Go Beyond CPI

A local game advertiser who expanded overseas previously asked me how overseas mobile game companies are normally operated. He wanted to understand why it was possible, through the players’ payments, to make way for the successive emergence of SuperCell, Zenga, Kabam, and others which, via one hot product, turned into “unicorns,” or successful game companies.
I answered him by saying that there is a qualitative difference in purchasing quantity domestically and overseas.

This difference may result in Chinese game developers reaching their breakout point soon, obtaining profits, and hitting the market, but through the middle stage towards the later period of the game, a situation where not continuing to purchase quantity will result in being unable to maintain users will occur.

At the same time, in comparison with CPI, overseas advertisers pay more attention to LTV, which is also the lifetime value per user acquired. As long as the LTV of each group of users is differentiated, developers can offer a price by means of the users’ LTV.

So today, I would like to explain to you what LTV is:

The 20/80 rule that everyone knows says that twenty percent of a company's staff produces eighty percent of the company revenue.
This rule is widely used in the allocation of human resources. Every company wants to find out that twenty percent of its staff and let them make the best use of their talents to gain more income for the company.
Later, we discovered that the same situation actually exists in the gaming market. Twenty percent of the players, or even five percent of the players, contribute to eighty percent of the company revenue.
Therefore, can we optimize that five percent and twenty percent?

In the early CPM(Cost Per Mille) environment, whenever each user sees a particular ad in channels such as FB, charges would start at once, which is the CPM everyone is familiar with. This charge model of CPM is widely used in various channels.
Later, CPC(Cost Per Click) and CPA(Cost Per Action) are derived, directly setting the advertisers’ final targets as pay per click or installation. This is also the mainstream channel at present and the favorite to be used with FB.

Now, almost all Chinese developers are concerned about obtaining the biggest quantity level while maintaining the CPI(Cost Per Installs). Many advertisers are very proud to say that as long as the CPI meets the conditions, our budget has no upper limit.

For example, in the United States, as long as the CPI of moderate to severe SLG games remains at $10, advertisers hope to have hundreds of millions in quantity of Installs. This is because after calculations made for each team, every country and every group orientation has a corresponding price.

However, this has a very fatal shortcoming:

First: If there are a lot of SLGs running in the United States at the same time, in the case of identical audience, the CPI of the game will rise sharply. If the advertisers still want to keep improving real income, those who are capable will increase advertising budget and improve its game quality. Or they will slightly limit their offer price, reduce the quantity level, and maintain meager profits.
Second: If a massive washout of quantity occurs in a channel within a specific time, the price and quantity level will suddenly soar or decrease.
Third: If the game is a new one, and without knowing the market price and user recovery period, one starts to purchase quantity, this is likely to lead to an advertiser’s failure to make ends meet.

Therefore, this is part of the most obvious shortcomings when purchasing quantity, with CPI as the benchmark: we cannot objectively differentiate market conditions, and most of the time we must follow the effects of market price fluctuations.

In addition, each user's LTV is completely different. If we can view ads based on income, then we can clearly discern exactly which users are of the highest quality. For example, when we found that the average LTV of one Targets user in the United States was $30, couldn’t we can spend $20 or even $25 to make a purchase?

Yes, of course.

Thus, we believe that when advertisers set a fixed CPI as target, this is not because they do not understand the value of LTV, but because they do not know how to define a user's LTV or do not know how to optimize by means of the defined LTV.

For Chinese developers going overseas, they still need to understand the concept of LTV better to be able to go farther away and get rid of malicious competition from and even piracy by other advertisers, so that the value of the product itself may be transmitted to the truly effective user.

--

曾经有一个国内出海海的游戏广告主问我，他想要了解海外游戏公司平日是怎么操作的，为什么可以经由玩家付费，接连出现SuperCell，Zenga，Kabam，这类经由一款爆品就成为独角兽，或者成为一个比较成功的游戏公司。
我回答他说，国外的买量生态与国内有一个质的差别。

这个差别可能会导致国内游戏广告主更趋近于短期内爆发，获得收益，上市，却在游戏中期与后期出现一个不继续买量就无法维持用户的一个生态。

而同时，相较于CPI， 海外广告主更注重LTV。也就是每获取的一个用户终身价值。只要每一组用户进行一个LTV的判别，那么广告主就可以经由用户的LTV来进行出价。

所以今天，我希望跟各位说明一下谈谈LTV：

大家都所熟知20/80法则，就是说，一个公司百分之二十的人员，产出了百分之八十的收入。
这一个法则被广泛利用在人力资源的配置上。每一个公司都希望找出那百分之二十的人员，并且让这些人才更物尽其用，收穫更大的收入。
后来我们发现，其实在游戏市场也是同样的情况。百分之二十的玩家，甚至百分之五的玩家，创造的百分之八十的收入。
那么，我们是否可以优化那百分之五与百分之二十呢？

在早期CPM的生态中，每一个用户看到了特定的广告，渠道诸如FB就开始收费，也就是大家所熟悉的CPM。这个CPM的Charge model被广泛用于各个渠道上。
在后来，逐渐衍生出CPC与CPA，将广告主的最终指标直接设置为每一个Click或者每一个安装进行付费。也就是现在主流渠道，与FB所喜欢使用的方式。

在现在，几乎所有中国广告主所关注的都是维持CPI的情况下获取最大的量级，许多广告主都非常豪气地说，只要CPI满足条件的情况下，我们的预算无上限。

举例而言，在美国，中重度SLG游戏，只要CPI维持在$10的情况下，广告主希望可以有数百数千万的量级。 因为这是经过每一个团队计算出来后，每一个国家，每一组定向，有一个相对应的价格。

然而，这有一个非常致命的短板：

第一：如果同时有很多SLG在美国进行跑量，在受众相同的情况下，这个游戏的CPI就会直线上升。广告主如果还希望不停地提高真正收入，能做的网游戏内加广告，提高游戏质量。或者稍微限制自己的出价，让量级降低，维持微薄的利润。
第二：如果一个渠道在特定时间内出现大规模洗量的情况，价格与量级会突然间飙升或者减少。
第三：如果这个游戏是一个新游戏，在不知道市场价格与用户回收週期的情况下就开始买量，很有可能会导致广告主入不敷出。

所以，这是以CPI为基淮，进行买量时的一部分最明显的短板：我们无法相对客观判别市场行情，而大部份时候必须跟随市场价格波动所影响。

而用户的LTV则完全不同。如果我们可以根据收入进行广告检视，那么我们可以完全清晰的辨别哪一部分的用户是最高质量的用户。举例而言，当我们发现在美国某一个Targets的用户平均LTV为$30，那么我们是否可以花费$20，甚至$25进行购买呢？

当然愿意。

因此我们认为，当广告主以固定CPI为目标时，并不是因为广告主不瞭解LTV的价值，而是不知道该如何定义一个用户的LTV，或者不知道该如何经由定义出来的LTV进行优化。

对于中国出海型的广告主，应该还更需要理解LTV的概念，才可以进一步走长走远，摆脱其他广告主的恶意竞争甚至盗图，让产品本身的价值传递给真正有效的用户。