小白学数据分析-----> 13个要重点关注的数据指标[社交游戏,翻译自国外blog] part_1

简介: 1.流失[Churn] 每个月离开游戏的用户量,有时候也选择用每周来衡量。举个例子,比如一款游戏在月初有100人在游戏,其中70个人在那个月结束后仍旧留在游戏中,那么我们就说流失率为30%,因为那个月中30个人从最初的100人中离开了游戏。

1.流失[Churn]

每个月离开游戏的用户量,有时候也选择用每周来衡量。举个例子,比如一款游戏在月初有100人在游戏,其中70个人在那个月结束后仍旧留在游戏中,那么我们就说流失率为30%,因为那个月中30个人从最初的100人中离开了游戏。

流失率也被用来分析一个玩家离开游戏的可能性。比如,一个游戏100个用户,其中30%的用户离开[30%流失率]。那么就意味着离开的可能性为30%,同样换个角度,也意味着,留下来的可能性为70%。所以如果我们要计算那个月结束后有多少玩家仍旧留在游戏中,那么我们就可以这样计算:留下来的百分比*月初的用户量,即70%*100=70,也就是说,该月结束时,有70个玩家留下来继续游戏。

如果我们计算两个月后,有多少人还留在游戏中,我们可以这么计算:70%*70%*100=49人,就是乘以两次留下来的百分比,进而计算两个月后仍有多少人在游戏。

把流失率作为一种流失可能性对待,能够帮助我们估计平均一个用户在游戏中生命周期长度。等式如下:

1/% churn =Ave.player LifeTime

举个例子,我们的月流失率为30%,那么我们的用户生命周期长度为:

1/30%=3.3 month

这点对于我们而言很重要,尤其是我们要计算平均每个用户对于我们的价值量时,这里后续会提到一个LTV(用户生命周期价值)。

如果我们忽略一个社交游戏或者其他游戏第一周的流失率,我们会发现周与周的流失率一直在5%到15%之间。每周5%的流失率对应每个月大概20%的月流失,每周15%的流失对应每个月50%的流失率。

2.平均每付费用户的收益[ARPPU]

平均每付费用户的收益,这个一般是按照月来计算的。换句话说,平均每个玩家花费了多少钱(注:大多数玩家是不花钱的,ARPPU仅仅计算那部分花钱的用户情况)。计算如下:

月总收入/月付费用户数

通常这是存在一些标杆的比如

3.平均每活跃用户收益[ARPU]

平均每活跃用户收益,如同ARPPU一样,这个也是每个月计算一次,ARPU的计算如下:

月总收入/月独立用户(月活跃用户)

依据账户方法论,这个指标能够告诉你从来自于衣服类道具购买的收入,该如何在用户生命周期中被均摊,同样的比如消耗性的,能量性(红、蓝药)的道具也能立刻通过此加以认识。但我认为这有一点小复杂。

Zynga的财报中我们看到他们每个月的ARPU大概是$0.40,当然这是他们全部游戏的一个计算值,实际上不同类型游戏的表现比另一些货币化的更好。

  • Casual Social Game: A casual game is designed for anyone, including those without prior gaming experience. Such as Farmville, Cityville, Bejeweled, Words with Friends. ARPU around $0.10 – $0.20
  • Virtual Currency Poker and Casino Games: Traditional gambling games that only allow players to play with virtual currency, such as Zynga Poker, Slotomania. ARPU around $0.25 – $1.25
  • Mid-core Social Game: Typically more investment is required to succeed. Tends to be more competitive in nature, and players can be punished for not playing well, such as Mafia Wars and Backyard Monsters. ARPU: $0.25 – $1.25
  • Virtual Worlds: Online worlds where players create avatars and interact in realtime, such as Habbo Hotel, Club Penguin, Runescape, and Puzzle Pirates. ARPU around $0.84 – $1.62

4.生命周期价值[LTV]

生命周期价值指的是平均每个玩家的消费金额。LTV包括了付费与非付费玩家。LTV计算如下:

ARPU*玩家留存游戏中的平均月的数量[玩家登录游戏平均月数量即 平均生命周期]

比如,如果我们的ARPU=$0.5,玩家总计在有5个月每个月都登录过游戏[平均生命周期=5month],那么LTV就是

$0.5*5=$2.5

刚才我们说到我们使用流失率来计算用户平均生命周期,实际是很有效的,你大概在一个月后就能知道平均用户会呆在游戏中多久,也包括他们会花多少钱(ARPU),而且我们也能够知道每个用户的生命周期价值是多少(LTV)。通过此,你就可以了解,每个用户对于你的价值是多少,进而你也就明白了广告投放你要为新用户话费多少。

5.K因子[K-Factor]

病毒增长衡量标准。计算如下:

感染率*转化率

所谓转化率指的是当感染后转化为新用户的情况。

如果K-Factor是1,那么就意味着每个玩家都能带来另一个玩家到你的游戏中,游戏不增长,不下降[用户量];

如果K-Factor是小于1,那么游戏现在进行的营销将会耗尽玩家;

如果K-Factor是大于1,那么你的游戏时按照指数级增长的。

其实,极少的游戏能够做到K-Factor大于1的情况,这也是为什么Zynga在2011年的Q1花费了4000w美元在市场营销上面。所有的游戏需要营销支持发展。当然成功游戏与失败游戏之间的区别就在于成功游戏花钱在营销上,但是却能够从获得玩家身上同样赚取利润。这也是接下来的定义为什么这么重要。

6.CPA

获得每个用户的花费,用于衡量把一个用户导入到游戏中的花费。CPA的衡量多种形式,推荐的形式如下:

新访问者--->注册--->完成新手教程--->变成真正意义的玩家

因此,广告联盟计算是:

总花费/带来的新玩家

比如花费1000美元在Google Adwords,获得1000玩家,那么CPA就是:1000/1000=1美元CPA

未完待续

 

原文:

 

1. Churn

 

The percentage of users who leave your game each month, or sometimes measured as the percentage who leave each week. For example, if a game that has 100 users at the start of the month, and 70 of those users are still playing the game at the end of the month, then we would say the churn rate is 30% because 30 of the original 100 people left that month.

The churn rate can also be thought of as a probability that a player will leave. For example, imagine a game that has 100 players and a 30% of any player leaving (30% churn). That 30% chance of leaving could also be thought of as a 70% chance of staying. So if we want to figure out how many players will be left at the end of the month all we multiply the chance of staying by the number of players at the start of the month. So 70% x 100 = 70 players at the end of the month.

 

To calculate how many will be left after two months we can simply do it twice, 70% x 70% x 100 = 49 players after two months.

 

Treating churn as a probability allows us to estimate how long the average person plays your game. The equation is simple: 1 / % Churn = Ave. Player Lifetime. For example, with our 30% monthly churn rate we find that 1 / 30% = 3.3 months average player lifetime.

 

This comes in important later when we want to calculate how much the average player is worth to us, or the LifeTime Value (LTV).

 

Ignoring the first week (we’ll talk about that in the Onboarding definition) a social game or virtual world will typically see week to week churn around 5% to 15%. A 5% weekly churn is equivalent to roughly 20% monthly churn. While 15% weekly churn is equivalent to 50% monthly churn.

 

Phew, that was a long one, don’t worry, the rest are shorter!

 

2. ARPPU

 

Average Revenue Per Paying User, usually measured each month. In other words, how much money does the average customer spend (most of your players will never spend any money, ARPPU only includes those who spend money). It can be calculated as total monthly revenue divided by total monthly paying users. Some benchmarks:

 

 

Club Penguin, on the other hand, has subscriptions and no micropayments. Their ARPPU is somewhere around the $6 mark.

 

3. ARPU

 

Average revenue per active user, and like the ARPPU this is also measured each month. The ARPU is calculated by dividing the total revenue for the month by the total number of unique players for the month. Sort of. Depending on the account methodology used it could be said that revenue from the purchase of a virtual clothing item should be amortized over the players lifetime, where as energy and consumables can be recognized immediately. I think that’s just making it all a little too complicated!

 

From the Zynga IPO filing (link) that the average revenue per user per month is around $0.40. Of course, that’s across their whole portfolio of games. In practice, different types of games monetize better than others.

 

  • Casual Social Game: A casual game is designed for anyone, including those without prior gaming experience. Such as Farmville, Cityville, Bejeweled, Words with Friends. ARPU around $0.10 – $0.20
  • Virtual Currency Poker and Casino Games: Traditional gambling games that only allow players to play with virtual currency, such as Zynga Poker, Slotomania. ARPU around $0.25 – $1.25
  • Mid-core Social Game: Typically more investment is required to succeed. Tends to be more competitive in nature, and players can be punished for not playing well, such as Mafia Wars and Backyard Monsters. ARPU: $0.25 – $1.25
  • Virtual Worlds: Online worlds where players create avatars and interact in realtime, such as Habbo Hotel, Club Penguin, Runescape, and Puzzle Pirates. ARPU around $0.84 – $1.62

 

4. LTV

 

The Life Time Value is the average amount of money spent by each player. The LTV includes paying and no-paying players. To calculate the LTV you multiply the ARPU by the average number of months a player stays in your game. For example, if the ARPU is $0.50 and the average player lifetime is 5 months then the LTV is $0.50 x 5 = $2.50.

 

Earlier we used the Churn Rate to calculate the average player life time. This is really useful. After only 1 month you know roughly how long the average player will stay in the game, and you know how much they spend on average (ARPU) and therefore how much each player is worth over their lifetime (LTV). Through knowing how much each player is worth you can figure out how much you can afford to spend on advertising for new players.

 

5. K-Factor

 

The measure of viral growth. It’s calculated by multiplying the Infection Rate by the Conversion rate. The conversion rate is when the ‘Infection’ turns into a new user.

 

A K-Factor of 1 means every member is bringing one additional member to your game, your game is not growing nor is the game declining.  A K-Factor of less than 1 means that without ongoing marketing your game will run out of players. While a K-Factor greater than 1 means that your game is growing exponentially.

 

It is very rare that any game will ever have a K-Factor greater than 1. That’s why Zynga spent $40,000,000 on marketing in Q1 2011. All games need marketing to continue growing. Of course the difference between a successful game and a failure is that a successful game spends money on marketing but still makes a profit on each player they acquire. That’s why the next definition is so important…

 

6. CPA

 

The Cost Per Acquisition is a measure of the cost of bringing that user to your game. The CPA can be measured in different ways. We recommend measuring the CPA as the cost to convert a new visitor from the homepage into someone who has registered, finished the tutorial, and become a player. So the CPA for an advertising campaign can be calculated by dividing the total spend by the number of new players. If we spend $1000 on Google Adwords and get 2000 new players then our CPA is $1000 / 1000 = $1.00 CPA.

 

 

 

相关文章
|
18天前
|
数据挖掘 PyTorch TensorFlow
|
1月前
|
数据采集 DataWorks 数据挖掘
提升数据分析效率:DataWorks在企业级数据治理中的应用
【8月更文第25天】本文将探讨阿里巴巴云的DataWorks平台如何通过建立统一的数据标准、规范以及实现数据质量监控和元数据管理来提高企业的数据分析效率。我们将通过具体的案例研究和技术实践来展示DataWorks如何简化数据处理流程,减少成本,并加速业务决策。
167 54
|
17天前
|
机器学习/深度学习 数据挖掘 TensorFlow
🔍揭秘Python数据分析奥秘,TensorFlow助力解锁数据背后的亿万商机
【9月更文挑战第11天】在信息爆炸的时代,数据如沉睡的宝藏,等待发掘。Python以简洁的语法和丰富的库生态成为数据分析的首选,而TensorFlow则为深度学习赋能,助你洞察数据核心,解锁商机。通过Pandas库,我们可以轻松处理结构化数据,进行统计分析和可视化;TensorFlow则能构建复杂的神经网络模型,捕捉非线性关系,提升预测准确性。两者的结合,让你在商业竞争中脱颖而出,把握市场脉搏,释放数据的无限价值。以下是使用Pandas进行简单数据分析的示例:
30 5
|
1月前
|
存储 数据挖掘 数据处理
DataFrame探索之旅:如何一眼洞察数据本质,提升你的数据分析能力?
【8月更文挑战第22天】本文通过电商用户订单数据的案例,展示了如何使用Python的pandas库查看DataFrame信息。首先导入数据并使用`head()`, `columns`, `shape`, `describe()`, 和 `dtypes` 方法来快速概览数据的基本特征。接着,通过对数据进行分组操作计算每位顾客的平均订单金额,以此展示初步数据分析的过程。掌握这些技能对于高效的数据分析至关重要。
32 2
|
1月前
|
数据采集 机器学习/深度学习 算法
"揭秘数据质量自动化的秘密武器:机器学习模型如何精准捕捉数据中的‘隐形陷阱’,让你的数据分析无懈可击?"
【8月更文挑战第20天】随着大数据成为核心资源,数据质量直接影响机器学习模型的准确性和效果。传统的人工审查方法效率低且易错。本文介绍如何运用机器学习自动化评估数据质量,解决缺失值、异常值等问题,提升模型训练效率和预测准确性。通过Python和scikit-learn示例展示了异常值检测的过程,最后强调在自动化评估的同时结合人工审查的重要性。
55 2
|
18天前
|
机器学习/深度学习 数据挖掘 TensorFlow
从数据小白到AI专家:Python数据分析与TensorFlow/PyTorch深度学习的蜕变之路
【9月更文挑战第10天】从数据新手成长为AI专家,需先掌握Python基础语法,并学会使用NumPy和Pandas进行数据分析。接着,通过Matplotlib和Seaborn实现数据可视化,最后利用TensorFlow或PyTorch探索深度学习。这一过程涉及从数据清洗、可视化到构建神经网络的多个步骤,每一步都需不断实践与学习。借助Python的强大功能及各类库的支持,你能逐步解锁数据的深层价值。
36 0
|
1月前
|
数据采集 数据可视化 算法
GitHub星标68K!Python数据分析入门手册带你从数据获取到可视化
Python作为一门优秀的编程语言,近年来受到很多编程爱好者的青睐。一是因为Python本身具有简捷优美、易学易用的特点;二是由于互联网的飞速发展,我们正迎来大数据的时代,而Python 无论是在数据的采集与处理方面,还是在数据分析与可视化方面都有独特的优势。我们可以利用 Python 便捷地开展与数据相关的项目,以很低的学习成本快速完成项目的研究。
|
1月前
|
数据采集 数据可视化 算法
GitHub星标68K!Python数据分析入门手册带你从数据获取到可视化
Python作为一门优秀的编程语言,近年来受到很多编程爱好者的青睐。一是因为Python本身具有简捷优美、易学易用的特点;二是由于互联网的飞速发展,我们正迎来大数据的时代,而Python 无论是在数据的采集与处理方面,还是在数据分析与可视化方面都有独特的优势。我们可以利用 Python 便捷地开展与数据相关的项目,以很低的学习成本快速完成项目的研究。 今天给小伙伴们分享的这份Python数据分析入门手册本着实用性的目的,着眼于整个数据分析的流程,介绍了从数据采集到可视化的大致流程。
|
28天前
|
SQL 数据挖掘 Serverless
SQL 窗口函数简直太厉害啦!复杂数据分析的超强利器,带你轻松攻克数据难题,快来一探究竟!
【8月更文挑战第31天】在数据驱动时代,高效处理和分析大量数据至关重要。SQL窗口函数可对一组行操作并返回结果集,无需分组即可保留原始行信息。本文将介绍窗口函数的分类、应用场景及最佳实践,助您掌握这一强大工具。例如,在销售数据分析中,可使用窗口函数计算累计销售额和移动平均销售额,更好地理解业务趋势。
40 0
|
28天前
|
SQL 数据可视化 数据挖掘
SQL 在数据分析中简直太牛啦!从数据提取到可视化,带你领略强大数据库语言的神奇魅力!
【8月更文挑战第31天】在数据驱动时代,SQL(Structured Query Language)作为强大的数据库查询语言,在数据分析中扮演着关键角色。它不仅能够高效准确地提取所需数据,还能通过丰富的函数和操作符对数据进行清洗与转换,确保其适用于进一步分析。借助 SQL 的聚合、分组及排序功能,用户可以从多角度深入分析数据,为企业决策提供有力支持。尽管 SQL 本身不支持数据可视化,但其查询结果可轻松导出至 Excel、Python、R 等工具中进行可视化处理,帮助用户更直观地理解数据。掌握 SQL 可显著提升数据分析效率,助力挖掘数据价值。
35 0

热门文章

最新文章