How to Start a Bank: Kakao Bank's Case study (World's fastest growing digital bank)

 

Kakaobank's timeline

Date	Description
June 2015	Announcement of Internet-Only Bank Licensing Plan (Financial Services Commission, Financial Supervisory Service)
October 2015	Submission of Preliminary Authorization for Internet-Only Bank
November 2015	Completion of Preliminary Authorization for Internet-Only Bank (Kakao Bank, K Bank selected)
January 2016	Establishment of "Kakaobank Preparatory Corporation"
April 2017	Kakaobank gets Final Authorization banking license
July 2017	Begin operation

 

What does it mean to start a bank, and where should we begin?

 

I joined Kakao Bank in November 2016 during its preliminary approval phase and worked until August 2020 as a practitioner in charge of partnerships and new businesses in the strategy department. I would like to share my experiences during this period on "How does a new bank come into being?"

 

 

We had never had the experience of building a bank before

After the establishment of the preparatory corporation, our office was set up in Pangyo in March 2016. Employees from Korea Investment & Securities, Kakao, and Kookmin Bank began to convene and initiate discussions on service ideas. While we had not solidified any specific business plans, the banking system, including core banking, wouldn't significantly vary depending on the nature of the business. Consequently, the process of bank establishment began simultaneously with the initiation of dispatching Requests for Proposals(RFP) for system construction.

 

The engineers from Kakao were eager to engage in 100% in-house development. But, creating a core banking system was a proposition of an entirely different magnitude.

 

 

To give a brief introduction to the banking system, when a customer requests a transaction through the bank app, this request is received and relayed by the channel system, then delivered to the account system (core banking), which serves as the transaction ledger.

 

The reason for having this bridge, called 'channel system', in the middle is because it must be absolutely flawless since it stores the original records of transactions.

 

At that time, Kakao Bank adopted the Jeonbuk Bank's model developed by LG CNS based on x86, Linux, and Java.

 

We had to quickly introduce a model that had already been verified, and the core of customer service was the channel system rather than the account system. On the other hand, K Bank(South korea's another Internet-only bank) chose Unix and would switch its account system to Linux a few years later.

 

 

 

In March 2016, the development of the core banking system began, and the people at Kakao Bank started gathering ideas, jotting them down on post-it notes and pruning as they went along.

 

Many ideas that were initially planned in the preliminary approval application were dropped at this stage. A prominent example was the 'Kakao Bank Universal Points', a seemingly attractive idea that would give back more benefits to customers in the form of shareholder benefits instead of interest, if they so desired.

 

However, we all ended up asking the fundamental question, "Would this truly be a beneficial reward for the customer, or just a nice-sounding marketing gimmick?" Through such deliberations, we all came to the conclusion that a new bank should save customers' time, assist them in raising funds quickly and at a low cost, and provide an easy and enjoyable saving experience.

 

From this essence, product and service planning began.

 

 

Is it necessary to input the account password every time when making a transfer?

In 2016, the existing internet and mobile banking services of traditional banks were honestly not too bad.

 

Therefore, we needed to differentiate ourselves.

 

As a latecomer, we reevaluated everything from scratch. A fierce internal debate at Kakao Bank at that time was whether to operate PC-based internet banking or just an app. These days, people don't use internet banking much, but as of 2016, individual customers used internet banking more, with a ratio of 100:70 for internet banking to mobile banking.

 

However, simply thinking about it, the fact that a bank has to provide internet banking was a fresh point that we didn't even consider because it seemed so obvious. Honestly, we were already spending billions on system construction, and it would have cost about 2 billion won to create an internet banking site through outsourcing at the same time.

 

This issue sharply divided employees from financial and non-financial backgrounds. Daniel (CEO Yoon Ho-young), who came from Kakao, and Yan (former co-CEO Lee Yong-woo), who came from Korea Investment & Securities, usually made decisions smoothly while respecting each other's expertise. However, they opposed each other sharply on this topic.

 

After a fierce debate, it was decided to only operate a mobile app. Since it became a single channel, all efforts were concentrated to redraw mobile banking from scratch.

 

At this point, three decisive factors were determined.

 

1) Issue of entering a bank account password when transferring: Whether it was internet banking or mobile banking, it was a very natural procedure. But no one could explain why this was natural, and why it had to be done. There was no content in the Banking Act or the Supervisory Regulations that required confirmation of the bank account password when transferring. It was a custom of existing banks. So we removed the step of entering a bank account password in the transfer scene.

 

2) Encourage '1 won certification' instead of video call for non-face-to-face real-name verification: Because there was no face-to-face branch, non-face-to-face identification had to be done when opening a new account. If the video call method was chosen, it was possible to attract customers who did not have a bank account or who were not good at using internet/mobile banking. Although it seemed like a business-friendly decision, from the customer's point of view, a service delay would definitely occur because they had to wait a long time for a video call with an employee. Therefore, we chose the reverse transfer method (a method of depositing 1 won in an existing bank account and verifying through a transfer statement), which was a very successful choice.

 

3) Implementing the screen with 100% native code: It was too natural in the internet industry to implement the screen natively, but even at that time, a significant number of bank apps were implementing the screen through an assembly framework for development convenience, providing a user experience where the screen did not transition smoothly or touch stuttered. Kakao Bank decided to give a natural feeling that the screen transition or touch is smooth as if it was sliding by implementing 100% with native code. It was too obvious, but it was not obvious at that time.

 

UX differentiation was led by those from Kakao, and employees from Korea Investment & Securities and KB led the differentiation of financial products, each demonstrating their expertise.

 

 

Let's distribute 3 million won to the entire nation: Loss Leader Strategy

 

"I heard Kakao is distributing 3 million won to each citizen."

 

As the preliminary business plan spread to the banking sector, there were talks among bankers that "Kakao is misinterpreting banking," and "Aren't they trying to do high-interest card loans?"

 

However, Kakao Bank's emergency loan had a purpose. Basically, since it was a guarantee loan from the shareholder, Seoul Guarantee Insurance, it did not financially affect Kakao Bank, but as a new bank, it was able to accumulate its own CSS experience in small amounts. We set the minimum cut-off (review) criteria with Seoul Guarantee Insurance, and expanded coverage to the extent that almost 90% of the entire nation could receive it.

 

As it is now and was in 2017, it was a virtuous product that brought customers using cash service and card loans to the banking sector, and I think it was a product that helped Kakao Bank build its own CSS experience muscle.

 

Isn't providing good products to customers the real innovation?

 

 

The Licensing Process to Concretize Preliminary Authorization

 

If the preliminary approval contained macro plans and innovation, the main approval was a process to finalize and specify the contents that were proposed as draft.

 

Have you heard of CPC(Central Point of Contact)? It refers to an electronic system for reporting to financial authorities, and commercial banks have a duty to report important indicators every day.

 

Understanding unfamiliar work procedures and processes, defining over 300 business reports based on bank account standards, regulations for each task (organizational regulations, internal control and compliance monitoring system regulations, board of directors and governance, consumer protection system, etc.), financial plan, human and physical facility plan, risk management plan, etc. Preparation beyond preliminary approval was required.

 

So, after the system development and main approval preparation after the preliminary approval in November 2015, we applied for the main approval in January 2017.

 

 

Can We Succeed?

After three months of review following the main approval application, we obtained the main approval in April 2017, and conducted a CBT for three months targeting employees of the shareholder company. The reaction was not bad.

 

We were working on marketing strategies for the remaining time and started discussing with shareholders like eBay Korea on how to promote.

 

Around that time, K Bank started operations and gathered about 300,000 members in a month, doing rather well. I remember Daniel(Kakaobank's CEO) passing by and asking, "How many do you think will join us?" to which I replied, "Well... we might be able to gather 2 million within this year?"

 

 

We need to look into the lives of Koreans through a microscope

 

At the time, most internet-only banks worldwide had not properly settled in the market, and cases of meaningful success were 1) Internet banks based on offline points of contact that have distribution/sales channels: Japan's Seven Bank (Japan's top internet bank in the Seven-Eleven group, generating 90% of operating profit from non-interest income based on ATM fees), Germany's BMW Bank, which primarily handles loan business linked to BMW sales, 2) European banks targeting customers with poor access to existing banks due to a high number of immigrants, 3) Banks of platforms such as WeChat and Alipay that have secured a strong customer base (Webank, Mybank, etc.).

 

Kakao Bank fell into the third category with a customer base, but in 2016, KakaoTalk was even more of a messaging app.

 

Korean customers' app usage characteristics (e.g., using KakaoTalk for messaging, Coupang for shopping, Naver for search, etc. according to purpose) were a bit different from the customer base that solves everything in one super app (e.g., Southeast Asia's Grab, China's WeChat), so we were not overly optimistic. On the other hand, UK's Monzo and Atom bank, which rely on payment processing and interest margin as their main sources of income, could not avoid continued deficits.

 

Therefore, considering the success story in Japan, we signed an MOU with Lotte Group ahead of the launch, and for a while, I was dispatched to Lotte's office in Jung-gu, Seoul, to focus on implementing a service that allows easy payment to Kakao Bank accounts in the Lotte distribution network.

 

Looking back now, if we drill down into the characteristics of Koreans' lives and app usage, non-interest income is good, but it's a side menu, and the main dish is the intrinsic value of a bank, the interest margin. I think it's more important to provide good loan products, good deposit products, and meet the essential banking needs and unmet needs of customers.

 

Especially since the lives of Japanese, where cash payment is central, and Koreans were somewhat different, I think those who start new challenges will be able to clearly provide differentiated value if they look into our own lives, although benchmarking is good..

 

 

The Brand Value of Kakao that Had to be Acknowledged

Even though we were using the name Kakao, I thought it wouldn't be easy to acquire customers because, unlike Kakao Pay which is embedded into KakaoTalk App, Kakao Bank is an independent app. Having had experience of struggling to acquire and maintain users at my previous job, Toss(Viva Republica), I think I looked at it from a more realistic perspective.

 

As a separate corporation from Kakao, we had to purchase KakaoTalk emoticons at the same price as other companies due to fair trade principles, and we also paid the same cost to use Kakao Friends characters on check cards. I personally thought that the Kakao Friends design on the check card would definitely hook customers, but I was really skeptical about the emoticons.

 

 

"Would people really go through the trouble of taking a photo of their ID to open an account just to get an emoticon?"

"Wouldn't it be a better way to encourage account opening if we give the 2,000 won per unit that we pay to Kakao to the customers instead?"

 

These were some scientific thoughts I had, but in reality, as customers who opened accounts started to use the limited edition new emoticons, it became an issue and had a great resonance and effect in attracting a lot of users who chat on KakaoTalk, stimulating the desire to try it without having to explain the value of account opening in terms of money. It was a moment when I realized again that what's needed isn't sexy logic, but ultimately, the reaction of customers and the market is everything.

 

 

 

 

 

On July 14, 2017, the confidentiality was lifted and I was able to announce the Kakao Bank Friends Check Card to those around me.

 

I had an intuition about success as it spread tremendously and resonated.

 

So, at 7 a.m. on July 27, Kakao Bank started its first business. The 2 million members we expected to achieve by the end of the year were achieved in just 5 days in terms of membership and in just 13 days in terms of account openers.

 

All the marketing plans we had discussed with shareholders became meaningless and were put on hold.

 

There were so many customers that account opening and loans were not smooth, and even other bank apps became unusable due to problems occurring at external institutions linked such as NICE Information Service(Credit Evaluation Company).

 

By the end of the actual year, we had opened 5 million accounts, received 5.19 trillion won, and issued 4.76 trillion won in loans, becoming the fastest-growing internet bank in the world. K Bank, which launched earlier, only reached 5 million customers in early 2021, three years after Kakao Bank. If you are aiming to enter the internet-only banking market in the future, I think you will be able to attract customers relatively easily if you have secured a customer base in any form, tangible or intangible.

 

 

Shareholder Composition and Capital Increase Capacity

As of early 2018, Kakao Bank had undergone four rounds of capital increase due to rapid growth, and the total paid-in capital reached 1.3 trillion won. We had to increase capital urgently due to unexpected loans just a week after launch. Despite the notification that we had to shoot hundreds of billions of won within a week (...), the nine existing shareholders continued to participate without real rights. Even the government (Ministry of Science and ICT), which seemed to have a complex decision-making process, participated in the capital increase quickly.

 

On the other hand, K Bank had a significant obstacle in decision-making as it had as many as 20 shareholders. Also, KT could not become the largest shareholder, and although it attempted to increase capital several times similar to Kakao Bank, it ultimately failed and even experienced a temporary closure of operations with loans suspended for almost a year. Eventually, BC Card took over as the largest shareholder instead of KT and resumed operations, but the competition with Kakao Bank was already over.

 

For a consortium aiming to enter the internet-only banking market in the future, it seems necessary to refer to the composition of shareholders and the ability to increase capital. The time for capital increase comes very quickly due to stronger loan demand than expected. You need to establish a specific plan that can withstand continuous capital increases for several years, almost to the level of commitment. Looking at the case of K Bank, I think the stability of capital raising determines more than half of whether to grant permission.

 

In the case of Kakao Bank, it was possible to operate stably because Korea Investment Holdings, a financial holding company, continued to endure capital increase with a stake of up to 58% until the amendment of the separation of banking and commerce exception law.

 

I hope you always think about the worst case.

 

On the other hand, just because you have strategic investor (SI) shareholders doesn't mean that collaboration automatically takes place. You need to think about how to collaborate specifically. In the case of Kakao Bank, we parted without any collaboration with Netmarble, Korea Post Office(Ministry of Science and ICT), Yes24, etc.

 

 

Chatbot and Interactive User Interface

 

 

Just 45 days after the opening of Kakao Bank, over 1 million inquiries were received at the customer center. Until the opening of the second customer center, not only the customer service part but also all employees of Kakao Bank were tied to responding to customer inquiries, but they were reaching their limits.

 

The Customer Service Part decided to introduce a chatbot to deal with such VOCs and to solve the immediacy of responses to customer inquiries.

 

Fortunately, Kakao was creating a chatbot integrated with KakaoTalk called 'Kakao I Open Builder,' and I was in charge of the deal with Kakao as a strategy manager. The reason why a deal was needed was because the number of VOCs of Kakao Bank was so overwhelmingly large that there was no comparison in Korea.

 

However, the size of the homework given to me was not easy. It was a negotiation that the other party had no reason to do from their point of view. Those who have experienced many deals will know how vague this kind of thing feels. It's hard to explain what it is, but I had no choice but to push it in all directions, and it became the opportunity to change the business strategy of the main body of Kakao. If you absolutely need a deal, I hope you make it happen by doing things that the other party does not expect.

 

It was a pleasure to be recognized by a colleague.

 

The chatbot team in the Customer Service Part continued to optimize the content provided as answers, improving the quality to the level of natural language conversation beyond a simple dialogue form, and increased the usage proportion of the AI counseling chatbot to 60%, enhancing the immediacy of solving customer problems 24/7 and providing higher quality counseling with the secured available resources.

 

These internal experiences became an opportunity to introduce an interactive interface to Kakao Bank's own product beyond the counseling chatbot. Due to the nature of the internet-only bank without a face-to-face channel, there is a psychological anxiety of customers such as disruption of the loan process, so an HCI (Human Computer Interaction) approach can be effective.

 

 

 

Achieving 10 Million Customers in Just Two Years After Launch

 

Even if we were using the Kakao brand, if the products were not competitive or the services were not usable, we would have definitely been rejected by consumers.

 

From my experience, banking business is definitely a very good business if you manage soundness effectively, but for those preparing a new bank, rather than focusing solely on achieving financial goals, if you repeatedly think about what the real vision you want to achieve through the new bank is, and what you ultimately want to solve, I believe you will be able to achieve both social contribution and good results.

 

Do you need advice? Please contact me jayme@kakao.com