Review of the 6th GEO School Training (2026-06-20)

Review of the 6th GEO School Training (2026-06-20)

Hello, I’m marketing instructor Kim Bo-kyung.
In this second review, I’ll summarize the latter half of the 6th GEO School—the hands-on, practical part where you actually get to work and generate revenue.
If Part 1 was about “why AI search matters and where to focus”, this post is about “so how do you build it?”

In this Generative Engine Optimization (GEO) training, I shared exactly the methods I’ve applied with real clients—from content automation and keyword scoring to domain strategy.

AI content—how do you write it to get results?

There’s still a lot of resistance to AI content.
But most of my clients have achieved results with AI content.
That’s because Korea still doesn’t have enough content accumulated on Google, so well-made AI content is in a phase where it gets cited well.

Most people only think as far as “publishing content,” and then get stuck at rewriting.
If you look at the actual mix of content production methods: about 7% is written only by humans, about 7% is written only by AI,
and about 85% is a hybrid of humans and AI.

I also go with mass publishing first, then polishing through rewriting.

Here’s the core principle of rewriting.
Gather links to a few competitor pieces that rank at the top and train a tool like NotebookLM on them,
then ask, “Please identify the purchase intent of people searching this keyword.”
Then you’ll see why the existing top posts are ranking.

Rewriting works best when you do it by filling in what’s missing (the gaps) compared to your content.

How do you make competitor keywords your own?

In this review, the part that made practitioners’ eyes light up the most was keyword scoring.
Here’s how it works.

1) Enter a competitor’s website into Ubersuggest.
Paid tools (Ahrefs·Semrush) cost $100 a month, which can be a burden, but Ubersuggest can be used for free as well.

2) Then thousands of keywords that bring organic search traffic to that competitor will pour out.
Export them as-is.

3) Paste this keyword list into Claude or ChatGPT and ask it to “score them as A, B, C, or D based on purchase intent.”
– I used the example of my wife buying a bag.
– Keywords with product names like “Chanel bag” are searches from people ready to buy.
– These transactional keywords are Grade A. On the other hand, keywords that are just looking for information are Grade C or D.

So what were the actual results?

Health functional food OEM
We pulled about 2,600 keywords from a U.S. competitor, then auto-published 500 that we judged to have purchase conversion potential over three months. As a result, consultation inquiries increased from 10 per month to nearly 100.

Paper box shopping mall
After consulting, traffic jumped 20–30x.
However, I also shared honestly that Korean e-commerce tends to convert relatively weakly on Google.

The only difference is that companies with brand awareness can do well with AI content alone, while newer brands need to raise the overall quality of their content even more.

To do GEO well, you need to build a solid process.

Here’s a summary of the hands-on portion of the 6th GEO School.

– AI content performs when it’s a combination of mass publishing + rewriting.
– It starts with scoring competitor keywords into A, B, C, and D to set priorities.
– If you miss the basic infrastructure—domain age, sandbox, sitemap—even great content won’t get visibility.

The essence of GEO isn’t flashy tools,
but a refined process that starts with competitor keywords and flows into auto-publishing and rewriting.
With this one principle, I grew a $250 website into $12 million in annual revenue.

How should we build our company’s work processes? How do we internalize an automation system?
If you’re 고민이시라면 언제든 문의해 주세요.

I’ll make sure to repay you with high-quality consulting and training services.