应用商店优化 (AS0) 2020 大会：回顾与学习

时间：2020 年 5 月 13 日

地点：在线会议

内容：专注于专家级应用商店优化讨论的独家活动

第三届应用商店优化大会于 2020 年 5 月 13 日在柏林举行，来自 68 个不同国家/地区的 1,400 多人通过线上参与了本次会议。本次会议的目标是在 COVID-19 带来的挑战下，继续分享关于应用商店优化的知识。应用商店优化社区提供了非常多的insights，无法全部总结，但在本文中，我们收集了您应该了解的 13 场演讲中的主要内容。

1. Covid-19 对移动行业的影响

在开幕致辞之后，Phiture 的联合创始人 Moritz Daan 快速概述了 COVID-19 对移动应用行业的影响：

• 下载在 全球范围内总体增加，Apple App Store（3 月 13 日左右）和 Google Play Store 均是如此。
• 周末季节性已不再明显，因为工作日和周末融为一体。例如，即使在工作日，游戏现在的下载量也非常大。
• 从 COVID-19 的影响中获胜的类别是商业、健康与健身、医疗、教育和新闻，而失败的类别是旅游、体育、导航、天气和书籍。
• 根据牛津大学的数据，类别中注意到的变化与政府应对政策的不同顺序直接相关。例如，美国旅游类别下载量的急剧下降是由于国家旅行的停止。
• 关键词搜索流量从一天到另一天发生了巨大变化，一些重要的关键词消失了，而另一些关键词变得非常重要。
• 应用必须调整其策略以适应这种情况，因为 Apple 要求在考虑 COVID-19 的情况下，在艺术品和创意中反映社交疏远指南。
  应用创意中反映的社交疏远指南”>
  在应用创意中反映的社交疏远指南。

开幕致辞为之后各种各样的应用商店优化演讲和小组讨论奠定了良好的基础。这些演讲涵盖了应用商店优化许多不同的重要方面，但其中许多都讨论了本地化这一主题，这似乎是 2020 年应用商店优化的热门话题。

2. 构建正确的仪表板来衡量 ASO 影响和成功

Nadir garouche，高级增长顾问 – tilting point

由于应用商店优化活动的成功取决于内部和外部因素，因此衡量和统一不同的 KPI 对于充分了解应用的性能至关重要。在 App Store Connect 中，Tilting Point 的高级增长顾问 Nadir Garouche 建议使用“App Units”而不是“Installations”，因为它们在您的分析中可能更相关。“App Purchase Date”是一个有用的过滤器，用于创建在给定日期或时间范围内安装您的应用的用户群。但是，由于您在 App Store Connect 中仅限于两个维度，因此您需要获取您的数据并创建您自己的自定义仪表板，以进行深入的分析，包括每个主要商店指标的图表。为了将提升归因于关键词优化，他建议将搜索安装与关键词可见性指标进行比较。

Tilting point 的主仪表板 – 在一个视图中聚合主要商店指标。

在 Google Play 上，保留率非常重要，必须加以考虑。分析保留率和流量之间的相关性可以带来有价值的insights。此外，崩溃可能会对您的可见性产生负面影响，但 Google Play 可能需要 30 天才能反映出来。

最后，为了在运行 ASA 活动时报告效果，Nadir 建议从 App Store Connect App Units 中减去 MMP ASA 安装*。另一种选择是使用 App Store Connect 销售额/按 IAP* 减去 MMP ASA 销售额/IAP。通过并排查看 App Store 效果图表和 Apple Search Ads 效果图表，您可以发现趋势并查看指标如何在同一时间段内演变。

*MMP：移动衡量合作伙伴
*应用内购买

3. 实现应用商店优化、效果营销和品牌之间的一致性

Natalie drozd，营销经理 – hily

如何结合跨团队知识来创建每个人都会喜欢的产品页面？虽然应用商店优化、品牌和付费 UA 团队具有相同的长期目标，但在短期内它们可能会有所不同。Hily 的营销经理 Natalie Drozd 解释说，为了解决这个问题，每个团队都必须专注于其核心职责，但总体上要共享信息、提出问题并倾听其他人的意见。首先，应用商店优化团队必须采用“可发现性优先的方法”，通过选择具有信息丰富的创意且相关且高流量的关键词。其次，关于付费和效果营销，目标是通过采用投资回报率优先的方法来降低价格，最大化 LTV。第三，对于品牌，该方法围绕使命和价值观，最终目标是在人们的脑海中留下深刻的印象。

虽然这些目标中的每一个都定义明确，但她强调团队之间存在重叠，协作是关键。提出好的问题可以产生新的想法并创造价值。例如，应用商店优化团队可以从品牌和付费团队那里学到很多关于效果最佳的视觉效果、广告设计、品牌定位和身份的信息。

框架：结合从意识到着陆页的跨团队知识。

4. 使用 NPS 流程优化您的评分和评论

Alex gorius，高级增长顾问 – phiture
Alice muir，增长顾问 – phiture

Phiture 的 Alex Gorius 和 Alice Muir 谈到了如何最好地使用 NPS（净推荐值），这是一个用于衡量总体体验的指标，或者换句话说，用户幸福感，范围从 0 到 10。然后，Alice 解释了使用 NPS 的不同好处：

1. 通过更好地了解批评者来最大限度地减少客户流失。
2. 通过提示推广者将您的应用推荐给他们的随行人员来提高知名度。
3. 通过鼓励推广者对您的应用进行排名来提高应用评分。
4. 通过识别产品问题（例如平台或本地化问题）来改进您的应用。

Alex and Alice then highlighted the importance of collecting and gathering data on NPS over time to better understand detractors and leverage promoters. Through a NPS Dashboard built on Google Data Studio, they showed an example of how the data can be broken down and organized to get key insights about potential areas of improvement and better understand how different types of users respond to the app.

此外，他们还强调了在构建 NPS 时找到不同类型指标之间正确平衡的重要性，展示了 Google Play 上的一个真实案例。最后，Alice 提醒我们 Apple 在鼓励评分方面非常严格，因此，请小心 iOS 的 NPS 设计调查。

转化和幸福感的触发因素的影响。” src=”https://apptweak-blog.imgix.net/images/2020/05/14/screenshot-2020-05-14-at-15.15.37.png?auto=format,compress&q=75&w=1200″ alt=”触发因素对覆盖范围、烦恼、转化和幸福感有不同的影响。这取决于它是否发生在 NPS 之后，在多少次播放之后。”>
当触发因素发生在 NPS 之后和多次点赞之后时，会产生更好的结果。

5. 小组讨论：应用商店优化与品牌

Louis tanguay – 应用增长峰会
alexandra lamachenka，增长经理 – SkyScanner
catherine sibirko，应用商店优化专家 – Flo health
ian Irving，高级数据分析师 SEO 和应用商店优化 – BBC
mathew Parry，应用商店优化经理 – babbel

Panellists confirmed a very large part of their app installs come from only a few keywords, mostly branded keywords. However, ranking well on your most important generic keywords is still extremely important, as some users are looking for them without a brand in mind. It can be challenging to choose between generic and branded keywords, in particular on iOS where limitations of characters are more restrictive. A good practice is to test and monitor search volumes, rankings and other relevant metrics.

Everything must be aligned across channels but also internally. The ASO team needs to collaborate with product and branding. New features and positioning can influence the app’s title and metadata. During a brand activation campaign, you should also make sure everything is synchronized between channels.

Finally, a question was asked about whether we should attack competitors branded terms? All of them say no because it’s not allowed by the stores, but also because it was not effective when one of them tried some time ago. For well known branded terms, competitors might pay a lot of money but it seems to have a limited impact.

6. 日本的应用商店优化：了解如何打开日本市场。

Adzsa Hatano, ASO Expert at Phiture shared tips and insights for a successful keyword optimization and conversion rate optimization for the Japanese Market. As there can be up to 5 different keywords for one word, she highlighted the importance of researching and collecting all types of written versions of a keyword used in Japanese. Based on her experience, Japanese users tend to search with Hiragana on the App stores so keywords in Hiragana tend to have high Search Popularity. But, she recommends using the formal written version of a keyword in the visible assets (title, subtitle, short description, etc) regardless of Search Popularity. She also added an interesting remark that sometimes Japanese users find English keywords to be cool so it does not necessarily make sense to translate every single keyword.

As for creative assets, she emphasized the importance of understanding the seasonal events in Japan and leverage them by adapting and tweaking icons and screenshots accordingly. Also, she shared some small tips that can make the difference in Japan including the use of Kawaii, animism and text bubbles, as well as highlighting social proof and features in the first 3 lines of the long description. Finally, she focused further on screenshots and explained that successful imported apps have localized theirs by catering content and features to Japanese users, used Japanese influencers and models when possible, and followed category trends such as landscape screenshot for dating apps.

屏幕截图示例。” src=”https://apptweak-blog.imgix.net/images/2020/05/14/screenshot-2020-05-14-at-15.18.15.png?auto=format,compress&q=75&w=1200″ alt=”屏幕截图通常具有高信息密度，它们突出显示社会证明并使用季节性活动。”>
屏幕截图中塞满了许多元素，并突出显示了社会证明。

7. 关于有机提升力量的公开问答

Gabe Kwakyi，联合创始人 – Incipia
Marina Corby，数据分析经理 – Incipia

The term “organic uplift” is often used in the mobile industry. But what does it mean and how to factor it into your ASO strategy? The organic uplift, also known as “k-factor”, represents the number of organic downloads gained from one marginal paid download. It happens mostly hours or days after a paid download but it can be recognized weeks-to-month later. Gabe Kwakyi and Marina Corby explained that there are different types of organic uplifts, divided into two categories: user-level organic uplift, coming from different kinds of brand searches, and ecosystem organic uplift, coming from non-brand search, top chart rank, app list rank or brand awareness.

为了对您的变量（付费安装、App Store 功能、创意）和输出（例如：有机安装）之间的关系进行建模，可以使用不同的统计技术，每种技术都有其自己的可解释性和复杂性。 模型越简单，获取输出并应用它就越容易。 Therefore, in order to grab the low hanging fruits, they recommend to start simple and increase gradually the method complexity based on your data. You should determine your business needs, look at the data and gather all the factors you can. To run tests, ideally, no UA campaigns should be conducted so there is no noise in your data that could mislead the results. A good idea might be to go to a market that has a low and stable baseline, as it will help the algorithm to understand what is happening. Finally, new apps on iOS often have a better organic uplift, as they benefit from high visibility during the first 7 days. They closed by emphasizing that Organic uplift pre-COVID-19 is not applicable anymore and that we should all spend some time to figure out what is the new normal.

Incipia 描述了用户级别或来自生态系统的有机提升。

8. 小组讨论：应用商店优化与 Apple Search Ads 之间的协同作用

主持人：David Llewellyn，EMEA 营销总监 – appsflyer
小组成员：Maryia filimanchuk，首席营销经理 – apalon
小组成员：Quoc Anh luong，移动营销经理 – OMIO
小组成员：Thomas petit，移动增长专家

Many factors must be considered when looking for the best keywords to target, and lots of insights come from Search Ads. Some keywords will have high search traffic, some other high conversion rates or better monetization channels. A good practice is to create different keyword groups to see which ones perform better and where. A winning strategy might be to differentiate and go for low competition keywords.

ASA is not an end in itself, as Apple is not favouriting apps that put a lot of budget on ASA. While Apple will increase rankings on some keywords if download velocity due to ads is high, using ASA specifically is not proven to have an added value compared to other networks. Marketers should find the sweet spot when looking at the relationship between investment and visibility; ASA can help sometimes to bring downloads on some keywords, or it can give you a heads up in new markets. Finally, the golden rule is to target relevant keywords only, those who will give both visibility and retention. Buying traffic for some keywords that are not consistent with the value proposition of the app will not help. Therefore, there should be a synergy between keywords and landing page or screenshots. The best results will be achieved only if ASA and ASO are well integrated, which implies to think broader than just about the keywords.

9. Apple Search Ads (ASA) 堆栈：介绍运行 ASA 的备忘单。

Andrea Raggi, Search Ads Lead at Phiture, highlighted the benefits of Apple Search Ads and provided a cheat-sheet for how to run them effectively. He explained that Search Ads increase visibility, discoverability, and drive downloads as it allows you to advertise your app to 500 million weekly visitors in 60 countries on the Apple App Store. He proceeded by explaining the key terms and sharing a strategic framework with the different layers to consider for developing Apple Search Ads:

1. Campaign: aims at properly organizing your account and helping better leverage the different tasks.
2. Ad group：帮助定制和制定策略。解决特定受众或优化效果，这有助于应用商店优化措施。
3. Keywords: can help reach scale, optimize performance and discover new keywords with the search match. It also helps you find some search terms that work better than others and that can be transferred to your ASO efforts.
4. Tools: Support strategy through the study of the market, technical improvement and understanding of what the competitors are doing.
5. Supporting insights: Enables to grasp the big picture. ASA improve keywords ranking at the organic level.

These 5 supporting layers are the foundation of ASA strategy for which the goals are to help scale the channel and optimize its performance. He finally added that the strategy does not need to cover everything at the same time and recommended to use it to improve the ASO conversion.

Phiture 的战略 Apple Search Ads 框架的 5 层。

注意：于 2020 年 5 月 14 日发布 ASA Stack，旨在为应用商店优化 Stack 社区提供所有可用资产的概述，以在 Apple Search Ads 中制定详尽的策略。

10. 我们如何开始内部本地化团队以进行应用商店优化

VLMedia 的产品合作伙伴专家 Seyma Cam 向我们介绍了他们如何建立一个有效的内部本地化团队，以将他们的应用本地化为 34 种语言。她解释说，他们在决定在哪些地理位置进行本地化时，考虑了 2 个主要因素：

• Where their users were and from where the most downloads and reviews/ratings were coming from.
• Where their competitors were localized and an in-depth analysis of their localization strategy in those markets.

一旦选择了本地化市场，Seyma 解释了其公司为建立内部本地化团队而采取的步骤：

• Find people that speak English and the foreign language targeted and have experience with mobile applications.
• 让每位母语员工负责研究文化和语言趋势，并了解该目标市场中的最佳应用商店优化实践（关键词优化和创意）。
• 开发一种工具，将所有翻译集中在一个地方，包括应用内副本、商店列表和营销。

She stressed the importance of experiments and choosing the right keywords for that specific market based on both research and the different metrics. Also, she explains how localization goes beyond ASO efforts and that the company has also localized promotions, email marketing, paid ads and everything else. Overall, localization allowed her company to increase revenue, retention rate and in-app purchases. It proved to bring positive results in both App Stores as the app, once localized, became more engaging and fun for the users.

转化率、保留率、购买和许多其他指标产生了积极影响。”>
本地化对转化率、保留率、购买和许多其他指标产生了积极影响。

11. 自动化本地化：通过每周发布周期为 40 多种语言创建无缝流程

Ilia kukharev，应用商店优化负责人 – aviasales

Localization has huge benefits, such as better visibility for your app, monetization with Ads, more chances to get featured worldwide or unexpected boosts. However, as you keep developing your app, translating it in many languages on a regular basis can be extremely time-consuming. Ilia Kukharev, Head of ASO at Aviasales explained that by automating the process, the company managed to reduce the time needed for translations from 15% to just 1% of their work time. How? By using by using Crowdin a cloud-based solution that streamlines the localization process. You can set up a project, upload your files, and invite translators to work on them. When new files need to be translated, freelancers or a translation agency will get an email. They need to log in and translate new stings. This way, you can focus on the quality of the translations rather than the logistics of managing them.

应用的更新、翻译、推送和构建的自动化。” src=”https://apptweak-blog.imgix.net/images/2020/05/15/result-1.png?auto=format,compress&q=75&w=1200″ alt=”Aviasales 自动化了翻译过程，并将其任务的工作时间减少了 15。”>
Aviasales 自动化了翻译过程，并将其任务的工作时间减少了 15。

12. 处理 Google Play 实验中的假阳性 A/B 测试结果

Simon thillay，应用商店优化负责人 – AppTweak

我们自己的应用商店优化负责人 Simon Thillay 分享了 Google Play 实验固有的一些问题如何影响测试结果的可靠性和可信度，然后提出了一种旨在提高结果可靠性的新测试方法。

• 无法控制来自不同流量来源的用户的行为以及商店每周的季节性变化，这会产生大量的统计噪音，这些噪音会以不同的方式影响每个 A/B 测试样本中的转化，从而导致假阳性结果。
• 在每个测试样本中获得足够的下载量是达到足够统计功效的关键。
• 在您的测试中复制您的新变体，以便您具有 A / B / B 结构将帮助您比较两个 B 样本是否获得相似或不同的结果。来自两个 B 样本的相似结果很可能是可信的，而在每个 B 样本中发现不同的结果意味着您可能获得了假阳性结果。

AppTweak” src=”https://apptweak-blog.imgix.net/images/2020/05/14/screenshot-2020-05-14-at-17.58.37.png?auto=format,compress&q=75&w=1200″ alt=”我们为新的 A / B / B 测试格式推荐的规则 – AppTweak“>
我们为新的 A / B / B 测试格式推荐的规则 – AppTweak

Simon 最后要求观众中愿意的成员与他联系以分享他们的结果，因为他将尝试汇编一些关于这种新方法的好处的统计数据，这些数据也将在我们的博客上很快详细介绍。

13. iOS 和 Android 上的评分库

Sebastian gallese，创始人 – haystack reviews

众所周知，要求评分非常重要。但是，当您的评分机制是硬编码的时，进行实验会变得很困难。Haystack Reviews 的创始人 Sebastian Gallese 解释说，为了解决这个问题，开发人员可以利用现有的库，这将使他们花费更少的时间来设置所有内容，而花费更多的时间进行实验。例如，可以在 Android 上使用“skent/Amplify”，在 iOS 上使用“Armchair”，或者使用 Flutter 使用“skyost/RateMyApp”。Sebastian 还指出，虽然 Android 尚未提供本机提示，但已经发现了一些时间前包含本机提示机制的测试应用，营销人员应为该功能很快进入商店做好准备，并因此确定最佳触发机制。

我们希望您像我们一样喜欢这次会议！如果您想了解更多关于AppTweak的信息，请随时查看我们的其他博客、预订演示或开始免费试用！