SYSTEM NOTICE

Auto translation by AI. Be sure, accuracy, nuances and authorial intent may not be fully reflected.
見出し画像

[Claude 4 Experience Report] When I held a 'Tool Development Meeting' with 3 people (+α) including a Google employee, the strategic level was on another dimension

*The statements made by the Google employee (fictional) appearing in this article are personal opinions and do not represent the official views of Google.

It goes without saying

Introduction: What is Claude 4?

Has everyone used Claude 4 yet?

It is the latest AI model released just on May 22, 2025, and it allows for usage like "calling multiple experts at the same time".

This time, I had it read all my past articles and asked it to "come up with ideas for a tool that will sell" and...

The meeting started like this...

At first, I was only talking to an engineer, but honestly, the accuracy of the proposals was at a housewife level... Like this proposal.

If this is it, the existing "Writing Consultation Room" is more practical.
It even has a built-in perspective change function. If you're interested, buy it!

I immediately shot it down👇

I quickly called in the marketing manager...

With this, a Google engineer and a marketing manager were summoned.

What's more, even my Dark Mihiro barged into the meeting, leading to an outrageous turn of events...


Meeting Member Introductions

The manga can be made with this free tool👇
(I just had it read the article and said, "Introduce the characters in 4 panels")

👩 Mihiro (Me)

  • 21 days of note history, 700 followers, a realist with 2,400 yen in sales

👨‍💻 Tanaka (Google Engineer)

  • L5 level, expert in technical implementation

  • "I can make a 3,000 yen tool!" - a dreamer

A trial version of this already exists...
It's currently available for a bargain price of 100 yen that would even surprise Google👇

👨‍💼 Sato (Google Marketing Manager)

  • Data-driven thinker, market analysis pro

  • Strategic, saying 'Let's build a recurring revenue model'

😈 Dark Mihiro

  • Cold-blooded analyst, my alter ego

  • Merciless, saying 'You guys are too soft'


Round 1: The Google team's 'too dreamy' proposals

Tanaka (Engineer): How about a 'Gradual Price Strategy Optimization AI'? For 12,000 yen!

Sato (Marketing): How about an 'AI Personality Branching Evolution Platform' for 50,000 yen?

Me: '...Um, I'm doing my best just to sell a 500 yen tool😅'

The wall of reality

  • My highest sales: 500 yen x 3 items

  • Many of my followers are beginners at note

  • Actual data on price resistance: Buyers dropped to one-third when going from 100 yen to 500 yen

Sato-san and Tanaka-san both toned it down immediately...

Round 2: A realistic strategic shift

Once I showed them my actual performance data, the Google team got serious.

My true strategy was revealed

Actually, by my third article after debuting, I was calculating to generate sales:

  1. Conducted a 'I'll buy your 100-yen article project' on Threads

  2. Purchased 3 articles to learn (Investment: 300 yen)

  3. Learned how to create GPTs from them and developed a 'Perspective Changer'

  4. Turned it into an article as a 'way to return the favor for what I was taught'

  5. Calculated that at least the original article author would buy it

The calculated Threads post I wrote based on my desires looks like this👇
(This might also be what they mean by 'Threads × note'…)

Sato (Marketing): 'Isn't this a perfect method for getting your first sale with a 300-yen investment!'

Tanaka (Engineer): 'That's a highly reproducible method...'

Revised tool proposal

  • Tool for getting your first sale with a 300-yen investment(500 yen)

  • Checklist for selecting articles to purchase(9-item evaluation)

  • Template for a 'returning the favor' article

  • Investment recovery simulation

This is realistic!

Round 3: Black Mihiro descends, 'Too soft'

Just by interjecting a few questions, the meeting proceeds on its own….

Blah, blah (it proceeds on its own), and even the development policy was decided.

Realistic Mihiro's honest impression: Hmm…. Something feels off.

Honestly, it doesn't feel right, so I'm playing my trump card here.
Since it's the Black version 2.0 that was upgraded the other day, it should be able to go toe-to-toe with a Google employee. I've called it in.

Black Mihiro: 'You guys are too serious. Let me teach you a darker method'

Black Mihiro's Strategy

Core: Instead of targeting people who have never bought an article, target people who have a habit of buying.

Specific Method:

  1. Create a list of people who have purchased others' articles

  2. Perform structural analysis of their worries and needs

  3. Create a product that hits the mark for one person

  4. Expand horizontally to groups with the same worries

Sato (Marketing): "This is... a super advanced marketing technique."

Tanaka (Engineer): "Technically speaking, the CVR (conversion rate) is likely to be abnormally high."

The terror of the 'Structural Tool'

The analysis method proposed by Black Mihiro:

【ターゲットAの完全分析例】
- 表面的悩み:「記事が読まれない」
- 深層悩み:「自分に価値があるか分からない」
- 真のニーズ:「小さくても成功体験が欲しい」
- 最適価格:200円
- 最適タイトル:「私も3記事目まで0円でした」
- 購入タイミング:絶望した夜

Me: "This is... completely mind-reading 😱"

The true greatness of Claude 4

Claude 4's capabilities I realized during this meeting:

1. Simultaneous management of multiple perspectives

  • Engineer perspective

  • Marketer perspective

  • My alter ego (Black Mihiro)

  • Each maintains consistent expertise

Of course, in the background, all my past article data,
Black Mihiro's strategy, and personality have already been installed.

2. Correcting the gap with reality

  • Initial dreamy proposal

  • Immediate strategic pivot after showing the data

  • Finally breaking it down to an executable level

3. Proposing even 'dark strategies'

  • Methods a typical AI would never suggest

  • But legal and effective

  • Proposals with a deep understanding of human psychology

Conclusion: Decided to start with a personal version

Me: 'That's an amazing strategy, but it's too dark and scary... I'll start by making it for myself first.'

Black Mihiro: 'Wise choice. Let's start by acquiring the weapon first.'

Sato (Marketing): 'Experiment and verify the results before commercializing. It's a Google-style MVP approach.'

The value of this article

What you will gain

  • Practical ways to use Claude 4

  • Strategic thinking at the level of a Google employee

  • Step-by-step product development process

  • How to think about realistic pricing

Note: Some strategies are at a 'sealed' level

Black Mihiro's 'structured analysis' method is so powerful that I have withheld the details this time.

If you are interested, please let me know your thoughts in the comments.

Meeting concluded...

For those who want to try Claude 4

For those who want to hold an 'expert meeting' like this one:

  1. Prepare concrete data(In my case: past articles, sales performance)

  2. Specify multiple experts(Engineer, marketer, etc.)

  3. Clearly state realistic constraints(Budget, time, skill level)

The 'dialogue with reality' is where Claude 4 truly shines.


Claude 4 was an even crazier tool than I thought.

Why don't you all try having a 'meeting with Google employees' in your own field of expertise?

You'll surely find strategies you couldn't come up with on your own✨


P.S. This article itself is my specialty of 'turning failure into content.' Even the process of the meeting becomes a product... this might also be a technique I learned from Black Mihiro😏

#Claude4 #Google #StrategyMeeting #AIUtilization #BlackMihiro #noteStrategy


いいなと思ったら応援しよう!

みひろ 「応援いただけると嬉しいです!ツール開発用のAIサブスク契約に使わせていただきます。今一番欲しいAIツールは、Gensparkです♪」