More than Half of the Web is No Longer Written by Humans

First, let’s look at the numbers.

According to a 2024 study (by Amazon and Rice University), approximately 57% of low-quality translated text on the web is estimated to be AI-generated. Another survey indicates that the proportion of AI-generated content is even higher when limited to .com domains in English-speaking regions. What was nearly zero five years ago has now surpassed half.

What does this mean?

It means that the “competition” in search results has exploded. Moreover, this competition operates with zero labor costs, works 24/7, and can produce hundreds of pages in a single day. While a small business takes half a day to create one page, AI can churn out 100 pages on the same topic in that time.

The rules of the game for “being found in search” have fundamentally changed.

So, can small businesses no longer win? In conclusion, the answer is that “the way to win has changed.” In fact, there are positions that only small businesses can take.

—

Where is AI-Generated Content Increasing?

The important point is that the increase in AI-generated content is not uniform.

The aforementioned study reveals that there is a significant difference in AI generation rates based on the web’s “connection structure”—that is, how many sites link to a given page.

• Peripheral pages (isolated pages with few links): The AI generation rate is the highest and is rapidly increasing.
• Central pages (reliable pages with many links): The AI generation rate is low, and the increase is gradual.
• .gov and .edu domains: The proportion of AI-generated text is quite low.

In other words, AI is primarily mass-producing “isolated, low-quality pages that are not linked by anyone.” Conversely, “trusted pages” that are naturally linked from local communities, business partners, and industry organizations have not yet been encroached upon by AI.

This is where the opportunity for small businesses lies.

—

The Era Where 3 Million Yen in SEO Efforts is Now “Finished” for 50,000 Yen

Let’s shift our perspective a bit.

In the past, if small businesses wanted to implement SEO, the going rate for outsourcing was around 300,000 to 500,000 yen per month, totaling 3 to 6 million yen annually. Keyword research, competitive analysis, article creation, and internal linking design—all were done manually by experts.

Now, using ChatGPT or Claude, keyword research can be completed in minutes. Competitive analysis, article drafts, and structural proposals can be done at nearly zero cost. With AI tools costing 20,000 to 30,000 yen per month and a few hours of work from an in-house staff member each week, most of the tasks that used to cost 3 million yen annually can now be covered.

Costs have dropped to one-hundredth of what they were.

This is both a threat and a weapon for small businesses. While large companies are busy getting approvals for outsourcing, a small business owner can say, “Get this done this afternoon.” The speed of decision-making is the greatest weapon of small businesses.

However, there is one problem here.

—

“Pages Written by AI” Are No Longer a Differentiator

The ability to mass-produce articles using AI is now accessible to anyone. This means that pages created solely by AI have no value. If 100 companies generate articles on the same topic with the same prompt, there will simply be 100 similar articles on the web.

Google is also taking action. In the core update of March 2024, it explicitly included “mass-generated low-quality content” in its spam policy. It has repeatedly stated that the criterion is not whether content is AI-generated, but whether it is “valuable to users.”

So, what constitutes “valuable content”?

It is something that AI cannot write—”primary information that only your company possesses.”

—

What is the “Primary Information” That Small Businesses Can Win With?

Let’s think concretely.

Take a local sheet metal processing company as an example. AI can endlessly generate explanatory articles on “what sheet metal processing is.” However, the following information cannot be written by AI:

• **A field failure story about “the cause and countermeasures when bending a 1.5mm SUS304 plate with an R3 bend.”
• “We can deliver this processing in 3 days because we have this equipment.” Specific delivery times and justifications.
• “10 minutes by car from the △△ industrial park in ○○ city. Prototype meetings will be held in person.” Geographic information.
• Photos of actual processing cases and testimonials from clients.

All of these are “primary information.” They can only be generated in your company’s environment. No matter how advanced AI becomes, only you can write about your factory’s processing achievements.

And this type of information is precisely what Google is trying to evaluate as “E-E-A-T (Experience, Expertise, Authority, Trustworthiness).”

—

How Small Businesses Can Use “Intent-Driven SEO”

Recently, the concept of “G-SEO (Generative Search Engine Optimization)” has been proposed in academic research. In an era where AI summarizes and answers search results, such as Google’s SGE (Search Generative Experience) and Bing’s Copilot, it is optimization to get your company’s information picked up by AI’s answers.

Among these, a method called “RAID” is gaining attention, which involves breaking down user search intent and structuring content accordingly.

However, to be honest, small businesses do not need to implement this method exactly as described in the papers.

The essence is simple. Imagine specifically “what customers want to know when they search” and write the answers on your page. That’s all.

For example, for a local construction company:

• × “We provide high-quality homes” (AI can write this. Anyone can say it.)
• ○ “The utility cost results for a house built in ○○ city with insulation grade 6. The average electricity bill for a customer after one year was 12,000 yen per month.” (Primary information. Specific. Directly answering search intent.)

Pages like the latter are more likely to be cited as “trustworthy sources” when AI generates summary answers. This means they will be advantageous in both traditional SEO and AI search.

—

So, What Should We Do?

Let’s narrow down what small businesses should do to three points.

1. Use AI as a “Drafting Machine”

Have AI create article outlines and drafts. However, do not publish them as they are. Always add your company’s primary information—numbers, case studies, photos, customer testimonials. Add 20 points that only your company has to the 80-point draft created by AI. This 20 points will determine life or death in search.

2. Win Small with “Location Name × Specialized Terms”

There is no need to aim for first place nationally with “sheet metal processing.” It is enough to take first place with “○○ city sheet metal processing SUS short delivery time.” AI-generated content focuses on “broad and shallow” keywords. Niche keywords that combine location and specialization still have little competition. Securing these is the standard strategy for small businesses.

3. Create One “Linkable Page”

As the earlier study shows, AI-generated content is increasing on “isolated peripheral pages.” Conversely, if you have even one “trusted page” that is linked from local chambers of commerce, industry media, or partner sites, you can stand on a different layer from the mass-produced AI pages.

Specifically, create detailed pages explaining your company’s technology or local industry reports and get them introduced by industry organizations or local media. This is easier for small businesses than for large corporations. If the president mentions at a chamber of commerce meeting, “We created this page,” that can lead to a link.

—

In an Era Where the Web is Filled with AI, “Human Traces” Become the Greatest Weapon

In a world where more than half of the web is AI-generated, there are things that conversely gain value.

It is information about “what this person has actually done in this place.”

AI generates patterns. However, what happens in your company’s environment lies outside those patterns. That “outside the pattern” is the information that search engines—and customers—truly seek.

What happens when costs decrease? As the cost of content generation approaches zero, the value of “things that cannot be generated” increases. That includes primary information, field experiences, and trust relationships rooted in the community.

Small businesses inherently possess these.

The only question is whether they are putting it on the web.