作者：Jeff Grubb

比起艺术，手机游戏更倾向于科学，这也是许多开发者更注重创造性而缺少利用数据能力的问题所在。同时这也是OtherLevels会提供海军训练营般速成课去帮助一些工作室更好地了解自己玩家的原因。

OtherLevels是一家面向手机游戏开发商的服务型公司，他们专注于研究如何保持玩家对于免费游戏的兴趣。该公司通过识别行为模式然后使用各种信息去瞄准特定人群而将他们重新带回游戏中。在执行了这些技巧后，像《水果忍者》的开发商Halfbrick便增加了34%的活跃玩家数，并且像《疯狂喷气机》等游戏的收益也得到了100%的提升。为了帮助像Halfbrick这样的工作室，OtherLevels创造了一个软件工具包和一个“FastStar”程序。开发者可以独立使用这一软件工具包，并且该工具包也让OtherLevels能够通过12至16周的课程去帮助游戏开团队学会如何更好地使用重要数据。基于FastStart课程，OtherLevels还能够与像Halfbrick等工作室维持一种持久的关系。如今的手机游戏价值已高达348亿美元，所以保持玩家长期的兴趣便是获取收益的关键方法，这也是像OtherLevels等公司想努力的目标。

Jetpack Joyride (from crsky)

手机游戏的用户粘性指代的是玩家一周多次回到自己喜欢的应用中。但根据OtherLevels的总经理Brendan O’ Kane，这同时也是一个需要像Halfbrick等公司花大量时间去思考的复杂理念。

他说道：“Halfbrick需要考虑一些不同的用户粘性元素。他们知道自己必须面对opted-in用户和opted-out用户。”

在这里，opt-in指的是那些愿意接收来自应用推送通知的玩家。这意味着游戏可以告诉他们“嘿，你已经很久没出现了。回来吧，我们会给你一些免费的货币”或其它相似的内容。Hlafbrick承认在与OtherLevels合作前他们根本不知道如何与这些玩家进行对话。

Halfbrick的首席市场营销官Nicholas Corneliu声称：“（游戏邦注：在与OtherLevels合作之前）我们通过游戏内部渠道所进行的交流是不同且分散的。我们希望OtherLevels能够帮助我们处理信息策略中一些相矛盾的问题，并进一步强化用户终身价值且降低我们的玩家获取成本。”

O’Kane表示，FastStart便能在此发挥作用。

他说道：“FastStart是一种结构化程序。我们进行了许多分析与测试，所以我们拥有这样一种基于事实的方法去应对可能发生的情况。我们会大量使用像控制群组这样的技巧。我们也会追踪各种内容。”

在展示了工作室是如何获取精确数据后，OtherLevels向我们展示了改变特定信息所创造的正面和负面结果。例如，如果推送通知能够提高10%的消费，但却会导致15%的玩家离去，这便不是一种有价值的方法。

O’Kane说道：“我们是与消费者一起努力做到这点，我们也致力于分析自己的技巧和一些最佳实践方式。我们的目标是他们可以依赖于自己想做的事并真正去执行它们，并将我们所分享的这些技巧带到实践中去。”

有时候像Halfbrick这样的工作室会将OtherLevels当成持续咨询对象。即当你在考虑所有的变量时，你必须考虑何时该重新去吸引那些已经离去的玩家。

O’Kane说道：“我们对不同类型的信息，不同信息节奏都进行了许多次的测试。开发者总是会问这样的问题：三天后，四天后，我们是否该伸出援助之手了？拥有反复回到游戏中的用户以及拥有消费能力很高的用户之间有什么区别？也许用户们只是感到疲倦了。我们不该反复追随他们以将其带回游戏中。也许对于特定用户群体来说较缓慢的节奏会更合适。我们进行了许多测试去理解什么方法能够持续发挥作用以及怎样的方法具有最佳实践性。”

这些都是很重要的理念。大多数开发者可能会独立进行思考。但一旦他们看到策略中小小的变化对于整个玩家基础所产生的连锁作用，他们便会为此而惊讶。

Cornelius说道：“到目前为止OtherLevels的最大发现便是意识到像信息语调，频率，节奏和发送时机等方面的小小改变以及具有预测性的分析能够有效帮助我们接触到更多用户。”

最后，OtherLevels指出开发者在此所面临的最大挑战便是提醒自己不该放弃那些未接收推送通知的玩家。

O’Kane发现，尽管有机会与离开的粉丝进行交流很棒，但这却不应该是以为那些活跃于游戏中的玩家创造信息为代价。FastStar便尝试着将开发者带向使用游戏内部信息与玩家保持交流的理念中。

他说道：“我们经常谈到‘100%的用户’，而对于我们来说这便意味着你已经拥有沉浸于你的游戏中的玩家。他们中的有些人可能很乐于接收不同类型的信息，所以他们是opt-in玩家。但是其他人可能并非如此。而你需要做的便是着眼于整体用户并以此创造活动和方法去与所有的用户进行交流，而不该只是瞄准那些愿意接收信息的用户。”

（本文为游戏邦/gamerboom.com编译，拒绝任何不保留版权的转发，如需转载请联系：游戏邦）

Why Fruit Ninja developer Halfbrick went to an analytics boot camp

JEFF GRUBB

Mobile gaming is more a science than an art, and that’s a problem for some developers that are heavy on creativity and lacking in a capacity to crunch through data. That’s why one company is offering a boot camp-style crash course to help studios better understand their players.

OtherLevels is a services company for mobile developers that specializes in keeping players engaged in free-to-play games. The company does this by recognizing patterns in behavior and then targeting certain kinds of people with various types of messages to get them to return to a game they enjoy. After implementing these techniques, Fruit Ninja developer Halfbrick saw a 34 percent increase in active players and 100 percent increase in revenue for games like Jetpack Joyride. To help studios like Halfbricks (along with ZeptoLabs and online gambling site Bwin), OtherLevels has a software toolkit that and a “FastStart? program. The former is something developers can use on their own while the former has OtherLevels running a game’s leadership team through a 12-to-16-week course to help them better surface and act on important data. On top of the FastStart classes, OtherLevels can also form a lasting partnership with a developer like the one it has had for the last three and a half years with Fruit Ninja developer Halfbrick. Mobile gaming is worth $34.8 billion, and keeping engagement high is a key piece to capturing that revenue, which is why a firm like OtherLevels is thriving.

Engagement in mobile gaming is the buzzword that means a player is coming back several times a week to their favorite app. But it’s a complex idea that companies like Halfbrick spend a lot of time thinking about, according to OtherLevels managing director Brendan O’Kane.

“Halfbrick has thought about different kinds of engagement,” he told GamesBeat. “They know that they have to deal with the opted-in audience and the opted-out audience.”

In this context, opt-in refers to the player who says yes to receiving push notifications from an app. This means a game can say “Hey, you haven’t been around for a while. Come back, and we’ll give you some free coins” or something along those lines. That leaves the other kind of player who is nearly impossible to reach outside of the game. But either way, Halfbrick admits that it was not great at speaking to either of these people prior to working with OtherLevels.

“The communications we were sending through in-game channels were disparate and fragmented [prior to working with OtherLevels],” Halfbrick chief marketing officer Nicholas Cornelius claimed. “We looked to OtherLevels to help us tackle the inconsistency issues in our overall messaging strategy, chiefly to drive stronger customer lifetime value and lower our player acquisition costs.”

O’Kane says that this is where FastStart can really make a difference.

“FastStart is a very structured program,” he said. “We do extensive analytics and testing so we can have a fact-based approach to what’s going on. We’ll use techniques like control groups extensively. We track everything.”

After showing the studios how to get accurate data, OtherLevels shows them how changing certain messages can have both positive and negative outcomes. For example, if a push notification is driving a 10-percent increase in spending but producing 15 percent fewer players, it probably isn’t worth pursuing.

“We doing this in conjunction with our customers, and we’re sharing techniques and best practices,” said O’Kane. “The goal is that they can then take that, and depending on what they particularly want to do, they can pick it up and run with it and apply some of those techniques on their own.”

Sometimes studios, like in the case of Halfbrick, keep OtherLevels around in an ongoing consultancy position. This makes a lot of sense when you consider all of the variables you have to consider when trying to reengage a lost player.

“We do a lot of testing in terms of different types of messages, different cadence of messages, playing with user journeys,” said O’Kane. “Developers are always asking things like: Should we reach out after three days? Should we reach out after four? How does that vary by comparison if we have, for example, someone who’s been a high frequency user, or someone who might have been a high spending user? Perhaps they’re just tired. It’s inappropriate to go chasing them three times to jump back in. Maybe a slower cadence for a particular class of user. We do a lot of that type of testing to understand what’s going to work and what’s the best practice.”

And those are the big ideas. Most developers can probably think through a lot of those on their own. But once they see how a tiny shift in strategy can have a huge ripple effect across their entire player base, it can get overwhelming.

“So far the biggest [revelation with OtherLevels] is realizing how small changes in things like message tonality, frequency, cadence, and timing and better predictive analytics are helping us reach a stronger percentage of users per campaign,” said Cornelius

Finally, OtherLevels points out that one of the biggest challenges it has with developers is reminding them not to give up on players who don’t opt-in for push notifications.

O’Kane notes that while it’s great to have the opportunity to talk to lapsed fans, it shouldn’t come at the expense of crafting messages for people who are actively in the game. And FastStart tries to walk developers through the idea of keeping in contact with players using in-game messages.

“We often talk about ‘100 percent of audience,’” he said. “That, for us, is a reflection that you already have people engaging with you. Some of them are happy to receive messages of different types. They’ve opted in. Others aren’t. But you should look at your whole audience and build your campaigns and build your programs to work with that whole audience, not just the portion that’s ticked the box and said, I’m happy to be messaged.”（source：venturebeat）