“Prompt Engineering” Works on Humans Too—A 72-Hour Observation Report on a Top Salesperson
Summary at the start
I, a CEO with a technical background, took on the challenge of sales, which I am not good at.
I shadowed a top salesperson (Mr. O) for three days to observe and verbalize his thoughts and actions.
I installed these as a “self-prompt” and put them into practice immediately; as a result, I closed a deal in my first sales meeting (N=1).
Conclusion: By observing the actions of excellent people and turning them into “prompts,” anyone can update their own abilities.
Hello, this is Serikawa.
Since becoming CEO, a new mission has been added to my role, which was previously focused entirely on technology as CTO. That mission is “sales.” To be honest, I had a huge complex about it. I didn't really know how to do it, and I had never succeeded when I tried in the past. I felt that things were not looking good if I continued like this.
In the midst of this, a major turning point occurred during a business trip to Tokyo the other day.
I had the opportunity to shadow Mr. O, our company's top producer and a highly skilled salesperson, for almost three days. I decided to explore the secret behind why he is so trusted and able to achieve such results by witnessing his interactions with clients right beside him.
Day 1 & 2: Observation and Discovery — Why does Mr. O sell without “selling”?
For the first two days, I focused on thoroughly observing Mr. O's sales style and verbalizing its essence in my own way. Watching from his side, the reasons why he is liked by clients and why it naturally leads to contracts became clear in an interesting way.
Surprising ability for self-disclosure: He openly talks about his recent minor failures even to people he has just met. This instantly relaxes the atmosphere and makes it easier for the other party to speak their true feelings.
“Non-selling” sales: Mr. O does not directly pitch what we want to sell. He keeps it strictly on the level of casual conversation, saying things like, “We’ve been doing this experiment lately.” Then, the client becomes interested and asks questions, saying, “Could you tell me more about that?” This structure of “drawing it out from the other party” was the key.
Strong commitment to Win-Win: In the position of contract development, it is easy to become passive, but Mr. O always tries to build an equal relationship with the client. He clearly demonstrates the value that “it’s meaningless unless both parties are happy” and is committed to providing convincing prices and value.
Attitude of engaging with honesty: He honestly conveys negative information, such as when a budget is tight. That is precisely why clients can feel safe speaking their true feelings. The relationship of trust that “it’s okay to be frank with this person” led to smooth work progress. (This might be due to his natural cheerfulness, but I felt that it is also a large part of what he is conscious of.)
Emphasis on long-term trust relationships: He values each and every project and builds relationships through past achievements that make people think, “I want to ask Mr. O again.” It is truly the accumulation of a “trust balance.”
While taking notes during my observation and analyzing his principles of action, I suddenly realized something.
“Isn't this just like prompt engineering for AI?”
Prompts give clear instructions and context to AI to elicit excellent output. I formed a hypothesis: if I structure Mr. O's thoughts and behavioral patterns as a “prompt” and install them in myself, couldn't I achieve the same results?
Day 3: Practice — Installing the “Pro Sales Prompt” into myself!
Then came the third day. I reminded myself of the “pro sales prompt” I had articulated by the previous day and finally embarked on a sales meeting led by myself.
Specifically, I was conscious of the following “self-prompt”:
【商談開始プロンプト Ver.1.0】
トリガー：クライアントとの挨拶直後
アクション：
1. まず自分の最近の小さな失敗談（事前に用意）を軽く共有し、場を和ませる。
2. 売りたいサービスの話はせず、「最近こんなチャレンジをしていて…」という切り口で雑談を開始する。専門用語は使わない。
3. 相手からの質問を待つ。質問が出たら、熱意を持って、しかし分かりやすく説明する。
4. 常にWin-Winを意識し、対等な目線で話すことを忘れない。
制約：絶対にこちらから商品を売り込まないこと。
I repeated this in my head over and over, and then it was time to put it into practice.
And then... surprisingly, in just a 30-minute meeting, I was able to elicit a positive response from the client, along with the words “Please send us a quote,” and even secure a concrete appointment for the next meeting!
Initial meeting closing rate: 1 / 1 (100%)
Of course, since this is an N=1 (sample size of 1) result, I cannot claim everything based on this alone. But for someone like me who had never achieved results in sales before, it was an incredibly significant breakthrough.
After the meeting, I immediately conducted a review, identified what went well and what needed improvement, and updated my “self-prompt” to Ver. 1.1. I am convinced that continuing to run this verification loop is key.
Why was I able to replicate it? The key is “self-prompt engineering”
Through this experience, I realized that excellent skills are by no means the exclusive domain of a few geniuses, but can be replicated and installed to some extent through a cycle of observation, articulation, and practice.
High-speed cycle of observation → articulation → practice: Instead of just watching the actions of excellent people, thinking deeply about “why they do that,” defining it in your own words (= turning it into a prompt), and putting it into practice in a short period (within 72 hours in this case) makes the learning easier to retain.
The power of clear instructions (prompts): Humans are the same as AI; we can move without hesitation when we have concrete action instructions rather than vague goals. A “self-prompt” is a blueprint for intentionally designing your own actions.
Speed of the verification cycle over success rate: Rather than aiming for perfection from the start, running the “try it first → improve it” loop at high speed accelerates skill acquisition.
This is nothing special. You could say it is just re-framing the age-old wisdom of “imitating someone who is good at what they do” through the lens of “prompts” in the AI era. But this perspective became the catalyst that dramatically changed my own actions.
Are you setting your own limits?
After reading this article, some of you might have thought, “It sounds interesting, but it has nothing to do with me,” or “That’s only possible because it’s you, Serikawa-san.”
But I want to ask you: “Are you setting your own limits?”.
Everyone must have someone they admire, a colleague they think is “amazing,” or a boss they respect. Why not observe what makes them great and why they are able to achieve results with a slightly higher resolution, and try stealing their essence?
The first step to “self-prompting” you can take today
Observe the person you admire for 24 hours: Let’s consciously observe that person’s words, actions, and thoughts. (Social media, videos, past statements, etc., are also fine.)
Turn the behavior you want to emulate into a 3-line prompt: Try turning the question, "If I were that person, how would I think and act in this situation?" into a short set of instructions.
Read it aloud first thing in the morning, and try it immediately: Tell yourself that prompt, and try putting it into practice, even in a small situation.
Call for Action
If you try this "self-prompt engineering" experiment yourself, I would love for you to share your results or thoughts with me.
Feel free to post about it on X using the hashtag #SelfPrompt.
Feedback from everyone will fuel my own next verification loop, and it might even become a new strength for the entire company.
Let's rewrite our own limits together.
"Limits can be rewritten."
no plan inc. CEO Aoi Serikawa