"No AI at Touchpoints": Our Website Development Policy

Hello. This is Lake and Moon Design.

We’ve been getting a lot of questions lately asking, “Do you use AI?” The answer is “Yes.” It saves time, helps with research, and, to be honest, it’s a tool we rely on quite a bit.

However, as part of our policy for website development, there’s one thing we’ve decided we simply won’t compromise on. That is, we do not use AI for client interactions—what are often called “touchpoints.” In all other areas, we use AI quite actively. Today, I’d like to explain this distinction as concretely as possible.

To put our approach simply, it boils down to these four points.

• We do not use AI at touchpoints
• Actively use AI wherever it saves time outside of touchpoints
• We won’t use it if it involves taking on unnecessary risk
• Do not ignore uncertainties

It’s not a simple choice between “using AI” and “not using AI.” It’s about where and for what purpose we use it. That’s what we prioritize.

Our Approach to “Where to Use AI and Where Not to”

Broadly speaking, we’ve organized it as follows: We don’t use AI at customer touchpoints. We actively use AI behind the scenes—in areas that lead to time savings and efficiency. And we decide not to use AI in situations where the risk is high or where we don’t fully understand how the system works.

Let’s discuss each of these points in a bit more detail.

What Exactly Is a “Touchpoint”?

A “touchpoint” refers to any point of contact between a company or brand and a customer (or someone who might become a customer in the future). While it’s a term commonly used in the world of marketing, the meaning itself is simple: it refers to every single “moment a person comes into contact with the brand.”

The moment you happen to catch a glimpse of an ad. The moment you open a website. The moment you see a social media post. The moment you walk into a physical store. A phone conversation. The moment you open a package you received. We experience countless touchpoints in our daily lives without even realizing it.

And touchpoints are not merely about the “number of interactions.” Each one is an opportunity to build impressions in the customer’s mind—whether it’s “I like this brand” or “This doesn’t quite feel right.” Surprisingly, how the experience felt tends to stick in people’s memories more than whether the design was polished or not.

Touchpoints Before, During, and After Purchase

When viewed along a timeline, touchpoints can be broadly divided into three categories.

Before a purchase is made, touchpoints include casually seeing the brand on social media or in online ads, finding the website through search results, or hearing word-of-mouth recommendations from acquaintances. The role of this stage is to “raise awareness” and “spark interest.” If the first impression here isn’t very good, it becomes difficult to get customers to proceed to the consideration stage.

At the moment when a customer actually makes an inquiry or signs a contract, interactions via forms or email, the wording of quotes and contracts, and how the customer is treated during meetings all come into play. Since this is a time when customers often feel a great deal of tension and anxiety, it’s important to prioritize giving them a sense of reassurance—that “a real person is taking care of them”—rather than treating it as mere administrative processing.

Touchpoints continue even after the transaction is complete—from follow-ups after delivery to after-sales support and the customer’s ongoing experience with the delivered product. Whether a customer feels they “made the right choice” depends not only on the moment of delivery but also on how the company engages with them afterward.

Touchpoints exist both online and offline

Touchpoints aren’t limited to just websites. Online, they include homepages, social media, email newsletters, and review sites. Offline, they encompass business cards, brochures, store interiors and exteriors, customer service, phone interactions, and mailed materials.

In creative fields like ours, touchpoints include not only the “products we create”—such as designs, photographs, copy, illustrations, websites, printed materials, and logos—but also the “way we interact” with customers, including customer service, the atmosphere of the store, the demeanor of staff, and how we treat customers. When you list them all out like this, you realize that touchpoints actually cover quite a broad range.

Why Are Touchpoints So Important?

The reason we should value touchpoints isn’t simply because “they create a better impression.”

If each individual touchpoint provides a positive experience, those experiences accumulate and lead to the perception that “this brand is trustworthy.” Conversely, even a single instance of sloppy service can instantly shatter the trust built up to that point. People naturally want to share positive experiences with others. The quality of touchpoints directly influences the rate at which customers recommend the brand to others.

If a brand’s “authenticity” is maintained at every touchpoint, the brand’s identity becomes clearer. Inconsistent impressions are hard to remember. Precisely because we live in an era where it’s difficult to differentiate products and services based on features alone, I believe the quality of the experience at each touchpoint is what ultimately makes a brand the preferred choice.

That’s why we don’t want to just “tweak” our touchpoints haphazardly—we want to design them with a clear purpose. And to be honest, we don’t want to leave that design and execution entirely up to AI.

We want people to create the things that people touch and that move their hearts

Saying “we don’t use AI for touchpoints” isn’t simply a matter of “we don’t use AI to create designs.” What’s important is why we don’t use it.

The things customers pick up, look at, read, feel, and what moves their hearts—that’s why people think about them, create them, and respond to them. That is our core value. This isn’t a rejection of AI—we simply believe that the nature of the work we want to entrust to AI is fundamentally different from the work we want people to take responsibility for.

We do not believe that a brand’s “authenticity” can be reproduced simply by learning from past performance data or trends in customer preferences. The nuances of words that emerge during conversations with customers, a single color chosen after much hesitation, or a balance deliberately thrown off—it is the accumulation of these detailed decisions that defines “authenticity.” This is a different kind of work from efficiently arriving at the “right” answer.

While speed is often seen as a virtue, when it comes to touchpoints, I feel there is value in taking the time to think things through. This is because ideas that emerge after discarding numerous options—rather than those that come to mind immediately—often leave a lasting impression on clients.

What we’re really selling isn’t the website itself

We call ourselves a “website design firm.” But I don’t think what we’re actually delivering is the website itself.

What we value most is listening to our clients face-to-face and solving each individual’s problems based on the authentic insights that emerge from those conversations. In other words, what Lake and Moon Design sells is “communication” and “problem-solving.” The website is merely one tangible result of that process.

Precisely because AI is evolving so rapidly, we believe there is value in “things only a person can tell you.” There’s no information better than a person’s voice. That’s why, rather than fearing AI, we aim to coexist with it effectively while using the human voice as our starting point. That is our stance.

So, do we not use AI? → No, not at all—in fact, we do use it

If you’ve read this far, you might think, “So Lake and Moon Design is a company that doesn’t use AI.” Actually, that’s not the case at all—we use AI quite a bit.

We actively use it for everything except direct customer touchpoints.We rely on it without hesitation for tasks like research and preparatory work for organizing information, brainstorming with a sounding board, drafting initial versions of text (which we never submit as-is), streamlining simple or repetitive tasks, assisting with coding and debugging, organizing internal documents and meeting minutes, conducting competitor analysis and market research, and managing schedules and tasks.

Our approach here isn’t to “replace work that people should be doing with AI,” but rather to “use AI to free up more time for people to think.” We don’t let the time saved by AI go to waste. We use it to deepen our discussions with clients, expand the range of proposal options, and review the finer details.The goal of streamlining is to free up more time to focus on the tasks that only humans can do.

We don’t use AI just to take risks

Just because AI is convenient doesn’t mean we’ll adopt a “if AI can do it, let’s just give it a try” approach. We’re particularly cautious about the following:

Do AI-generated designs or illustrations resemble existing works too closely? Does the text simply replicate phrasing that seems familiar from somewhere else? Since disputes over intellectual property rights are often irreversible even if discovered later, we thoroughly verify everything at the time of generation.

We also do not casually enter client information or the details of projects that have not yet been made public into external AI services. Even if a tool is convenient, we avoid using it if we’re unsure how the information we enter will be handled.

Even if AI-generated output seems plausible at first glance, the details can sometimes be incorrect. Especially for areas requiring specialized knowledge, we don’t take it at face value but include a step where a human reviews it.Our mistakes aren’t just our own problem—they also affect our clients’ brands and trust. That’s why we assess the magnitude of risk not by “how we’ll be perceived,” but by “how it will impact our clients.”

We have a very clear rule: if we look at something generated by AI and think, “Well, I guess this will do,” we’ve decided not to release it. We protect ourselves. We don’t cut corners with our clients. These are the two things we absolutely won’t compromise on.

“It’s too much trouble” is no reason to use AI

The mindset we consider most dangerous when it comes to AI is thinking, “It’s a hassle, so let’s just let the AI handle it.” I understand the desire to take the easy way out. But if this becomes our sole criterion, the line between what’s acceptable and what isn’t will collapse in the blink of an eye.

The standard we want to use isn’t “whether it’s easy,” but “whether it’s reasonable.” Why are we letting AI handle this part? Why is a person doing that part? We need to be able to explain it in our own words. If we can’t do that, I believe our judgment is still too superficial.

Conversely, avoiding AI even when there’s a rational reason to use it is actually a waste. Not using it simply because “I feel a little uneasy” might just be a case of refusing to try something new out of prejudice. That’s precisely why we want to be able to clearly articulate both the reasons for using AI and the reasons for not using it.

Don’t leave “I don’t know” unaddressed

This mindset goes hand in hand with “not taking risks.”

If you ask AI, “Make this for me,” it will return something that seems plausible. But how does it work? Why does it behave that way? Is it safe? To be honest, I find it scary to use it without understanding those details.

For example, let’s say we’re building a website and have an AI create a contact form system for us. In that case, we always make sure to verify how the system works, where the entered information is sent, how personal information is handled, whether there are any security issues, and what happens if unexpected input is received.

This isn’t limited to contact forms. It applies to payment systems, user registration processes, and integrations with external services. Especially when we ask AI to help with the “back-end mechanisms,” we make a point of not basing our judgment solely on the end result.If there’s something we don’t understand, we ask the AI again, review the documentation, or check with an expert. It takes effort, but we believe that not settling for “it seems to be working, so it’s fine” ultimately leads to protecting our customers.

“It works, so it’s OK” is not good enough. If there are issues with the code or systems generated by AI, we as humans need to notice them and fix them. Having AI create something and leaving everything entirely up to AI are two completely different things.

Using AI Isn’t Just About “Making Life Easier”

For Lake and Moon Design, AI isn’t a tool meant to eliminate human jobs; it’s a tool that allows people to spend their time on the work they truly want to focus on. We use it where it’s useful. But we don’t use it where it isn’t. If we don’t understand it, we simply don’t use it. Our criteria are extremely simple.

How We’ll Engage with AI Going Forward

AI technology will continue to evolve. Just because we decide “not to use it” today doesn’t mean that decision will remain unchanged forever. As the technology becomes more reliable, we intend to reevaluate where we draw the line.

However, there is one thing we want to keep unchanged: never making decisions based solely on the fact that something is “convenient.” We will maintain our commitment to continually asking ourselves, on a case-by-case basis, whether a particular choice is truly in our clients’ best interest.

How Is This Different from a Site Created Entirely by AI?

Recently, there has been a rise in services that allow you to create a website in just a few minutes simply by giving instructions to AI. If you’re looking for affordability and speed, I think that’s certainly a valid option.

What sets us apart from those services is that we don’t compete on speed or price. Templates automatically generated by AI tend to result in websites with similar structures and similar wording. That isn’t necessarily a bad thing in itself. However, the key difference is that it’s difficult to capture the unique “essence” of a specific brand within those templates.

At Lake and Moon Design, we don’t simply plug information into a template; we craft websites by carefully refining the brand’s unique language and atmosphere, step by step. That’s why we invest time in consultations and never leave the copywriting or design entirely up to AI. To be honest, our approach might not be the best fit for those who prioritize speed and affordability above all else.However, for those who value the “essence of this store” or “essence of this company,” we believe our approach is a better fit.

Frequently Asked Questions

Q. Do you create websites without using AI at all?
No, that’s not the case. We do utilize AI for research and to streamline our workflow. However, the parts that clients see—such as design and copy—are handled by humans.

Q. Will production costs be higher since you don’t use AI?
We do not charge extra simply because we don’t use AI. For us, this isn’t a “special service”—it’s simply our standard way of working.

Q. Does it take longer to deliver compared to sites automatically generated by AI?
Compared to AI-powered automated generation services, the time spent on meetings and production is longer. In exchange, we conduct multiple rounds of consultation to work with you in finding the words and designs that truly reflect your brand.

Q. Do you avoid using AI for coding and system development as well?
We actively utilize AI in those areas. However, we don’t simply accept something as “good enough” just because it works; we ensure we understand the underlying mechanisms and perform a final manual check before delivery.

Q. Do you think the scope of AI use will expand in the future?
I think it’s possible. Our current boundaries are based on a judgment made in light of current technology and circumstances. If AI technology or reliability changes, we’ll reassess our approach.

Precisely because we’re in the AI era, we want to value human-made work

Finally, let me summarize our four guiding principles once more.

• We do not use AI at touchpoints
• Actively use AI wherever it can save time outside of touchpoints
• Do not use AI if it involves risk
• Don’t ignore what you don’t understand

Precisely because AI is becoming an increasingly familiar part of our lives, we need to consider what tasks to entrust to AI and what tasks should be handled by people. I feel that thinking about this will become increasingly important going forward.

At Lake and Moon Design, we won’t just chase convenience—we’ll continue to work effectively with AI while cherishing the value of what people convey to one another.