At the start of our twelfth fiscal year of operations, Mercari embraced Back to Startup and AI-Native. These topics demand fundamental changes in our organization and product that use AI as the cornerstone of all our work. At the same time, they give us opportunities to re-engage with the spirit of high-speed trial and error and the bold challenges that we experienced as a startup.

Amid this transformation, we took a minute to ask the company CTO playing a key part in our plans to become an AI-Native company how they see the ensuing changes and what they think of them. We sat down with @kimuras to take a deep dive into the current state of our efforts.

*This article was composed and written using AI tools based on the day’s audio recording. This English version was translated from Japanese using a custom generative AI model interface trained using Mercari style guides, glossaries, and past Mercan articles.

The strategic relationship between AI-Native and Back to Startup

—Mercari has embraced Back to Startup and AI-Native as topics to pursue this fiscal year. Could you share what each of these means to you?

Let me start by explaining why we need to implement Back to Startup. For Mercari to continue achieving disruptive growth, we need to return to the high-speed experimentation and bold spirit of challenge that defined us in our startup days. However, as our organization has expanded to over 2,000 members, approval and decision-making processes have of course become more complex. A short while ago, we noticed that our overall speed had started to slow down. This organizational challenge was part and parcel of our rapid growth. Yet the emergence of AI is fundamentally changing the situation.

The key point is that AI-Native exists as a natural extension of Back to Startup. Rather than simply introducing AI as a tool, we are positioning AI as a core premise of how we run our organization. This approach allows us to achieve the speed and decisiveness we aimed for under Back to Startup while maintaining our current scale.

AI has the power to infuse speed—normally possible only at smaller startup companies—to large-scale organizations as well. Through AI-Native, Back to Startup is more tangible and sustainable.

—When you talk about fundamental change, what kind of change do you envision specifically?

At AI-Native companies, the boundaries between job roles will shift significantly. It will become common for a backend engineer to handle frontend work as well, or for a product manager to take on design responsibilities. In the past, the boundaries of expertise made it necessary to divide tasks between specialists—for example, one person for frontend and another for design. Now, with AI supporting coding and design work, individuals will gradually be able to work across multiple domains.

As a result, a new workstyle emerges—what we might call “self-collaboration.” Conventionally, collaboration has involved people working together on a specific project, but from now on, we will see a two-stage process in which many people join in the planning phase for discussion, and a small expert team then drives production forward rapidly. With AI support, Back to Startup will become more tangible and effective.

Background and strategy behind the creation of the AI Task Force

—— Could you tell us more about how Mercari formed and operates the AI Task Force?

We prioritized speed and decided to form it after just a few days of discussion. Out of roughly 2,000 employees company-wide, we selected about 100 members and defined three roles: enablers, owners, and PGMs (program managers).

Enablers lead AI promotion in each domain, owners are decision makers for business improvement, and PGMs handle overall coordination. What’s important is that this is a six-month focused initiative. We start by conducting a comprehensive review of company-wide operations and ROI analysis while keeping the impact on business growth minimal.

The primary goal is to create momentum through “quick wins”—small successes achieved in a short time—and embed a self-driven culture of AI utilization within each domain. We believe that moving forward agilely rather than aiming for perfection is the core of how our organization should operate from now on.

—Please tell us about the background and strategy behind launching the task force.

We had been introducing AI to the company gradually for about a year before launching the task force. It started with the Eliza Team, which focused on LLM utilization, and we made significant progress in applying AI within our products. LLMs have dramatically improved several features, such as listing items from images, search engine results, and recommendation functions.

Since the start of 2024, thanks in part to the leadership of AI initiative driver Gomi Hayakawa (@gomichan), AI performance has improved exponentially. We began to see tremendous potential not only for use in our product but also for improving entire business processes. However, improving individual tasks alone does not lead to greater company-wide output. Unless we adopt AI across the entire company to improve workflows, we cannot achieve the fast-paced AI utilization envisioned by Back to Startup.

That’s why we decided to invest significant resources and budget into AI, knowing it would somewhat impact short-term business progress. We are conducting a company-wide audit of operations and prioritizing the shift to AI-Native of areas with the highest ROI.

Moreover, even the audit documentation itself will become a valuable asset for the future. We plan to use it as a kind of roadmap to show how much of our operations we can rebuild around AI and then continue implementing it across all domains even after the six-month audit. This is because the actions that can realistically be completed in six months are limited to certain areas identified in the audit.

Redefining experiences and Mercari’s competitive strategy

—You announced not only the formation of a task force but also a reevaluation of both the user experience and our employees’ workstyles. Are these two areas interconnected?

Our basic premise is that without redefining our workstyles, overall output will not increase. We need to advance both product improvement and productivity enhancement in tandem.

We are aiming to create a “zero-tap” user experience. It will be different from the existing step-by-step process of opening the app, selecting a category, and setting search conditions. We aim to create a fundamentally new experience that allows users to complete any listing by just pointing their smartphone camera, or to find items using conversational language that includes giving the brand, the average price, and some other details, like “not too worn.”

On the other hand, in terms of workstyles, having individuals take on multiple roles reduces communication costs and accelerates decision-making. For instance, if someone in a development team has legal or security knowledge, they can assess risks on the spot. This dramatically increases the speed of hypothesis testing.

—As Mercari undertakes these transformations, how do you perceive the company’s unique competitive advantages?

First, services with a significant market share will be able to maintain a strong position even in the AI era. This is because having data and users already in place provides a major advantage in this new age. However, these elements alone will not be enough.

Mercari’s true strength lies in its culture of actively taking on new challenges. The synergy between the Task Force’s efforts and Mercari’s culture of encouraging innovation has further increased company-wide AI tool usage. This is a trait of our organization that other companies cannot easily replicate.

In addition, the strength of our learning cycle and our All for One mindset are key differentiators. Looking back, during our expansion to the U.S., we dispatched more than half of our employees, and when we established Merpay, two-thirds of our employees were involved. This collective focus on a single goal during major transformations is what I believe to be Mercari’s unique competitive advantage.

The qualities needed to succeed in the AI era

—As organizations and workstyles evolve, what kind of talent and mindset do you think will be required going forward?

The most important thing is that the strong-willed people among us survive and continue to produce results. By “will,” I mean a strong determination to shape our service in a certain way and deliver specific value to users.

In an age when everything is optimized, it’s possible for one person to achieve ten times more than before. In an environment like that, small teams will naturally consist of people who all have strong willpower.

In other words, it will become harder to thrive without leadership qualities. Leadership means having the ability to make decisions and envision the future. In the fast-paced AI era, this ability will be extremely important.

What will matter most will not be your background—for instance your past career—but whether you possess the ability to lead and make decisions. And on the other hand, if you are strong-willed, AI will support you, making it possible for anyone to succeed. It’s inevitable that a divide will open up between passive and proactive individuals.

Going forward, it will be essential not only to rely on past experience but also to actively harness new capabilities. As people work alongside AI, those with strong willpower will find greater opportunities, while different approaches will lead to different levels of success.

Refining existing assets and growth strategies

—How do you think AI will transform Mercari’s growth, and what kind of future do you envision?

For companies like Mercari, which possess sizable assets, maximizing those assets yields higher cost-effectiveness. I also believe this can have a greater impact on society. While creating new businesses is important, carefully refining and enhancing each of our existing assets will be the key strategy in the times ahead. Launching new businesses in a thoughtful and strategic manner will lead to greater efficiency and broader social impact.

So how do we go about refining them? The answer lies in dramatically improving operational efficiency. The first step is to update our mindset. We must break free from conventional thinking that dictates what we can or can’t do because something requires too many people or is too expensive. To prepare for the coming era, we need to shift our mindset toward seeing what will be possible going forward.

For instance, an international expansion that once required 30 people for one market could now potentially be achieved with just six. Improvements in efficiency could make things once thought impossible entirely feasible, such as simultaneous global expansion or authenticity checks for all items.

This shift in mindset is at the very heart of growth. More than simply improving efficiency, the impact of fundamentally changing the framework for our ideas is one of our most powerful strategies.