As the previous article “2020 Game Company Transformation Tide” said, “change” has become the theme of this generation of game companies, and Grapeman is also studying various companies in the transition period.
In these companies, Heartbeat is a unique sample. Six months of resignation tribute, unlimited holidays, Slack… Heartbeat has been promoting its own system and culture. However, from February 2017 to now, they have not been able to come up with self-developed products.
According to the financial report, Xingdong’s research and development expenses have been increasing: 318 million in 2019 and 220 million in the first half of 2020 (a year-on-year increase of 68.2%). Calculated according to this growth rate, their research and development costs in the past two years may be several hundred million. Where did all this money go? Are they playing tricks?
So some time ago, I interviewed several producers, project management leaders, and bosses that moved my heart. I also joined many of their Slack and Confluence (team communication and collaboration tools, you can view historical information at will), trying to understand their self. The status quo of research business. What I am most concerned about is, when a company wants to try UE4 and do global high-quality projects, how many barriers do they have to cross?
“In the future, I want to have 300 colleagues with an annual salary of 3 million.”
The biggest impression left me on this trip is that the labor cost of Heartbeat is too high, even so high that it does not hesitate to sacrifice profit margins.
Interviews with Xindong CEO Huang Yimeng and partner Dai Yunjie were arranged on the terrace on the top floor of Xindong A1. The temperature in Shanghai that day was about 10 degrees Celsius, but inside the glass house, a vertical fireplace ignited a long tongue, and there was an unactivated barbecue next to it, all of which made the room look warm.
This area is a heartbeat “high-end talent recruitment zone.” Huang Yimeng had repeatedly eaten here with dozens of bigwigs, “Flicker” and they joined the heartbeat. “It may take 1-2 times to recruit a big boss. After the big boss is hired, I will recruit other people, and I will continue to accompany them to dinner. This way, keep recruiting and keep accompanying…just like an MLM organization.”
Huang Yimeng said that in 2017, they think the difficulty of transferring to UE4 is about the same as that of Cocos transferring to Unity. “We didn’t use Unity when we were doing RO (“Ragnarok RO: Guarding Eternal Love”), but the game looks good.” As a result, in 2018-2019, we were tired of the process, development pipeline and personnel quality. , Several UE4 projects have all reached a deadlock.
The management realized that there is still a huge gap between “want to upgrade” and “can upgrade” and “know how to upgrade”, so they decided to ask a few big players with a 3A background to get the pulse, “spend money for knowledge “.
During the interview, I discovered that there is an implicit criterion for hiring people: The ability of the candidate should exceed the basic requirements of the position. In Huang Yimeng’s words, “recruit CTO to be the technical team leader and CEO to be the project manager.”
Xu Mudian, the producer of “Small Town” (formerly known as “Creation World”), told me that 6-7 people in their team have been producers or chief strategists of well-known companies and have led successful projects. His confidence in recruiting comes from Huang Yimeng’s promise: Despite investing resources, he would rather sacrifice profit margins and improve his self-research capabilities without considering short-term returns.
“Torchlight: Infinite” and “Project A” also received this promise. Hu Shaowu, the head of the former studio, told me: “The salary of the big cows is definitely high, but as long as people are competent, the money is worth it.” Xing Wanli, the head of the latter studio, said that he hopes that every team will There is a director-level leader. As a result, because too many talents have joined in the last six months, the team joked that the project should be called “Director Story”.
Such recruitment standards are very practical. In the first two years of development, “Project A” has been in a quagmire. Xing Wanli and I complained: “They all say that you need to polish more when you are doing games, but when your ability is not enough, it is not polished at all, it is just spinning around… you can’t even type the package and can’t see things. What are you talking about iteration?” After the technical director joined in April 2020, it took only one month for them to have a stable Japanese version.
However, the people recruited by this standard are also very expensive, not to mention that Xingdong is still trying to cancel bonuses and increase everyone’s basic salary. The cost of “Torchlight: Infinite” is expected to exceed 100 million; “Heartbeat Town” has spent almost 50 million in the past 3 years, but it may cost 70 million in the next year; “Project A” now has more than 130 people , The monthly cost is several million, and the total project cost is expected to exceed 200 million.
I asked Huang Yimeng, would it be too expensive to do R&D like this? His answer is very decisive: “The profit margin of the game industry is so high, we just sacrificed a little profit margin in exchange for the improvement of self-research capabilities and longer-term development advantages.”
In 2019, Viagra (President of Mihayou Liu Wei) said that excluding bonuses, there are more than 100 people with an annual salary of more than 1 million Mihayou. But now our goal is even bigger, and we have to brag: In the future, we must have 300 colleagues with an annual salary of 3 million.
“Drive the plane and change the engine”
Huang Yimeng and Dai Yunjie told me that recruiting top talents is not just to solve the problems of specific projects. They also hope that Xingdong can achieve a goal: “change engines while driving.”
The meaning of this sentence is that the engine cannot be replaced immediately, and the aircraft cannot be stopped. In the process of change, all projects must proceed normally.
Xie Tianmin was once the commercial director of Epic. He has met many UE4 teams and participated in the development and leading of many large UE projects. In 2020, he was invited to join Heartbeat and set up a project management studio (PMO) of 6-7 people, dedicated to improving the R&D efficiency of 13 research projects. His assessment of the current situation of his company is quite straightforward:
In domestic game companies, the development process of Heartbeat is actually pretty good, most likely in the top 20% of the echelon. But based on my understanding of Ubisoft, EA and Epic, in a horizontal comparison, the development process of the heartbeat can be said to be chaotic and chaotic…
The first goal of PMO is to sort out the heartbeat tool chain. For a long time, “smooth use” has been the only criterion for selecting tools for various projects. Some teams are using Git, some teams are still using svn, and there are even different teams in the same project using different code version management. Software… and Xie Tianmin told me that they hope that in the last 1-2 years, 1-2 teams will be able to use better tools proficiently, and then through personnel flow, more people will be used to more efficient tools and mode.
The second goal of PMO is to help you clarify the development logic and workflow of high-quality projects. “If the original painting cannot be drawn as requested by the plan, and the engine of the original painting is not done correctly, there will be continuous differences between the front and back links, and the quality of the final product will definitely be difficult to guarantee.”
These sound very basic, but the problem is much more complicated than imagined. There is a discussion in the heartbeat Confluence: “What if you shorten the iteration cycle, adjust the communication process, and simplify Jira?” I think this is a good idea just by looking at the form, but after reading the comment area, I realized that there are a lot of details. . For example, as a result of “temporary functions and temporary resources,” coffee machines with tens of thousands of sides may appear.
Hu Shaowu explained to me that the production pipeline of the 3A project is usually determined before the project is established, but they obviously cannot do this at the beginning of the project: “Some may be confirmed, some are not confirmed, and some are confirmed but the standard is not high enough. , And some need to be determined while doing… And even if the process and specifications are in place, if the quality of the personnel is not in place, the situation will not get better.”
On this point, Xing Wanli also cited an example. In order to reduce the character’s production cycle from 4 months to 3 months, they hope to add a middle mold between the white mold and the high mold. In addition to carving the high mold in the model material pipeline, it also allows the combat action pipeline to use the middle mold in parallel. Development. But can team members truly understand the upstream and downstream needs and deliver qualified things? This is the biggest risk.
In order to achieve “changing engines while driving the plane”, in addition to letting Daniel help “build wheels one by one,” the members of the existing team must also grow rapidly and even face elimination. For example, in the 80-90 team of “Torchlight: Infinite”, there are only a dozen members from the previous SLG web game team.
Xing Wanli described that the “Project A” aircraft not only changed the engine, but also changed the chairs, passengers and even the captain. In the last year, they have changed their main strategy, technical director, content director, and hired an executive producer to be responsible for project management, and he has to update his cognition wildly:
I went from a Japanese producer who manages everything to a European and American producer who pays attention to quantification, splitting and decentralization. Every 3 months, I feel like a fool before.
“Deliver first, then talk about others”
Xie Tianmin told me that before joining Heartbeat, he also found that this company had a terrible problem: “Everyone has too much time, and there is no bottom line.”
The project of “Heartbeat Town” was established in 2016. As a result, after Xu Mudian became the producer in 2020, he decided to overthrow the program, planning and art; “Project A” was established in 2017, and the core team alone formed nearly 1 Years, until now the game has not entered the mass production stage. In view of this, “Torchlight: Infinite”, which is planned to be developed in 3 years, may already be the most efficient large-scale project.
The strategy of the heart has explained this. Huang Yimeng said in a letter to shareholders before that, taking into account the help of exclusive content to TapTap, Xindong can develop products that have low ROI in the early stage or are difficult and require long-term polishing.
So I asked Huang Yimeng, so is TapTap a shield for everything? Long trial and error, frequent delays, and even the results after the project went live are not successful enough… Is there an excuse for everything?
Huang Yimeng admitted that he was really too tolerant of projects before: “I know every project, and I feel that everyone works hard and can fully understand why things are not done well.” But then they found that it is normal for them to do well in 3D games. If this continues, you may not be able to do anything. So from 2019-2020, Xindong began to pay close attention to project management.
Now all the projects of the heartbeat will first determine the final online time, and then push back each project node to ensure that everyone can align the goals. “After the time is clear, everyone will understand what help they need now, and it will be easier for the company to concentrate resources to support them.”
In my opinion, these are common sense in project management for large companies. But for a company that has just thought about these things, it may not be easy to completely change.
Perhaps to show determination, on Slack, Dai Yunjie described himself as “Project Manager” and set “Deliver first, then talk about others” as a signature. He explained to me: “Dash sings the red face, I have to sing the white face, to hold the company’s bottom line.”
Huang Yimeng said that “holding the bottom line” does not mean that investment must be strictly controlled or that projects that do not meet the standards should be cut off. But in order to reduce the risk of the rule of man, they hope that most projects can go online as soon as possible, and iterate while doing: “We don’t need to make games for the company’s evaluation. At that time, take a look at TapTap and look at player reviews to see if this project can be it is done.”
From a historical point of view, Xin Dong has had a similar experience: “Sausage Party” did not make money when it first launched, and after several editions, it remained only more than 20%. But considering that it can attract users to TapTap, my heart does not stop investing in it. Now after 3-4 years of operation, its monthly turnover has reached tens of millions.
However, this system also puts forward higher requirements for communication efficiency. In October last year, Dai Yunjie hoped that the game “Flash Party” could be tested online next week, but Huang Yimeng thought that the original December would be more secure. The two people started arguing on the Slack public channel… and finally forced the core members of the team. After expressing their thoughts, everyone quickly set a more rational time node. I guess if there is no Slack, this kind of thing will face a long wrangling.
“Aim at the blue ocean and work hard.”
In the interview, I also asked Huang Yimeng an unkind question: Since “changing engines while driving a plane” is so important, have they ever thought about simply replacing some producers, even Hu Shaowu and Xing Wanli?
Huang Yimeng’s answer is very rational: in the weekly chat, he can feel the speed of the production people’s cognitive improvement, and he can always learn new things, which is why trust is very important. And although the 3A boss has stronger realization ability, the existing producers are more familiar with the heartbeat product ideas.
Lao Xing (Xing Wanli) said something that makes sense: We decided very early that the producer should be rich, but after spending so much money, we found that we didn’t develop it, and thought that something might be wrong. Finally, after studying for a long time, I found that there is still only fertility…it is such a process.
The last time I interviewed Xing Wanli was in 2016. At that time, as a producer, he made an RO that was known to be “different from the sole of the foot to the top of the head” from other domestic MMOs: it had no sign-in, no main mission, no PK, and added a lot of collection and friendly social elements (such as players You can walk together holding hands), and abandon the paid design of routines, hoping to “make rewards for players with craftsmanship.”
Half a year after it went online, due to too much innovation, RO’s monthly turnover fell from 100 million to more than 10 million; but then they managed to resist the pressure and made a series of modifications, and RO also topped the best-selling list in many countries in Southeast Asia. After this, is the story of “Project A”.
Hu Shaowu is the oldest group of employees in the heart game business, and has participated in the page games “Heaven and Earth Heroes” and “Opening the World”. His most well-known product may be “Sweeping a Thousand Arms”-as the first self-developed IP mobile game, the team hopes to make it as long as possible. As a result, when it was launched in 2015, it was rated as B-level by the channel because of insufficient krypton gold in the early stage.
From the experiences of Xing Wanli and Hu Shaowu, you can understand the so-called “heartbeat product ideas.” Hu Shaowu said that they have several principles for establishing the project: first, the art style and product quality meet the standards of the global market; second, make games with a long life cycle (Huang Yimeng hopes that most projects can operate for more than 10 years); third , The project must be able to enhance the team’s R&D strength.
Several key projects of Xintong also meet these standards. “Torchlight: Infinite” is the same as the original, it is a dark-like game. But they hope to be more pure and creative than other Diablo-like mobile games, and the system framework is less constrained by revenue KPIs.
“Project A” is a two-dimensional style MMO, RPG is its focus. Xing Wanli thinks that traditional MMO is too complicated. He introduced its core fun like this: “I want to enjoy the story, I want to know these characters, and I want to change the world.”
“Heartbeat Town” is a life simulation game. According to the gameplay cycle shown to me by the producer Xu Mudian, it hopes to allow players to experience the fun of the daily gameplay of “Dongsen”, obtain new content through active or paid, and gradually form social relationships, and ultimately maintain the UGC atmosphere. “This category will not be an instant hit, but as long as users can raise it, it will have great potential.”
“Flash Party” is like a “Super Smash Bros.” that has a more refreshing rhythm and integrates the peripheral growth mechanism of “Wild Brawls”. The game is still in the stage of limited testing of the TapTap Bonfire Project. Producer Liu Jiali told me that their goal for the next version is to make it fun for ordinary players to play.
I experienced “Flash Party” for a while, and also watched the current versions of “Heartbeat Town” and “Project A”. They are a long way from the goals described by the producer and I, but at least they all look like that, and they are not routine.
There is just one thing that I am worried about: the core gameplay of “Small Town” and “Flash Party” are all related to Nintendo’s products. Huang Yimeng once compared the relationship between TapTap and Xindong’s self-developed products to the relationship between Switch and Nintendo’s first-party works. So now, will this kind of “Cosplay” cross the line?
But Xu Mudian said that this will not happen. He said that “Heartbeat Town” will make differences in gameplay similar to “Dongsen” (such as fishing) and add new gameplay methods such as 3D building, driving, gardening, and cooking. Their social mechanisms are also completely different;
Liu Jiali said that he never shy away from the core gameplay of “Flash Party” from “Smash Bros.”, but they will design differentiated heroes, create new gameplay mechanisms, and form their own style, “just like between LOL and DOTA. “Relationship.” Huang Yimeng also made a promise: “Although we have a low starting point, we have to learn from borrowing. But we will continue to iterate to make our own characteristics until we are recognized and praised by players.”
Make your own style-this is certainly not easy. Regardless of the level of industrialization, I am also a little worried about the design ability of my heart. But Huang Yimeng and Dai Yunjie told me that only by doing differentiated categories can they have the opportunity to overtake in corners:
Fighting MMO with Tencent NetEase is hopeless. We can’t cut the cake of the Red Sea market, but we have to find a track of our own, aiming at the Blue Sea, and devoting ourselves. Only by using these categories to complete the R&D and upgrading, we will not lose when we really need PK in 5-10 years.
Increasing the density of talents, driving the airplane to change engines, clarifying time points, insisting on differentiation… After more than 3 years of tossing and investing hundreds of millions of dollars, I finally clarified the rhythm of several key products.
“Flash Party” is going to start unlimited testing this summer, “Torch Light: Infinite” is expected to be launched at the end of the year, “Heartbeat Town” is expected to start a round of non-deleted tests at the end of the year, and “Project A” is expected in November Start the first round of testing…All the bullets will have to fly for a while, but Huang Yimeng feels that his heart-beating self-study is basically on the right path.
Huang Yimeng admitted that compared with the top R&D companies in China, there is still a gap of at least 3-5 times. They may still have 10 years to go before their ideal R&D strength. He said that now there are only 10 colleagues in the company with an annual salary of 3 million, which is far from the goal of 300. “It is possible to recruit all the top gamers in Shanghai.”
But things seem to be changing. In order to persuade the first and second place bosses with 3A backgrounds to join, Huang Yimeng spent half a year back and forth. But now he has the final say. If the candidate is willing to go to the glass house on the top floor of the company and chat with himself, the on-boarding rate is at least 50%. He feels that the “CEO selling art” approach has been effective because of his frequent appearances, but he also hopes that the industry will not misunderstand that the product is what he cares most.
When sorting out the material for this article, I saw a circle of friends of “Dream Tower Defense” producer Cai Jianwei: In 2004, the relevant department announced to the public that China’s lunar exploration program would go through three stages: “circle”, “fall” and “return”. , The whole process takes a total of 16 years. His comment was: “What does it mean to be a design for three to five years?”
To some extent, doing R&D also requires meticulous planning, long preparation and time, and the belief that goals will eventually be achieved. At this point, all companies including Xindong are no different.
Towards the end of the interview, I asked Huang Yimeng and Dai Yunjie how they felt about the recent years of R&D and upgrading. Huang Yimeng said that maybe 10 years later, he will only be able to express emotion; Dai Yunjie stared at me with his big eyes: “Since I want to go to the moon, I will go to the moon first. It is not yet time.”
Source: Game Grapes