How Rheo Makes Data-Driven Product Decisions Using Real-Time Dashboards [Case Study]

想象一下，您正在旅行，经过12小时的会议和电子邮件后，您返回酒店房间。

You flop onto your bed, and flick on the TV to relax before falling asleep and doing it all again the next day.

There is one issue- unless your favorite show is on at the time, you will find yourself idly thumbing through commercials and re-runs that you don’t even want to see in the first place!

This is the problem that Alan Cannistraro set out to solve by foundingRheo在2015年。他没有在想找到娱乐的时候随机搜索电视或YouTube，而是觉得您应该能够根据自己的口味和心情为您提供视频。

当然，他们面前面临一个重大挑战 - 当人们感到不耐烦，懒惰或需要找到令人愉悦的选择时，人们会使用自己的产品。

想象一下，您每天都有20分钟的公共汽车来上班。您想充分利用这段时间，但是您也不想在早上8点花费太多的心理精力！

This is where Rheo’s product team comes in. They need to make Rheo easy to use, but also smart enough that it will reliably serve the right videos at the right time. Otherwise, people will just go back to roaming YouTube.

由于他们具有积极的增长目标，因此RHEO产品团队监视了他们的实时数据TV，，，，desktopand移动的。他们每周衡量改进，如果转换漏斗中出现问题，他们可以立即开始寻找解决方案。这是团队如何将数据集成到他们的日常活动中。

公司

Rheo成立于2016年，是一款Apple TV应用程序，用于观看适合您目前心情的视频。您可以选择自己的心情，而不是扫描频道或搜索演出，而该应用程序将开始流式传输片段2-10分钟。就像潘多拉（Pandora）一样，您使用该应用程序越多，它就越好了解您的口味。它结合了人类和机器学习，为您提供最相关的视频。

2016年，该公司筹集了230万美元在风险投资融资中，2017年，扩展到网络和iOS。

People use their apps in a variety of settings:

1. 上班上班
2. 在用餐期间
3. 在寻求环境娱乐的聚会上
4. Travelling for business

Their channels are called Laugh, Inform, Learn, Taste, Spark, Move and Chill. For example, Inform carries the latest news, Move has extreme sports footage, and Laugh might have clips from late-night talk show hosts.

Rheo is a tech company creating a media experience for always-on screens. They compete with the traditional practice of flipping through videos or channels until something that fits the mood pops up. By prioritizing the user’s habits, tastes and feelings up front, they eliminate the search process and get right to the content.

The Challenge

Rheo uses a series of tools to analyze usage of their product. All of these tools prioritize data collection, and not data communication or analysis.

这些工具在跟踪Rheo应用程序中的数据方面做得很好。但是，在显示数据或组合多个数字服务的指标时，它们并不灵活。每个用户界面（UI）都是不同的，需要不同的分析。

管道对于Rheo的内部测量非常重要。当管道的不同部分来自不同的数据源时，产品团队将无法快速查看其指标的健康状况。在他们开始使用之前业务仪表板，他们将在每次会议中都在Chrome中提出多个选项卡，并一一进行服务。

同样,如果他们的首席执行官在一个投资者会议上,他could only use graphs within his slide deck. If potential investors asked for further details on any metric, he would not be able to reply immediately.

他们如何使用实时仪表板

RHEO主要在其业务仪表板中使用以下数据源：

Localytics

谷歌分析

iTunes连接

mailchimp

sendgrid

社交媒体（Facebook，Twitter）

自定义脚本/ Databox Abeplay体育appiosPI

They stay on top of their data via these methods:

Growth meetings 3x/week– The product team meets three times a week to make sure they are retaining new cohorts of users at higher rates than earlier cohorts. Users are broken up into iPhone users and Apple TV users since Localytics uses a different app ID for each. They display a series of性能仪表板连续确保每个漏斗都可以满足或超出期望。

例如，他们使用Google Analytics（分析）和iTunes Connect来创建与应用程序下载比率的页面浏览量。当他们在App Store中调整页面时，他们测量副本更改是否会影响该比例。

Weekly emails-产品团队成员每周收到一封电子邮件，与过去一周相比，KPI的变化。

Automated alerts-该团队已对其数据中的重大事件进行了定制通知。例如，如果转换率从一个漏斗中的悬崖上掉下来，可能会有一个错误。团队成员从中获得实时通知业务仪表板软件。或者，如果下载异常高，则可以调查原因。

他们监视的关键指标

Rheo uses metrics that are typical of online media companies:

• Daily Active Users (DAU)
• Monthly Active Users (MAU)
• dau/mau比率
• Time Spent per User per Day (their own calculations)
• Time Spent per User per Month (their own calculations)
• 观看用户的视频
• 观看了总视频
• Total Clips Played per Day/Week
• 整个平台上花费的总时间
• 28 day Rolling Churn

需要明确的是，“ 28天滚动”回答了一个问题：“过去28天，我们的用户中有多少百分比？”因此，如果您在八月的应用程序上有100个活跃帐户，那么9月有多少个不活动？该百分比与前28天相比如何？

搅动用户的定义是相对于应用程序类型的，但是在像Rheo这样的应用程序的情况下，搅动的用户可能会算作持续14天没有登录的人。

RHEO产品经理Gautam Banerjee说：

“保留是硅谷的国王。下载仅如果您的应用程序粘稠。”

因此，RHEO团队使用上述所有指标分析保留率。有些可以在数据库中显示，而另一些则无法通过Locabeplay体育appioslytics API提供。这些必须在本地学中进行分析。

Lessons Learned on Running a Data-Driven Team

这是了解Rheo产品团队使用的策略后，给我留下了深刻的印象：

大规模学习用户习惯：两个RHEO应用程序具有完全独立的用例。这创建了产品团队需要计划的两个不同的用户模式。这意味着两个具有单独基准的管道用于转换率。

On one hand, you might think that the Rheo team needs to track double the data compared to other startups. But, plenty of sites and apps have this challenge. For example, if you have a freemium SaaS app, you might have two profiles of users: the free/low-paying ones, and the ones with larger demands and budgets. Each will have separate needs from your SaaS product.

Ease of access to data determines level of attention:如果没有Databox的帮助beplay体育appios仪表板设计师tool. If they could not aggregate different sources into one place, the team would either meet less often or data extraction would become somebody’s part-time job.

保留是国王：If you want to quickly grow your company, you must focus on retention first. No amount of acquisition will help you overcome a leaky funnel. The Rheo product team measures at least 8 retention metrics simultaneously so they can discover which actions in their app create ongoing habits. Then, they can make more data-driven decisions on features and conversion points within the app.