Meet CTW’s Chief Data Officer, Liao Wei

Path to CTW?

This is our series on the team working at CTW.

Today, we’ve got our Chief Data Officer, Liao Wei. The work he and his team have done has transformed CTW from a successful game company to the No. 1 IP game platform company in Japan. It’s not just our marketing that’s become a high-performing data-driven machine. Our entire culture is transforming into one where data and AI support the autonomy, creativity, and motivation of our entire team!

You won’t find a more passionate leader, a more driven mind, than our Mr Liao.

What convinced you to join CTW?

Okay, that's actually a pretty funny story. So our Engineering Manager, maybe the third person to join the Tech Team at CTW, he basically approached me and we chatted about his new company. He told me, “I have a data project I’m working on. They want me to build an AI recommendation system.” He began describing the project and the challenges they were facing until I really expressed interest and began asking questions. Then, it became kind of a consultation between me and him about what could be done and what steps were needed. He mentioned that they were very interested in expanding their business and maybe even building another tech-focused company. So I met with him and he told me the truth. They wanted me to join them. They had an ambitious new direction they were planning for the company, and they wanted my help.

He’d lured me in with tantalizing bits and pieces until I was hooked! But it was all for the best in the end.

Some of the things that stood out to me from back then: They showed me Vivid Army, our biggest hit, and I decided to have a look, even though I basically didn't play games. So when I tried it, the next time I looked up from the screen was almost two hours later. And it was on a browser. I thought that was quite something, especially given I didn’t care much for games. Then, he said that, actually, the company didn't really do games until recently.?

I also remember being shocked by the overview of the business model. After working for Rakuten for 8 years, and another startup funded by Softbank, I was shocked by how much money a company could make doing just browser games.

Their vision combined with my interest in the project sealed the deal.

What was this project that enticed you?

I joined in 2020.?

There were about 40 people total, four of us on the Tech Team.?

So when I first started, actually, we weren’t working on the marketing at all.

But I was interested in it and I started talking with the marketing VP about how they were doing their advertisements. This was a little bit different from what I’d been doing at Rakuten, and I wanted to understand exactly how their advertising worked.

I asked a lot of questions regarding their spending decisions, how they measured performance, how they decided channels, that kind of thing. They told me that there was no set strategy for where or how the money was being spent. It was just huge volumes across multiple channels, basically covering all their bases.

I didn't actually talk directly with the CEO at this point. But the other leaders told me that the boss wanted no expenses spared in the quest for reaching new players. That was it. They didn't have any kind of feedback loops in place.

Their instinct was not to skimp on marketing but it wasn’t data-driven. They didn't know how it was going, they couldn’t measure cost performance, and they had no metrics at all. Return on investment was measured by revenue only.

So, in this first phase, I was just a question machine. In the most straightforward way, I wanted to understand how they were handling their information, what the decision-making process was, and what impediments were there. It was fundamental for me to prepare for what we planned to do.

Then, over a few days, I explained to them what they could do with their incoming data.

I went over the basics of data science. From data collection, pre-processing, organizing, and visualization, through the whole cycle of data flow — I made sure that they understood the entire process back to front. Then, I told them how we could use that data to calculate the likelihood of acquisition, to put a monetary number on individual user acquisition cost, and to also accurately estimate the lifetime value of each spender.

I explained how we would capture, sort, and define the data and what metrics could be used to calculate and predict performance.??

So once that was clear, we got the go-ahead, and began building the data warehouse. One and a half months later, we had the first AI recommendation system for Vivid Army, our biggest game. Within 3 months, the money saved on our pilot test was more than enough to cover the entire cost of our team for the year.

That was our proof of concept. Since then, the model has continued to learn over 2 years and it now visualizes, optimizes, and controls our entire marketing expenditure and return over hundreds of channels globally.

We get better and better performance on our advertising while continuing to reduce the cost of new user acquisition. In fact, our user acquisition is more cost-effective than any of the Big Tech companies, by a significant margin.

How has CTW changed as a result?

So, initially, we took time to win the trust of all our stakeholders. We started out by showing we could boost revenue, which — of course — impressed everyone. We proved that, from the data side, we could ensure that investment and return were controllable and that the company would have the highest chances of being profitable.

If we put in X amount of money into the ads, we would very probably get Y returns.

That's kind of the basic line we'd drawn.

Once that core was in place, and our leadership was convinced, it basically changed the business model of the company. Not only the business model but the culture too. Marketing was now a highly data-driven activity. People who had been spending huge amounts of money, stressing every day about the day that their channels started to fail them, now had a job they enjoyed. It was a pleasure to see them begin to experiment and optimize their understanding of the model and the new sources of revenue it could generate.

In the first stage, when I joined, the money would flow out and maybe make a decent return. Then, we created a feedback loop that got more and more accurate, first biweekly, then weekly, then daily, until it was basically real-time. Our marketing team had access to all the performance information of their ad spending, which they could see and understand and change their approach.

So that data-driven AI approach is now shifting to our games. If we’re cloud native, we can completely scale our game business. And again, we’re seeing the same kind of success.

We’ve taken our games from one market, with a few games, on a few channels, to nearly 15 games, across hundreds of channels, in multiple languages, with the same stability and performance.

We’re building a system that eliminates busywork, automates and scales, and allows people to use their skills and creativity to succeed rather than stress in the dark.

What’s the next step for the Tech Team?

So to explain this, let’s take a look at our banner generation tool. It can produce thousands of banners, all AI-optimized for the highest market performance, using photoshop elements made by our creative team. It boosts our creative productivity as if we had a creative team of hundreds of people.?

But we’ve been marketing focused until now. We want to bring those same kinds of benefits to everyone on the team, across every department.

Already, we’re working on translation tools that will dramatically cut down localization time. We’re aiming to get our games to all our markets as quickly as possible, increasing revenue and bringing better experiences to more players around the world.

We’re looking to give our game devs the ability to access the cloud platform in real-time and make updates to our games directly, with no lag, and greatly reduced bugs and errors.

The business side, the game planners, and the devs will understand how each of our in-game events is performing, where and why issues are arising, while also making it completely scalable and stable with every game we release in the future.

So it's not just about automating all our processes to succeed but also shortening the amount of time that our talent needs to spend doing things that get in the way of their work.

And this is the best part about working here.

We’re shaping and changing every aspect of the business.

If you go to any of the Big 4, you’ll be plugging yourself into their system. You’ll be part of the machine. Here, you’ll build the system. If you can do something, you get to decide how it’s going to work, how it’s going to be built. You can pitch new ideas on technologies that you think will improve our business and have the autonomy to create them yourself from scratch.

And with the speed that we’re growing, we need people who have experience taking companies global. Experience with technologies is good but we need leaders who can connect us with how customers and users think around the world, and bridge the gaps in our business to become a truly globalized world leader.

_________

For more information about our available positions, have a look below.?

We’re actively looking for talent in the following roles.

Front-end Engineers

https://lnkd.in/dAbMnXm7

Backend Engineers: Golang + Python

https://lnkd.in/gg-hGe3v

https://lnkd.in/g-v_Yfwa

Dev Ops Engineers

https://lnkd.in/grdpA4TQ

NLP Research Scientist

https://lnkd.in/ghMCPxVg

Data Engineer — Scala

https://lnkd.in/ghxFu4G8

Product Manager

https://lnkd.in/gC_fk9WB