Mastering Claude Code

Day 1: Loose Direction Setting and Trust Map Creation

Monday morning, I opened my laptop. LinkedIn, email, and a notes app were already open in the browser. "Starting today, I'm launching an AI automation business." But the question remains: what should I do first?

Most people make a mistake at this point. They try to define an ultra-specific niche first. Starting with something concrete like "I'll target dental clinics in Georgia that are wasting time on patient onboarding" actually locks you in place. You have no data yet on which markets respond, or which problems come up repeatedly.

Your direction on day one is more like a working hypothesis. It's not a lifetime commitment. This much is enough: "I help small and medium businesses automate repetitive, tedious work with AI." One sentence is all you need to start a conversation.

On top of this loose direction, prepare a few example problems. Things like lead follow-up automation, processing customer intake forms, syncing data between CRMs. You don't need to commit to these yet. They're just test cards to see which problems resonate in conversation.

Your second task on day one is to map your trust network. Open a Google Sheet. Put names in the rows, and fill five pieces of information in the columns.

[Figure 33-1] Trust Map Spreadsheet Example]

Your goal is to fill in twenty people. Include business-owning friends or family, work colleagues or managers, people you know from communities or Slack groups, and even second-degree connections, friends of friends. Something remarkable happens when you actually write this list down. You realize that your belief that "I don't know anyone" was an illusion. Most people underestimate their own networks.

This map matters because your starting point should be warm outreach. Statistically, warm contact and introductions convert far better than cold outreach. Trust already exists, and even borrowed trust from a friend of a friend is stronger than an email from a stranger. It's faster. It's less awkward.

Here's the big picture. Start with warm relationships, keep a broad scope, talk and find patterns, then narrow your niche later. Cold outreach is a tool you use when you're ready to scale, not when you're hunting for your first client.

Day 2 to 3: Five to Ten Warm-up Conversations

The goal for days two and three is not sales. It's conversation. All you're doing is having five to ten light conversations.

Be clear about what this conversation is. It's not a sales call. It's not pitching. It's a curious entrepreneur asking questions. You could start with a message like this:

There's a phrase you can't leave out of this message: "I'm not trying to sell you anything." That one sentence drops the other person's guard. People get defensive the moment they sense someone is trying to sell them something. Disarming that defense is where conversation begins.

Once the conversation starts, record the key insights. It doesn't matter if it's a notes app, a note-taking tool, or paper. What matters is the act of recording itself. You'll look at these notes again later. Here's what to capture:

The fourth item is surprisingly important. The words the other person uses to describe their problem become your marketing language later. You have to speak the business owner's language, not the technician's language, for resonance to happen.

If you truly have no business owners in your network, here's what to do. You still don't need to jump into cold outreach. Instead, ask like this:

The power of this one sentence is bigger than you think. It's not about selling your friend something, but asking for an introduction, so the awkwardness drops. And when you meet someone through that introduction, saying "OO introduced me" creates instant trust.

[Figure 33-2] Warm-up Conversation Insight Recording Template]

Day 4 to 5: Small Pilot Proposal

Spread out your conversation notes. You have five to ten conversations documented. Pick the person with the clearest pain point. High frequency, time-consuming, repetitive work. A problem where all three conditions overlap is the ideal pilot target.

Once you've decided, reach out.

This proposal works because there's almost no risk. From the client's side, there's no cost. If it fails, they lose nothing. If it succeeds, they save time. And because they're giving feedback in return, it feels like an equal exchange, not charity.

There's a principle you must follow when making the proposal. Keep the scope extremely small. Not "I'll automate your entire sales process," but "I'll build a workflow that automatically categorizes incoming leads and sends alerts to the right person." Focus on one action. You need to start small to finish, and you need to finish to prove it works.

Day 5 to 6: Build the Minimal Automation

The pilot proposal has been accepted. Now it's time to build.

There's a trap that's easy to fall into at this stage. The urge to build something technically impressive. You might feel like layering multiple API calls, adding complex branching logic, or applying the latest models. But the goal isn't to impress. It's to deliver results.

The standard for minimal automation, or MVP, is simple. "Did the client save time after this workflow started running?" If you can answer "yes" to that, you're done.

As you build, pay attention to the language your client used when describing the problem. If they didn't say "lead management" but instead said "organizing the daily flood of inquiries," use the same language in your reports. This language mirroring becomes a powerful tool in future proposals and sales messages.

[Figure 33-3] MVP Workflow Building Checklist]

The building process doesn't need to be complicated.

1. Define the problem in one sentence. 2. Make inputs and outputs clear. 3. Design the shortest path from input to output. 4. Build and verify. 5. Show the client the results.

That's all. Making it complex is something you do later in the scaling phase.

Day 7: Maintenance and Expansion Proposal

Day seven arrives. The pilot is running. Is it time to push hard for the close? No.

The goal for this day is not closing. It's deciding together what comes next based on the pilot results. A pushy tone can destroy the trust you've built in an instant.

The structure of the conversation is simple. Present two options.

Option 1: Maintenance. Keep the automation you've already built running smoothly. Fix it if something breaks, adapt it if tools or processes change. "I'll manage this automation over the next few months to keep it running well. If something breaks or needs adjustment, I'll respond right away."

Option 2: Expansion. Add one or two features on top of the pilot. Use the improvement ideas you discovered while building. "While building the pilot, I had an idea that adding this piece would make the results more stable. Should I scope it out and send you a proposal?"

When you present both options without pressure, the client naturally chooses the direction that fits them.

There's one advanced technique here. If you discovered related tasks while building, ask naturally. "While creating the workflow, I also noticed this process. Would it help to automate this part too?" This question is the language of collaboration, not sales. The client will feel that you're genuinely interested in their business.

Sometimes the pilot succeeds, but the client doesn't want additional work right now. That's not a problem at all. The strategy for this situation is explained in the next section.

The Order of Testimonials and Referral Requests

There is an order. Breaking this order breaks trust.

Discussion of maintenance or expansion comes first. Deepening the relationship is the priority. Next comes the testimonial. And last is the referral request.

If the client is satisfied with the pilot but doesn't want additional work, you ask like this.

Video testimonials are more powerful than text. When an actual business owner talks about their experience in front of the camera, it carries more persuasion than any marketing copy.

Only after receiving a testimonial, and only when the flow feels natural, do you mention a referral.

Let's think about why this order matters. Asking for a referral when you've done nothing yet feels awkward. Asking for a testimonial when you've done something but the results were poor feels even more awkward. The order follows logic: deliver value, deepen relationship, secure evidence, expand network.

Let's look at what to do if the pilot fails. You're not trying to sell something. You don't ask for a testimonial. You gather feedback, organize what you've learned, and start the next cycle.

[Figure 33-4] Flow diagram of testimonial to referral request]

Repeating the Seven-Day Cycle and Scaling to Cold Outreach

The Seven-Day Framework is not a one-time event. It's a loop you repeat.

In the first cycle, your messaging is clumsy. You have no sense of which solution fits which problem. In the second cycle, your message becomes shorter and more concrete. You start to feel which conversations gain energy and which lose direction. By the third cycle, you develop your own language. You learn how to borrow the customer's expressions, and the tone of your proposals becomes natural.

What each cycle provides is data. Confidence goes up, but the real value is in the data. Patterns accumulate about which industries respond, which problems appear frequently, and which message structures open conversations.

When these patterns have accumulated enough, usually after running two or three cycles, scaling to cold outreach becomes meaningful. This is because at this point you have built three things.

1. Evidence: a pilot or project that actually delivered results 2. Language: message structures the market responds to 3. Confidence: capabilities verified in real-world application

Cold outreach without these three elements is no different from someone sending hundreds of contextless emails a day. Messages sent without evidence, without trust, and without real connection are just noise to the recipient.

[Figure 33-5] Seven-Day Cycle Repetition and Cold Outreach Expansion Roadmap]

This was exactly the path followed by a practitioner in an AI automation community in the United States who secured their first client. They started with warm relationships, took a results-driven approach, built trust, delivered value, and repeated. Only after running two or three cycles did they build a prospect list and apply the same framework to cold contact. By then, it felt like operating a working system, not acting out of guesswork and anxiety.

Securing your first client is just the beginning of your entire business. To sustain and expand this relationship, you need to know how to handle the sensitive subject of pricing. Will you charge by the hour or by value? This choice determines your business trajectory.

Kim Kyung-jin, AI Expert and Attorney

AI Policy and Law Specialist · Former Member of National Assembly · Author of numerous works

If this book has stayed with you even for a moment, please support it so the next story can come into the world.

(Voluntary support requested - Account: Nonghyup 302-1096-0948-81 Account holder: Kim Kyung-jin)