What core data should product and operations personnel pay attention to in Internet projects?
|
1. Definition of Data Data is actually a bunch of numerical values. But these values are derived from user behavior statistics. Basic materials used to facilitate research and analysis by students who need to use data. 2. What data should we pay attention to? Here we use a mind map to simply list some core data: Core Data This mind map simply shows some common operational data, but if we look closely and summarize, we will find three types of data that all operations need: Channel data, cost data, and revenue data. Channel data is used to measure channel quality and channel effectiveness, which is determined by the target customer group and product characteristics of the product itself. If a financial product is launched in a channel such as a game community, its operating effect may not be very good, but if it is replaced with lottery or gambling, the effect may be very good; similarly, if the promotion and activities of games such as Legend are launched on a female community platform, the effect can be almost ignored, but if it is replaced with a Q-version mini-game, the effect may be very good. Cost data and revenue data will reflect the effectiveness of operations from different levels. Operations inevitably require costs. Operational efficiency can be improved through experience, proficiency, creativity and other means, but operational costs are inevitable and are generally proportional to operational results. It’s a very simple truth: there are two activities, one activity gives away 100 iPhone 6s, and the other activity gives away 1 iPhone 6, which one is better? Those who are engaged in operations must carefully evaluate the costs behind each operational action. The so-called "benefit" is not equivalent to "income". Gaining money is benefit, gaining users is benefit, and gaining word of mouth is also benefit. If we understand the three types of data: channels, costs, and revenues, we can set the data we need to obtain based on the characteristics of our own products. Let’s take Zuji as an example. What data does an application like Zuji focus on? From the product level, it will focus on: 1. Number of times the app is opened daily 2. Number of times each function is used and how often it is used 3. Number of times each tab is clicked and how often the corresponding pages are opened From an operational perspective, it may focus on: 1. The number of active users of the App every day 2. The amount of UGC generated every day (distinguishing between new and old users) 3. The amount of UGC shared to social media every day (while taking into account the amount of content generated by each user) 4. The number of new installations and newly activated users brought back by the shared UGC, etc. What we need to note is that these data points are not static. They will be adjusted according to the different stages of the product. If we assume that Zuji has a profit model, then the core data it focuses on will shift from content to revenue. At this time, conversion rate -related data will become important. The data we need is designed according to actual needs, and there is no completely universal standard. Of course, the more you do it, the more you will find that your sense of data has improved unknowingly, which is very important. 3. How to obtain data There are many channels to obtain data. Here are two recommended: Google Analytics and Baidu Statistics. You can learn more about more tools on your own. Using the analysis tool we can get the following: Record click information, including the percentage of links that can be directly generated without interacting with the website . Click distribution charts and heat maps can count user hovering and visualize user potential behavior. There are actually many ways to obtain data. The key is that as an operator, you must understand what kind of data is important and the relationship between these data. This is an interconnected process, not a single behavior. 4. How to analyze data Everyone has a different way of interpreting data. Here are a few summaries from the book: 1) First determine the accuracy of the data This includes the rationality of selecting data dimensions and the accuracy of data statistics. If the data dimension selection is unreasonable and the data statistical results are inaccurate, we may not be able to obtain correct analysis results. This is the foundation. 2) Identify the factors that affect the data A piece of data will be affected by many factors, both internal and external. Operations personnel should understand as many influencing factors as possible at all levels to facilitate our interpretation of the data within a relatively correct range. 3) Pay attention to long-term data monitoring In operational data analysis , month-on-month and year-on-year methods are often used to compare data. Simply put, the month-on-month comparison is the comparison between today and the previous day, the comparison between this month and last month, and the comparison between this quarter and last quarter; the year-on-year comparison is the comparison between the same day this year and the same day last year, the comparison between the same month this year and the same month last year, and the comparison between the current quarter this year and the current quarter last year. The month-on-month comparison helps us see short-term data fluctuations, while the year-on-year comparison helps us understand data fluctuations in the overall environment. 4) Maintain an objective perspective In the process of data analysis, objectivity is very important, and you should avoid choosing conclusions that are beneficial to yourself. We often make preconceived mistakes, which not only affects the accuracy of the data, but is also a problem of professional ethics and a very common problem in career development. 5) Be careful to eliminate distractions In actual work, we will encounter many problems, and these problems are interference factors. For example, in a relatively stable curve, a strong fluctuation suddenly appears at a point. At this time, we need to fully understand the cause of the fluctuation. If the cause cannot be confirmed, we should eliminate the fluctuation. Otherwise, it will be difficult for us to obtain a correct conclusion. 5. What are the career directions of data analysis? We are only talking about China here. in the country, positions related to big data are mainly divided into the following categories: Data Analyst: Using tools to extract, analyze, and present data and realize the commercial significance of data requires business understanding and tool application capabilities. Data Miner/Algorithm Engineer: Data modeling, machine learning, and algorithm implementation require business understanding, familiarity with algorithms, and proficiency in computer programming. Big Data Engineer: Using programming languages to implement data platform and data pipeline development requires computer programming skills. Data Architect: Advanced algorithm design and optimization; data-related system design and optimization. Vertical industry experience is preferred. Platform-level development and architecture design capabilities are required. Of course, there are definitely more than just these few positions in an industry. Here we only list the common and popular ones (you know~). VI. Conclusion Regarding the content of the data, an article cannot actually say much valuable things. It requires more thinking and summarizing from one's own work; of course, it is also important to communicate with peers and learn from the experience of predecessors in the same industry. Mobile application product promotion service: APP promotion service Qinggua Media advertising The author of this article is @Pao Ding Development Compiled and published by (APP Top Promotion), please indicate the author information and source when reprinting! |
<<: Case study: Steps to set up a Kuaishou information flow advertising account
>>: How many keywords should be placed in each unit of Baidu promotion?
Recommend
How to add words after tap on WeChat? Tutorial on adding text after tap
How can I add longer text after tapping on WeChat...
How to do user recall? 10 strategies!
In fact, whether it is recalling users who have a...
Xiaohongshu’s promotion strategies and methods!
Speaking of Xiaohongshu, what is the first thing ...
How did Momo gain tens of millions of users in just one year?
In June 2011, Boss Wang (Wang Li called himself B...
How to Write an Integrated Marketing Communications Plan (Version 4.0)
1. Those who do not consider the overall situatio...
Baidu promotion information flow video advertising display style - large picture video style
Style Overview Presentation location 1: Baidu APP...
What does it mean when a website is K? What to do if your website is blocked?
What does it mean when a website is K? What to do ...
How to improve user stickiness for Internet financial products?
In recent years, Internet finance has become very...
Free Qianjiang Road March 2022 Courses
Free Qianjiang Road March 2022 course resources i...
2020 China Advertising Agency Map
Recently, independent marketing consulting compan...
How to write an event plan? 4 basic elements to prioritize
The event background, purpose, theme, and time ar...
What is Server Load Balancing and how much does it cost to rent one?
What is Server Load Balancing and how much does i...
How to quickly build a marketing and promotion system for B2B products?
In the past two years, the SAAS product market ha...
Methodology for operating top short video influencers!
Short videos in the field of magic also attract p...
Why is no one buying products sold on Douyin live streaming?
During the 2019 Double 11 shopping festival, Taob...