Appsynth and open-finance enabler Brankas recently hosted an intimate roundtable dinner for top executives from banking and financial services. Below are a few highlights from our Q&A with guest of honor Dr Dennis Khoo, Managing Partner at allDigitalfuture in Singapore and former MD of TMRW Digital Bank. Dr Khoo is author of Driving Digital Transformation: Lessons from building the First ASEAN Digital Bank, which is available from leading bookstores like Kinokuniya or Amazon.
Background
Dennis is an accomplished digital business leader and speaker in innovation and leadership. Armed with dual experience in Information Technology and Banking, he has a unique advantage in developing and executing strategies for the transformation of businesses in a digital world.
Dennis was a senior banker who had previously run billion dollar businesses as head of consumer bank for Standard Chartered Bank and UOB in Singapore. He was responsible for many of the innovations in the consumer banking industry in Singapore, such as e$aver, XtraSaver (Winner of Asian Banker Best Deposit Linked account in 2007), Retail Bonds, Step-up time deposit, Pay-Any-Card (Winner of Asian Banker Best Payment Product in 2011), 15% Dining Cashback everywhere, 2-day Mortgage Service Guarantee, UOB Income Builder, One Account and UOB Mighty App.
A digital bank pioneer in ASEAN, Dennis was group head of TMRW Digital Group (UOB’s Millennial digital bank), where he was responsible for the strategy, growth and delivery of the TMRW Digital Bank. The first TMRW Digital Bank went live in Thailand in early 2019, and the second in Indonesia in 2020. TMRW won Global Finance Most Innovative Digital Bank in Asia Pacific in 2019. Dennis also led the joint venture initiatives into Fintechs like Avatec.ai and VUI Pte Ltd as part of the ecosystem to support the growth of TMRW.
Could you share one thing that you’ve achieved in your career that you feel particularly proud of?
The most challenging assignment in my career has been building TMRW. We built it from scratch. The majority of transformation projects in large incumbent companies actually fail, and with TMRW we just managed to push it over the line.
It was an incredibly difficult project, more so because the founding teams were originally Singaporeans. We came here and our Thai colleagues would ask us: why are you here, if you can’t speak Thai, how would you understand what it takes to build a bank for Thai Millennials? It was a very challenging assignment.
When we launched it here in Thailand, I think it was early 2019, we had an NPS score of negative six. And one of the reasons why we were negative six was that we had to determine a cutoff point and launch. The cutoff point was such that most of the menus were iconic, meaning menus all designed as icons, and that was kind of in line with what Thailand already had. But I decided not to do it for the drop-down menus because I thought we could do the icons later, and we were used to using drop-downs back home. However in banking apps all the drop downs in Thailand are icons. And so this was one of maybe 13 to 14 problems we had to solve, in five or six months after launching. And when we solved them, the NPS went from negative six to about 40.
In the end, the satisfaction of seeing it push through was very satisfying. Thailand is a pretty difficult market, but the project still gained an NPS score in the 40s in Thailand, and in the 50s and 60s in Indonesia.
Digital has definitely changed the way that we work. How do you nurture innovation in your teams and within the organization?
I specialize in innovation in complex incumbent companies. From an innovation standpoint, in a bank like UOB, for example, it is a difficult environment. When I joined they had been very successful at acquiring the right licenses throughout South-East Asia, but haven’t innovated sufficiently in digital terms.
There are a few things that are very important in driving innovation at large incumbents, but the most important is to break hierarchy. For example, when I joined the bank, someone told me that executive directors can only talk to other executive directors; they can’t talk to managing directors. I thought that’s silly, we should change that.
What should happen is that even the most junior person in a meeting should be told that they can “why”? And the most senior person must answer the question or make sure it’s answered. This immediately removes hierarchy or rank. If the “general” makes a suggestion and the “private” has a better suggestion, the “private” wins. This approach means that we immediately neutralise everybody’s rank. This means that the best ideas get done, not the voice of the most powerful person. This is very important in incumbent companies like banks. Such companies are very different from startups because they have an established position in the market. So you should learn everything from startups but copy nothing. Because when you copy wholesale, it mostly doesn’t work.
What new technology are you most excited about, in terms of how it’s going to impact the industry?
One of the propositions that I have worked on is providing great service. First of all, provide the best basics so that you don’t need customer service, but in the event that you do need customer service, serve the customer very well. The design in TMRW by UOB was such that I made the decision to manage all customer service through a chatbot first. I made this decision based on the principle that if you don’t implement this at the beginning, it becomes very difficult to wean customers off voice calls later down the line.
To mitigate the initial limited capability of the chatbot, we escalated enquiries to a human agent very quickly. So if the chatbot couldn’t understand the query in three tries, it would be escalated to a human. Each month, I would come to Thailand and look at all the chatbot queries which had been escalated to the agent. I would then ask the team to train the chatbot so that it was able to answer those queries in the future. It took us 18 months before we got to the stage where the Chatbot, called TIA (TMRW Intelligent Assistant) understood 95% of Thai queries, which is a very difficult language to master.
For example, Thai has no full stop, so you don’t know where a sentence ends and where the next sentence begins. We had to rewrite the natural language processing engine to handle Thai, which was an incredibly difficult task. It required a lot of groundwork. I would go down and ask, “How are you teaching it?” However, one of the problems in banks is that the seniors usually don’t get involved; they delegate the task. And, if you delegate a task without your involvement, it suggests that it is not overly important. The act of showing that the seniors are involved in the task is crucial in demonstrating its importance to the team, and especially so as the innovation in competitive banking environments is experiential more than breakthrough.
Today, with Generative AI (Gen AI), you can achieve this very quickly. But back in 2018, we didn’t have Gen AI, so we had to train step-by-step and code it into the system. The GenAI breakthrough will not only benefit the banking industry but also many support functions in the global services industry.
There are other advantages as well. For example, one of the greatest pet peeves in banking is calling up and having to repeat everything again when you call back later. In our design, the agent can see the last chat, so they can say, “Oh, I see. You must be calling about the query from last week.” This removes one of the most systemic problems in banking, which is repeating yourself.
The moral of the story here is that you cannot pivot to a digital bank from an analog bank. You must design ahead of time, and the design must be of a very good standard. You cannot pivot your way to great design and experience.
In regards to Open Finance, do you have any experiences where you have either opened up an API for third parties to integrate and innovate with, or partnered with third parties to bring their services into your core product?
I’ll answer this question from two perspectives. Firstly, instead of building your own software, why not use somebody else’s software to do the same thing?
Generally speaking, banks are not very skillful at writing new software. We do projects, and we do one project after another, and usually it’s not well documented. The next guy who comes in may understand how the code is written, or why it was written that way.
Given this context, when I was embarking on creating TMRW, I convinced my technology head not to write the software as far as possible. Their job is to focus on integration. For example, the chat module and the insight engine: how do you take transactions, consume the transactions and then produce insights?
Here’s an example of where using third-party software can be more efficient in solving a problem than writing it yourself. Let’s say you have two bank accounts: one that receives your salary and one that doesn’t. The account in which the salary isn’t deposited occasionally receives a charge when a debit fails because you forget to transfer money in. The bank knows that this is a regular occurring direct debit but they charge you for the transaction failing because it is your responsibility to ensure adequate funds.
However, this situation is easily solvable by notifying the customer prior to the event that their direct debit will fail and result in a charge. “Click here to top up your account”. We use third-party software like Personetics to do just that. That’s good service. But banks generally don’t do it because the paradigm is that the customer is responsible to keep the account funded.
On the other hand, there was a bank in Europe for whom we did a lot of benchmarking. When we started the design and met with them they said that they were considering whether to write or buy software. Three years later, they were still thinking about it. So, in my opinion, I don’t think there’s any point building your own software in B2B settings.
Banks have always used third-party software. In the past, a lot of banks used Silver Lake as their core system, but it’s often customized significantly, maybe up to 70%. So even though the label says Silver Lake, it is basically your own software. When you upgrade, you pay a lot to go from a lower version to the latest version, say version 20, and when you want to go from version 20 to 23, you pay another $10 million because of the high level of customization. So banks have always used fintech software. But customization is also very expensive. That’s something to think about.
Of the people using the bank’s API, which is the other part of the equation, nobody wants to do that in retail. In wholesale, it makes sense because big companies like Walmart want to use APIs to streamline their processing systems and resolve a lot of company problems, such as reconciliation. So I think API usage comes down to the individual use case. The use case is very clear in wholesale, but there is very little use case in retail for a company to use the bank’s API.
Lastly, can you tell us about something that you are working on now that you’re very excited about?
I’m not in banking anymore. However, there probably wouldn’t be anything exciting to say because frankly, the key to great banking is attention to detail. And as I often ask my clients, what’s the difference between a Mac and a PC? All the components are the same: keyboard, processor, memory, screen, etc.
The difference is that Apple is insane about attention to detail, and the problem with most banks is that they just don’t care enough about detail. For example, many bank seniors attend strategy meetings, but who goes to the process meeting? Nobody. They most likely send the junior staff there as a punishment! What I used to do was insist that the most senior person attend the process meeting.
Banking is about doing the simple things well. With this in mind, I have put together a playbook that is holistic in nature. It covers all business aspects, including capabilities, processes, leadership, customer service and more. For all complex and common companies, if you’re conducting a major transformation without a holistic approach, the probability of success is very low. So that’s what I’m excited about right now, trying to be more pervasive and spread this holistic approach far and wide. You can find the playbook at any amazon bookstore globally: A holistic approach to transformation and innovation in a complex world
We’d like to express our gratitude to Dr. Dennis Khoo for flying in from overseas to join us for the event and sharing his insights with everyone. Check out the Q&A with our other guest of honour Htin Hlaing, CTO of Wave Money, Myanmar’s leading digital wallet provider, here.
