In the past two years, the mini game has exploded, and many people feel that it has taken advantage of the trend, and for a while, it seems that countless teams have poured in, and the number of developers has exceeded 400,000.
But have you ever thought about what the mini-game depends on to keep going?
On July 16, at the 2024 WeChat Mini Game Developer Conference, many people talked about their thoughts in combination with the product. Among them, Zhang Jun, the head of R&D in Guangzhou Huanyu Kyushu, shared quite impressively. This company has teams such as Aipulei and Guangzhou Mia, and popular mini-games such as "Potato Hero", "Tower Defense Elf" and "Girl and the Sea" are all from this system.
The product they shared this time is called "Chicken Fleet Attack". It's a tower defense game, but it's rich in content, with a mix of cards + Roguelike + lawn mowing. Coupled with the chicken's Internet hot stalk elements, various uses of stalks, its recognition and playability are not low, and it is also a rare horizontal screen product in small games.
Shortly after its launch last year, "Chicken Fleet Attack" achieved good results, reaching the top 5 in the best-selling list of mini games. Since then, its results have been relatively stable, and it can return to the top 20 with the version update.
Source: Gravitational Engine
According to reports, they adopted a development method similar to Supercell, and the demo period was made by a team of 5 people, and gradually added people after verification. You may never have imagined that when adjusting the gameplay experience, the R&D team took into account the time when everyone's legs went to the toilet - squatting for 15 minutes and the legs were numb, and the length of a single game was too long; 9 minutes of leg numbness, this time is just right.
For the mini game market, Zhang Jun also mentioned that "moves" may reduce the CPI for products in the short term, but it is difficult to form barriers, and others can always learn them; If the mini game wants to develop for a long time, it may be more important to practice the "internal strength".
Here's what was shared (some of which have been adjusted for ease of reading):
Good afternoon, everyone, and I'm very happy to be on this stage again.
Why again? Because the "Potato Hero" shared last time was also done by this team, their core members have been running in for nine years, and they have been trying to do it on the subdivision track of tower defense.
The size of our team is not very large, but we make a lot of products. When we do the demo, there are basically no more than 5 people, and after the gameplay is verified, it will gradually increase to more than 10 people, which is a low-cost, rapid iteration, and rapid verification idea.
For so many years, we have been exploring the tower defense track, hoping to make tower defense products more fun and longer. This is mainly due to the fact that this category has several great advantages:
First, the audience is broad. In 2014 and 2015, when we did "National Democracy", we did research on the tower defense market, such as "Plants vs. Zombies" and "Defending Carrots", which have been downloaded more than 100 million;
Second, the threshold for tower defense is very low, and the gameplay itself is very intuitive and does not require too many operations;
Third, tower defense itself is a way to play with more elements, if you turn some defense towers into heroes, you can convert them into cards and combine them with cards for commercialization. This is something that is worth trying again.
The gameplay idea for "Chicken Fleet Attack" first came from a product I played four or five years ago, "Dungeon Maker".
I played several all-nighters, probably more than 100 hours. For the first time, I discovered the benefits of the meat pigeon gameplay, which can make the single-game experience of a level game very interesting and a very strong sense of growth.
Secondly, the meat pigeon has added a lot of randomness to the gameplay, and this randomness solves the pain point of the level when we originally did "The Whole People" - there is no way to set the difficulty of the level. Because no matter what I do, some people will say that your level is too difficult, or it is too easy to play.
But meat pigeon products are different, because there is a strong random factor, you can play but you can try again. Even if I hadn't played, I would probably blame bad luck and not feel like there was something wrong with my strategy. Therefore, the addition of the luck factor has greatly increased the game duration of the single-game gameplay, which is the same as the egg upgrade, even if I lose this time, I feel that I can play better next time. This is one of the reasons why we choose meat pigeons.
So why did it only last year? Because we haven't been able to find a good commercialization plan before. With the overseas success of Haibi's Legend of the Bow and Arrow and Bullet Shell Squad, we got some inspiration. It was just at the beginning of last year that the team had a schedule, and the project was officially launched.
After deciding on the gameplay, I'll talk about how our art style and subject matter came about. This is a big weakness of our team, because the main planning is based on values and gameplay, and they are not sensitive to this part.
In the past, several of our products had a relatively large disadvantage in terms of purchase volume, and they were more expensive than those of the same type, so we also put some effort into studying some competing products. In the end, we got a formula: "magical image + medium and heavy theme + relaxed and witty" painting style.
The magic image is to attract users and reduce the cost of customer acquisition; Medium-to-heavy themes are easier to commercialize, such as magic, Three Kingdoms or apocalyptic themes; The relaxed and witty art style is to let the user experience the game not be too heavy and be in a more relaxed atmosphere.
In this area, we have been studying art for a while, and we don't have a consensus on the word demonic nature - what is demonic nature?
Later, we briefly did some analysis, and felt that demonic nature has several characteristics: first, it is eccentric and anti-common sense; The second is that it looks funny, and can infect you with a certain emotion, so that you have a strong desire to share. Based on this criterion, we have some classic cases, such as "The Strongest Snail" and "The King of Salted Fish", which are some games that inspire us.
After defining the criteria of magic, how to find the subject matter and painting style? "Memes" basically fit these characteristics – fun, shareable, and able to express your emotions. For example, before "The King of Salted Fish", salted fish emojis were widely circulated; I'm a big fan of One Piece, and I also find the phone bug in it very interesting.
So we looked for what kind of subject matter and image we wanted to use. This chicken is a subject that we quickly settled. Because the popularity of chicks has been quite high in the past two years, many second creations and Internet memes are related to chickens, and it is a special image of life.
After that, we used the style that was better measured before, and compared and tested the guide of old users, and found that the image of the chicken had a decrease of about half of the CPI, so the image of the chicken was quickly determined.
After drawing a few of the first images, we picked two more contrasts. The first one is the Q version, which is quite interesting, but it's obviously not as interesting as the second one, because the second one is more in line with the magic we want. Its image comes from the screaming chicken that was popular in the first few years, looking stupid and silly, you really want to rub it twice, like Mr. Bean's silly feeling.
This theme and style of painting, in the stage of formal promotion and purchase, did reduce the cost of buying a lot for us, which can be regarded as a relatively successful attempt.
This stage took us a short time, and it was basically determined within a week. After I started polishing the demo, I completed the first demo in three months, and conducted a round of testing in the company, and the test results were very good, and everyone's feedback was very interesting.
From the gameplay test to the actual launch, it took another three months. In the past three months, we have made changes to the gameplay in the following three directions, and now I think this change is still necessary:
First, the game's initial single-game combat duration starts at 15 minutes, which is a reference to other apps or Steam games. We have made adjustments in this regard according to the usage scenarios and user attributes of Mini Game users.
Imagine that you want to play a game at work, go to the toilet and squat and open a handful, if you squat for too long, your legs will be numb. So after several adjustments, we found that our legs would not go numb after squatting for 9 minutes, so we adjusted the time to about 9 minutes. Of course, it is still necessary to ensure the existence of the fun of meat pigeons.
Second, we have greatly simplified the elements of the meat pigeon gameplay. The meat pigeon gameplay has also existed for a long time, and everyone is getting richer and richer, and there are many random factors in it, such as random buffs, menus, equipment, and events...... It's a lot, and it's very disruptive to Minigame users, so we've made some simplifications.
In the original version, there were only two options: choosing heroes and choosing weapons for heroes, which was actually enough, and it was completely able to express the fun of meat pigeons.
Third, better carding. The original intention of this game was to use our previous experience in making cards to give full play to the advantages of tower defense and meat pigeons. So when we started doing it, we didn't limit classes, and we could use multiple classes on the same battlefield.
After doing this, we found a problem: first, the balance of the meat pigeon is not very good, and the random weapons are not easy to follow. So we changed the version, made it a fixed class, and adjusted the balance and boss mechanics so that each class has something to play. In this way, users' demands for professions and cards will be stronger, and it will be more convenient for us to commercialize based on them later.
Here are some of the most important decisions to make before going live. This product, it may be because we did the first three steps correctly, so when we opened the paid test for the first time when we went online, the data was very good, and the weekly paid retention could be 70%, and the volume was quickly increased. It took less than 7 days for us to go from the first day of testing to more than 5 million. After the ramp-up, the data is relatively stable.
When I shared "Potato Hero" before, I shared how we took 6 months to go from bad data to the goal. In the project "Chicken Fleet Attack", we didn't have this pain at all, and basically felt OK on the line, which also proves that our R&D team has been growing.
After going live, we have a few clear experiences:
First, hybrid monetization. Our company's first few products are also mixed monetization, and we have an experience that it is best to think clearly about what to focus on, not both.
If you mainly do in-app purchases, try not to affect the experience of paying users, and do not invest too high-value props and values in advertising; If it's advertising-oriented, you can't raise the payment too high, and it's better to make some supplements.
"Chicken Fleet Attack" made it clear at the beginning, these two monetization methods cannot be both, the early stage of advertising is very fierce, but the sustainability ability in the middle and late stages is not particularly strong. So we decided to focus on internal purchases, supplemented by advertising, and also added advertising points to the beginning of the battle, for example, you can resurrect a hero by watching advertisements, but the back is removed, and the final effect is quite good.
Second, user stratification. In the process of operation, we updated some long-term gameplay and cultivation points, such as equipment, companions, and insect trapping, and the team also made some iterations.
We are not bad in long-term operation, which is the result of the joint efforts of R&D and distribution partners. Including the "Tower Defense Elf" we released before, we have a lot of experience in IP linkage, so we also found some linkage for "Chicken Fleet Attack", and it was up and running soon. Overall, this helps the product life cycle.
In the end, what really works in particular is actually giving players new mechanics and experiences. We have made some numerical cultivation points, but the improvement effect is quite limited, maybe around 10%~20%. But if we introduce a new hero that can give a new twist to the gameplay and allow users to find some new fun again, it works great.
Third, content is king. When making a new version, the most important thing is to think about what kind of new experience and content our version brings to users.
After talking about the advantages, reflect on the shortcomings and regrets.
First, this product broke out relatively suddenly, we only had two weeks when we went online, and this time node caught up with the e-commerce season of Double 11 and Double 12, and the ROI in the early stage was good, and I felt that it could return to the cost, so it broke out very quickly.
But this also brings a big problem, one is that the version content is seriously insufficient, and the other is that the underlying framework is not particularly good, when we update new heroes later, it is actually very difficult, and the team is also a little tired - we have to iterate on new content, and we have to do commercial optimization, and we need to have various access...... So in the first few months, a lot of core users poured in, and we didn't take the rest very well.
Second, when this product was designed, it didn't think about social networking, and it was still a pure stand-alone experience, which was also a big mistake.
There are a lot of single-player games on Steam, but no matter how good the single-player gameplay is, it will only be about one or two hundred hours for me personally. We may be able to do it for dozens of hours, but without socializing, users may get tired of it after a month or two.
Third, this framework is based on "Potato Hero", which is a chase server game, and the only social leaderboard when we did it was also invalid, because only in the first month, users can see the leaderboard when they come in. But in the next few months, the users who bought in basically couldn't pursue the rankings, which also greatly lost the possibility of us doing better commercialization.
Fourth, when we made this product, we didn't think about the medium- and long-term framework - we didn't think about our own "internal strength" well.
Regarding the future of tower defense games or games, I, as the head of R&D at the company, have been thinking about what kind of products will still stand in the future.
I am a martial arts lover, and if I compare the skills of a martial arts master, such as Tai Chi, Baguazhang, swords, guns, swords, and halberds, and dog sticks...... These moves can be so many and dazzling that they can bewilder. But the moves are easy to imitate, you can do one set of punches, and I can also play one set of punches. And the real internal strength lies in whether you have a strong body and whether you have strong practical experience. Only when moves and internal skills are combined can they exert a strong power.
When it comes to games, I think moves are a way for us to lower our CPI. For example, if you have a very novel theme, a particularly good painting style or a particularly interesting gameplay, and an engaging plot, including your product, which has generated a certain topicality, users are particularly willing to fission...... These are all moves to attract users. But this move doesn't constitute too big a barrier, especially in the current Red Sea stage, if you have a good move, others can use it immediately.
At this stage of the mini-game, it is the right solution to combine the moves with good internal skills. For example, we have a lot of products popping up this year, and they have come to a particularly high position, and I have carefully studied their team and products, in fact, they all have internal strength. In my opinion, there are internal functions similar to SLG, MMO, RPG, which can stretch the product life cycle very long, to the point of playing mechanics in years.
To sum up: the continuous loopable GVE gameplay will come to an end after two or three thousand hours of human and machine play, but people can keep playing. We are committed to constantly learning the moves, and we must also sink down and work hard to cultivate the internal skills.
In the future, we will make some attempts in tower defense. Tower defense is better combined with internal strength, and we are currently exploring SLG, MMO, and chicken gameplay. I hope that all the R&D partners of mini games will be able to invent and discover more novel and fun moves in the future, and at the same time, they can also cultivate their own internal skills and create more excellent products for Chinese mini game users, thank you.