,000 tools. Meanwhile, a survey reveals that 55% of companies that adopted AI tools regret their decision.

When these two figures are placed side by side, the underlying structure becomes clear.

The overwhelming increase in options has turned the act of “choosing” itself into a cost.

Large companies can afford to have dedicated AI teams that compare and validate dozens of tools, running PoCs over six months. However, small businesses with 10 to 50 employees do not have that luxury. By the time the CEO experiments with ChatGPT and thinks, “This might work,” another tool is already trending the following week.

The main point I want to convey in this article is simple: The greatest weapon for small businesses is not the “power to choose the latest tools,” but rather the “power to decline.”

—

Why Do 55% Regret Their Decision?—”Implementation” Has Become the Goal

When we break down the reasons for regret, they generally fall into three categories.

1. Inability to Utilize (Feature Overload)
They adopted an AI tool costing 50,000 yen per month but are using less than 10% of its features. The remaining 90% go untouched. In such cases, a ChatGPT Plus subscription at 2,000 yen per month would have sufficed—this is a scenario we have witnessed multiple times with our clients.

2. Misalignment with Business Needs (Lack of Clear Problems)
They only wanted the fact of “having adopted AI” without clarifying what business problems they aimed to solve. The tool is not at fault; the way the questions were framed is incorrect.

3. Rapid Obsolescence (Speed of Change)
A tool that was “cutting-edge” six months ago is now discontinued, or an open-source alternative with equivalent functionality has emerged for free. SaaS-based AI tools are particularly susceptible to this risk. The emptiness felt when told, “You can do the same thing for free” after spending 600,000 yen (10,000 yen per month for six months) can be fatal for small businesses.

What is common among these issues is that too much time and cost have been spent on the act of “choosing a tool,” leaving no room for the essential task of “transforming business operations.”

—

156 New Tools Weekly—What is Their Survival Rate?

Let’s think rationally. Out of the 156 AI tools that emerge weekly, how many are still actively used a year later?

According to data from Product Hunt, it is estimated that about 90% of launched AI products have virtually zero active users within 12 months. In other words, out of ten tools trending this week, only one or none is likely to survive until next year.

Knowing this “survival rate of less than 10%” can significantly change the decision-making process for adoption.

Instead of following the pattern of “A new tool emerges → try it immediately → adopt it,” consider this approach:
“A new tool emerges → wait three months → check if it’s still alive → then consider it.”

By simply waiting three months, you can automatically avoid 90% of the misses. This is the simplest implementation of the “power to decline.”

—

The “Power to Decline” for Small Businesses—Three Concrete Rules

Abstract discussions are meaningless. Here, I will share specific rules for adoption decisions that we actually propose to our small business clients.

Rule 1: “If it can’t be tested for under 5,000 yen a month, it’s too early.”

The most significant mistake in adopting AI tools is jumping into an annual contract for a tool that costs 500,000 yen upfront and 100,000 yen monthly.

In today’s world, most genuinely valuable AI applications can be realized within the range of ChatGPT Plus (approximately 3,000 yen per month), Claude Pro (approximately 3,000 yen per month), and Google Gemini (free to 2,900 yen per month). This includes image generation, text creation, data analysis, and meeting minutes preparation.

In fact, one of our clients, a local manufacturing company with 30 employees, reduced the labor hours for creating estimates from 40 hours a month to just 8 hours using only ChatGPT Plus. The cost was 3,000 yen per month. A dedicated estimation AI tool they previously considered cost 80,000 yen per month. The difference is 77,000 yen per month, amounting to about 920,000 yen annually.

Start by testing affordably. If it proves effective, continue using it. If not, discontinue. By following this order, most regrets will vanish.

Rule 2: “Ask first, ‘Can this be done with ChatGPT?'”

Many new AI tools are merely wrapper services that overlay a UI on top of ChatGPT or Claude. The underlying technology is often the same GPT-4o or Claude 3.5.

In other words, there are numerous cases where companies are paying for a specialized tool at 30,000 yen per month when they could achieve the same results with a general-purpose tool costing 3,000 yen per month.

The criteria for judgment are simple. When you discover a new tool, first try to replicate the same functionality using ChatGPT’s custom GPT or Claude’s project features. If you can replicate it, the tool is unnecessary. Only proceed to consideration if it offers unique value that cannot be replicated.

This filter alone can reduce the tools under consideration by about 80% based on experience.

Rule 3: “Do not adopt tools that require specific individuals to operate.”

AI tools that can only be used by “Person A” become a liability the moment they are implemented.

Small businesses often experience high employee turnover. Even if one person becomes proficient in using a tool, it becomes useless if that person leaves. This is the same structure as traditional Excel macros.

If you must choose, opt for tools that anyone can use with just a 30-minute explanation. Tools that are complex to operate, no matter how feature-rich, are unsuitable for small businesses.

—

What Lies Beyond “Declining”—A Structure Where Small Businesses Can Win

After reading this far, some may wonder, “Shouldn’t we just avoid using AI?” The opposite is true.

By declining, you can focus on what truly needs to be used.

Large companies require formal approvals for tool adoption due to their size, which can take six months for full deployment. In contrast, if the CEO of a small business decides, “Let’s use this,” everyone can start using it the next day. This speed of decision-making is something large corporations cannot replicate.

Therefore, “using a few tools deeply and quickly” becomes the winning strategy for small businesses.

Let me provide a concrete example. A local real estate company we support (with 8 employees) uses only ChatGPT Plus. However, they handle property descriptions, customer response emails, market analysis reports, and social media posts—all with this one tool. An investment of 24,000 yen (3,000 yen per month for 8 employees) resulted in a reduction of about 120 hours in administrative work per month. This equates to approximately 1.5 million yen in labor costs.

For 24,000 yen per month, they achieved the equivalent of 1.5 million yen in results. This is not a result of comparing 100 tools but rather the outcome of thoroughly utilizing one tool.

—

So, What Should We Do?

Here are three actions you can take starting tomorrow.

1. Take inventory of the AI tools you are currently using. How much are you paying monthly, and how many features are you actually using? Cancel any that are not being used by the end of this month.

2. Stop chasing information on new AI tools. Following 156 new tools weekly is the job of a large corporation’s research team. What small businesses should do is solve one immediate business problem with the tools they already have.

3. Implement the “three-month rule.” Even if you find a tool that interests you, wait three months. If it’s still trending and still exists after three months, then consider it at that time.

In an era where 156 AI tools are born every week, the strategy small businesses should adopt is not to “try everything” but rather to “ignore almost everything.”

Then, allocate the time and cost saved from not choosing to deeply utilize the one or two tools you do select. This is the most rational survival strategy for small businesses in an age of overwhelming choices.