In the process of work, we often encounter decision-making work, and objective data analysis conclusions and mature methodology can provide reference.
Decision-making is required in many places at work, such as product function determination, trial promotion and testing, product pricing, and marketing activity planning. How are these decisions generally made?
The more common entry points are data and methodology, which can be understood by talking about two cases.
Case 1: When discussing a product again, the leader asks whether the A function should be retained or eliminated. The product manager proposes an analysis based on the utilization rate of product functions and the click rate of functional modules: through customer visits and surveys, it is found that the online rate of function A is over 80% in the forefront, indicating that the customer selection needs in the early stage are very strong; but the actual application is the click rate of this module. but the lowest. We need to further investigate why the usage rate of this module is low, whether to further optimize and simplify the use, improve the learning documentation of this function, and guide the application.
This is a typical use of data to speak, according to the trend of the data, to draw conclusions directly.
Case 2:
One day, the leader convened all marketers for a meeting and asked how the product should be promoted (for example, the BI software of FineBI for enterprise users). One of the operators expressed his opinion: since the product has been formed and will be launched to the market, it is necessary to convene users to increase the number of users (similar to UV). To increase the number of users, you can contact channels to find cooperation, and you can also take online publicity to obtain trial customer information. After that, we need to consider the problem of user activity and retention rate, and do it according to this idea.
This is a typical methodology. Through methodology, problems are solved one by one.
The above two cases are both well-founded and well-founded, and the thinking is very clear. In daily work, most of them do the same.
Through one of the above methods, as long as it is clearly stated, the team will basically approve it. But I think that such analysis is often too "nerdy", thinking that as long as you take out the data and come up with a methodology, you can come to a conclusion, but in fact there is no actual human analysis or scene analysis, or such reasoning and argumentation does not It is not rigorous, and the factors considered are not comprehensive.
比如案例一，有市场经理就提出了，其实A功能之所以上线率很高但是使用率很低的原因，大部分是因为销售在向客户讲解产品功能的时候，把A功能作为亮点和卖点来吹捧，A功能确实很吸引人，也很“花里胡哨”，但实际场景是，技术人员需要花很多的时间去利用这个功能开发，但实际的利用率很低，后续维护成本也很高。
所以，这里认为数据/方法论=结论，其实有误区的，数据并不是全部，不能全依仗数据来说事儿，忽略人依靠业务经验对问题的洞察。应该是通过数据或方法论后，在结合人的实际分析（或场景分析），最后再得出合理化的结论。即：数据/方法论+人为分析=结论。
在我们平常工作中，强调数据事实，作为分析依据固然很重要，但这只是一部分，而不是全部。在实际的业务分析时，依靠业务经验的分析洞察往往更重要，不能到领导问你什么什么问题，你说等一下我去找一下数据。有时候太过追求这些并不是什么好事。
恰逢最近在看《从0-1》，越发觉得好的产品，好的策略不是依靠既有的事实或者通用的规律挖掘出来的，数据的大部分应用也是在对历史数据做分析，总结过去，统计规律。好的创意好的营销玩法，往往有时候是对人性的探索，对某些哲学理念的看透。
当然，这里不是在说数据或方法论不重要，他只是作为决策的一部分。这也是为什么会有数据挖掘学科，去挖掘深层规律，预测未来趋势。也为什么会有很多厂商、数据分析领域的人去“呼吁”业务人员去自己数据分析，尝试使用一些轻型的BI工具如FineBI去把玩手中的业务数据，切换维度做自主的分析，基于业务理解。
The above ideas are based on individuals, and may not be the same attitude from the perspective of the industry and enterprises. For example, why many companies spend a lot of money, manpower, and time to build a data analysis platform system. This system contains all aspects of business data, and also provides the basis for thousands of large and small decisions. Overall, it is to improve the accuracy of overall decision-making and prevent risks.