作者：Sarah Perez

手机应用开发者滥用推送通知，这种现象必须予以制止。虽然普遍看法是，持续向用户发送非关键性更新内容不是有效策略，但应用开发者似乎很难抵挡此免费营销渠道——选择性推送。频繁的通知信息能够促使用户将应用放在心上，促使他们重新启动应用，查看新内容，适时提醒他们，甚至更多—–或者，至少这是应用开发者的期望。但有用/友善与永不停息的垃圾信息之间存在明显界限。

推送通知能够让社交应用看起来颇受欢迎，否则它可能没有展示的渠道。手机持续响起，以告知我们有多少用户正在查看新应用。但用户能够接受的展示内容有限度。

关于某位好友订阅某服务，或发表评论，或“liked”某内容，或加你或关注你之类的通知鲜少是必要信息。这不是真实新闻。知道自己多一位粉丝并不会提高你的生活品质。这并不值得你将注意力从现实生活的手头工作转移，你的手头工作通常更加重要。虽然你也许会想要立即获得某些社交应用的相关信息，但你多半不希望在自己所安装的应用中被这些更新内容打断。

这只是一种类型，但社交应用通常是最严重的冒犯者。除社交类型外，还有很多各式各样的推送通知类型。查看如下列表：

娱乐

* 新内容（报道，专题）

* 时效性事件

* 日常专题（每日新闻）

出版业

*轰动新闻

* 本地通知

* 读者抽奖

* 服务信息

* 免费试用切换提示

* 订阅追加销售的单一版本

* 特别策划或问题聚焦

* 本地商业信息

媒体

* 新内容（完整事件、广告传单、奖励内容等）

* 提醒（表演事件、特别露面、现场事件）

* 投票（问题、提醒、结果）

* 将用户推向其他媒介（手机网站、电视节目）

零售业

* 征求产品评论

* 关于你综述及所追踪产品的评论

* 产品可用性/新风格

* 发货确认

* 订单查询期满

* 交易/产品销售

* 个人优惠券

* 建议

* 征求售后服务评论

* 发送品牌化娱乐/合作模式和顾客营销（视频和照片等）

* 基于本地数据的推送消息，如适合置身滑雪小镇用户的滑雪橇（游戏邦注：扩展至适合其他运动场所）。

社交网络

* 回应+所发布内容的反馈信息

* 用户互动

* 来自所追踪账号的新内容

* 功能更新

* 推进用户粘性的广播（新内容、你的内容得到推荐等）

B2B

* 新文章

* 即时支持

* 即时用户服务

上述内容中，有多少条你认为是关键信息？我想大概1个手指就能数完。即时这样，关于多久及何时看到这些内容，你依然有自己的若干限定标准。

推送通知滥用所存在的问题不仅仅在于，当你被毫无意义的内容打断时，你会感到非常恼怒，还包括用户将对整个推动消息机制本身产生怀疑。我想我们大概已心生怀疑。若应用初次询问是否发送推送通知时，你就予以否定回答，那么你基本就是持有此观点。用户不再相信开发者会合理运用机制，所以我们决定完全退出通知机制。

这非常遗憾，因为若把握得当，推送通知完全具有可行性。Urban Airship（游戏邦注：他们是通知功能提供商）表示，遵循“合理”通知消息规范的应用能够在内容下载后的6个月里将留存率提高1倍。“合理”推送通知的优点将与日俱增。合理推送通知能够在内容下载后的头一个月里将应用使用率提高67%，2个月和3个月期间分别能够提高74%和81%。

合理推送通知和用户具有高度关联性，着眼于满足需求。记住，是用户的需求，而不是应用开发者的需求。只有开发者，而不是用户才会关心应用的启动频率。Urban Airship首席营销官Brent Hieggelke表示，“很多时候，我们看到的是，推送通知只迎合应用或品牌目标，而非着眼于有意义的内容，这会很快流失用户。不要让推送通知打断他们的生活，因此你需要确保内容有价值。”

那么合理推送通知的具体例子有什么呢？他表示，Burton Snowboards就是个典型例子。他们通过向用户提供实时降雪量推送通知将用户同相关信息联系起来。其他公司则也许会采取不同策略，例如发送许多促销信息。这有可能会流失公司的忠实用户。Burton通过告知用户有用信息，树立自己的商誉及保持内容的相关性。

当然，最困难的部分在于，找出用户认为“合理”的内容，Burton Snowboards应用用户多半会关心实时预警。但每个用户都截然不同，从理想角度来说，他们应该在通知内容和通知频率上享有更多选择权。各应用的设置应融入相关控制装置。如果没有，那么开发者就会因为过于谨慎而犯错。毕竟，在界面上删除应用要比通过手机设备的深入开关设置操作简单许多。（本文为游戏邦/gamerboom.com编译，拒绝任何不保留版权的转载，如需转载请联系：游戏邦）

App Developers: Stop Abusing Push!

By Sarah Perez

Mobile app developers are abusing push notifications and it has to stop. Although it’s widely understood that it’s not an effective strategy to continually ping users with non-critical updates, app developers can’t seem to help themselves to the free marketing channel that is the opt-in push. Continual notifications keep the app at the forefront of users’ minds, entice them to relaunch the app to see what’s new, remind them of something they may have forgotten, and more – or, at least, so the app developer hopes. But it’s a fine line between being seen as useful and friendly versus a source of never-ending message spam.

Push notifications can give a social app an air of popularity it wouldn’t otherwise have the means to display. The phone continually buzzes to indicate how many people are checking out the new app. But there’s only so much showing off a user can take.

Notification that a friend of yours signed up for the service, or left a comment, or “liked” something, or added you or followed you, etc. is very rarely must-have information. It’s not actual news. It doesn’t improve your life to know you have one more follower. It doesn’t deserve the right to pull your attention away from whatever you were doing out there in the real world, which is almost always far more important. While there may be the rare social app where you want to know these things immediately (Facebook perhaps?), in all likelihood, you don’t want to be interrupted with these sorts of updates for every app you install.

That’s just one genre, but social apps are often the worst offender. Beyond social, there are all sorts of push notification types out there. See this list courtesy of Urban Airship:

Entertainment

* New content (items, features)

* Time-sensitive events

* Routine features (word of the day)

Publishing

* Breaking news

* Local alerts

* Reader sweepstakes

* Service messages

* Free trial conversion prompts

* Single-edition to subscription upselling

* Special section or issue highlights

* Local offers

Media

* New content (full episodes, teasers, bonus content etc.)

* Reminders (show times, special appearances, live events)

* Polling (Questions, reminders, results)

* Drive users to other media (mobile site, television program)

Retail

* Ask for Product reviews

* Comments on your reviews and the products you follow

* Product availability/new styles

* Shipping confirmation

* Shipping updates

* Shopping cart expiration

* Deals/sales

* Personal coupons

* Recommendations

* Ask for customer service reviews

* Send branded entertainment/co-op and shopper marketing (video, picture, etc.)

* Push based on location data ie ski gear to people who have been in ski town zip codes ( expand to fit additional sports locations)

Social Networking

* Responses + feedback on posted content

* User interactions

* New content from followed accounts

* Feature updates

* Broadcasts encouraging engagement (new contest, your content has been featured, etc.)

B2B

* New articles

* Real-time support

* Real-time customer service

How many of those items above do you actually think are critical messages? I’ll bet you can count them on one hand. And even then, I’ll bet you have a few qualifiers about how often and when you want to see them.

The problem with push notification abuse is not just that it’s annoying when you’re interrupted by something that has no meaning to you, it’s that it creates an environment where users become suspect of the whole push notification mechanism itself. I’d argue we’re already there. If you’ve said “no” to an app upon first launch when it asks your permission to send push notifications, then you basically agree. Users can no longer trust developers to use the system properly, so we’re opting out of notifications entirely.

That’s a shame because when done well, push notifications can and do work. Urban Airship says that apps following “good” push notification practices can actually more than double the retention rate over a six-month period after download. And the benefits of “good” push grows over time. Good push notifications drive 67% of app usage the first month after download, 74% of usage two months after download, and 81% of usage three months after download, they found.

Good push notifications, in case you’re wondering, are those that are highly relevant to the user and focused on meeting their needs. Remember, that’s the user’s needs, not the app developers’ wants. Only developers care about how often an app is launched, not the user. “We’ve seen time and time again, push notifications that solely serve an app’s or brand’s objectives rather than what is really meaningful can turn people off fast,” said Brent Hieggelke, CMO, Urban Airship. “It has to pass the family dinner test. Push interrupts their lives, so you have to make it worth it.”

So what is a an example of a good push notification? Burton Snowboards is a perfect example, he said. They found a way to engage users with relevant information by offering to push fresh snowfall alerts to customers. Another company might have taken a different approach, like sending a ton of promotional sale messages. This would risk alienating the company’s best customers (“Hey, I already bought the snowboard! Give it a break!”) Instead, Burton builds goodwill and stays relevant by telling customers something useful.

Of course, the hard part is figuring out what a user considers “good.” Seemingly, a Burton Snowboards app user probably does care about fresh powder alerts. But every user is different, and ideally, they should be given more of a choice about what and how often you ping them. There should be controls in every app’s settings for this. And if there’s not, then developers should definitely err on the side of caution here. After all, it’s a lot easier to X an app from the homescreen than it is to toggle a switch deep in the phone’s settings somewhere.（Source：techcrunch）