About the survival of growth products
increase! increase! increase! Business growth is the biggest problem faced by entrepreneurs every day. Whether your product is APP, or web, or applets, can only continue to maintain subscriber growth, in order to tell a good story to the capital markets, financing survive. Live to the final product, have a chance to profit.
In order to obtain subscriber growth, advertising can also be content marketing, social communication, sales push, or sustained focus on product optimization. Either way, we are faced with these questions:
- Operational activities, covering a number of users?
- How many users start using the product?
- How many users pay?
- How many users continued active?
- Next, we should focus on what areas? It is continuing operations? Or develop new features?
If you can not answer these questions, no doubt our operational activities or development is Mangrenmoxiang relies entirely on luck. To answer these questions, we may wish to look at means of this data-driven growth hacking.
Growth Hacker core idea
Traditional marketing strategies, such as running a TV ad covers how many people, how many people made a purchase after seeing the ad, how many people were shopping complex, there is no accurate data to measure, can only rely on the senior expert judgment based on experience . In the Internet industry, every product is new, unprecedented. Each product can not survive, how each operating results, do not have much experience to draw on, the results are uncertain.
GrowthHacking the rise in Silicon Valley start-up companies marketing tool, it is designed to use a small amount of the budget to get a huge amount of growth. Because of its high cost and effectiveness, it is ideal for start-up companies, which have been widely disseminated.
Growth Hacker's core idea is to index data, drive operational decisions, and optimize the product. Growthacker acquired by focusing on users, conversion, user retention, user recommendations, a series of core earnings and other indicators, as well as through various dimensions disassemble, analyze the next growth decisions. By Growth Hacking, create a closed-loop product growth strategy.
So how can we build a GrowthHacking architecture for their own products enabling it?
GrowthHacking of architecture
Growth Hacking 包含了数据的采集、存储、分析、报表、A/B test等系统，首先我们来看，传统的解决方案，搭建出GrowthHacking有哪些痛点：
搭建运营体系的痛点
搭建运营体系的过程中，常常面临以下问题：
- 缺少数据，数据散落在各个地方,有的是app数据，有的是web数据，有的是小程序数据，没有一个统一的架构来把数据采集到一个地方。
- 缺少一个分析平台。传统的策略，需要运维团队帮助搭建hadoop集群，需要专门团队持久运维。
- 离线跑报表，一晚上才能拿到一次结果，周期太长。手工跑一次，几个小时过去了，有什么新的想法，不能及时验证。严重影响运营效率。
借助云服务搭建的GrowthHacking技术架构
为了解决以上问题， 日志服务提供了日志采集、存储、交互分析、可视化的一整套基础设施，可以帮助用户快速搭建出来灵活易用的Growthing Hacking的技术架构，每天的工作只需要专注于运营分析即可。
Growth Hacking首先从数据采集开始，定义清楚要采集的日志内容、格式。把各个终端、服务器的日志集中采集到云端的日志服务。后续通过日志服务提供的SQL实时分析功能，交互式的分析。定义一些常规报表，每日打开报表自动计算最新结果，也可以定义报告，自动发送最新报表。全部功能参考用户手册
此外，除了日志数据的分析，还可以为用户定义一些标签，存储在rds中，通过rds和日志的联合分析，挖掘不同标签对应的指标。
日志服务有如下特点：
- 免运维：一次完成数据的埋点、数据接入，之后只需专注于运营分析即可，无需专门的运维团队。
- 实时性：用SQL实时计算，秒级响应。快人一步得到分析结果。
- 灵活性：任意调整SQL，实时获取结果，非常适合交互式分析。
- 弹性：遇到运营活动，流量突然暴涨，动动手指快速扩容。
- Cost: Common market analysis products, the use of packaging prices and limiting usage. Log service according to the amount paid, cheaper, and more powerful.
This log data collected by means of service delivery, storage and analysis infrastructure. Operators can re freed from the heavy data preparation work out, concentrate on using SQL to analyze data, configuration reporting, verification operation idea.
Began to build GrowthHacking system
Specifically, Growth Hacking architecture can be split as follows:
-
data collection
- Buried defined specifications, event content is defined to be collected, field format.
- By Android SDK , iOS SDK , Web Tracking and other means to the client Buried.
-
storage
- Select the region logging services.
- The definition of each log storage Project & LogStore.
-
analysis
- Analysis of the open road, the definition of regular reports, or interactive analysis.
- By analyzing the results, adjusted operating strategy, targeted to optimize the product.
Log-based service, you can complete Growth Hacking analysis strategy:
- Polaris defined indicators.
- Pull a new analysis.
- Retained analysis.
- Event analysis.
- Funnel Analytics.
- A user clustering.
- A/B test。
Log service, by defining a series of dashboards, the results of data analysis to precipitate. The next few articles will be followed by several strategies on how to achieve the above in a log service.
to sum up
This paper describes the overall architecture of Growth Hacking, after a series of articles will introduce step by step how to intervene data, how to analyze data.