Why do number lazy
Business demands
If you are in business, you are bound to encounter these problems
- Design drives business, where is the power point? How? How does the business agree with your proposal?
- The business thinks that there is a problem with a function point, and there is a conflict, how to convince the business rationally?
- What is the relationship between data growth on a page and design or some function points? Is there a causal link?
- How is the data processed, how is the data collected analyzed, and what data does the business’s Polaris metric relate to
- .
As the front end of the business, what I deal with most in my work is a variety of activities every year, and for activities, the most important thing is data.
The so-called “data” is raw material, but we need “information”, information is the basis for decision-making after analysis and processing. In the era of big data, data analysis ability is the most important decision-making basis for each growth team to improve user growth. From data, we can summarize the effects and shortcomings of activities to provide reliable support for interactive experience, visual design and business operation strategy.
However, in the past, we used to analyze data by relying on manpower to raise numbers, clean, analyze, draw and finally make analysis reports. The process is very tedious and extremely costly. As the saying goes, “if you want to do a good job, you must first improve your tools”, the business side urgently needs a tool that can set data analysis and visualization in one.
Existing solutions
There are already visual data tools on the market, and many of you are bound to wonder, “Why not use them? Would making another tool duplicate the wheel?”
So we did a round of research on internal and external data tools. The survey results show that most of the current tools can only provide some basic data display functions, as for visualization, how to use the data, how to analyze the data, there is basically a blank. Some tools have some simple comparison of indicators, but can not draw conclusions after comparison, still need to be handled by users, and the business side reported that these tools are basically unable to help them. And those commercial tools, most need to be expensive to use.
So we wanted to make a data analysis tool that was “good for business.”
Process and dismantling
First, we need to know exactly what our analysis process looks like. For ease of understanding, we split the whole process into several stages according to the life cycle of data analysis: before analysis, during analysis, and after analysis, through which we implement specific requirements one by one.
Analysis of the former
“The data quality is poor, the caliber is different, processing the data is really time-consuming…”
Before analysis, there are mainly the following pain points:
- Data caliber is not unified, processing trouble
- Large amount of data, inefficient
The business side has many data raising platforms, and the data formats and indicators from different platforms are different, which leads to the low efficiency of the business side in data processing. For example, there are nearly 200 items in the Double Eleven activity, and the data cleaning alone has reached 25 days. Usually, the data cleaning is finished, and the activity is over. Moreover, the data indicators of some projects are particularly different, which must be handled manually once.
Therefore, we need a cleaning tool that can uniformly clean tables from different sources into the same format for analysis and store the original data into the database.
The main process of data cleaning:
- All data sources are smoothed out in a standard format
- Parse page, floor, pit data
- Error checking
- Database entry
In the analysis
“How do you analyze the data? Am I going to study statistics?”
After obtaining the data, everything is about “how to analyze the data”. However, the professional data analysis process is very complicated. Even ABtest, which is often mentioned, is not a skill that ordinary people can master. Therefore, if we want to automate analysis, we must deconstruct the process of data analysis:
- Goal setting
- Index analysis
- Identify problems/advantages
- Draw conclusions/verify previous conclusions
After the analysis of
“There are too many analysis reports, retrieval is troublesome, writing reports is tired…”
At the end of the analysis, we need to collect, classify and store all the analysis results, and provide retrieval functions, which can directly extract reports through keywords or items. The “best” business feedback has report generation function.
A few lazy
Eventually, lazy counting was born. Lazy is a visual data analysis tools, is different from past data tools, a few lazy on the basis of the visual display, to subdivide the dimensionality of data from the project to the floor, pit, and provides the data cleaning, strategies, validation, conclusions export function, one-time data analysis through the whole process, provide a complete solution to the demand side.
Analysis process
1. Clean data
The sloth provides a visual desktop cleaning tool. Users can use the tool to clean data and verify data errors to obtain high-quality and standard data formats. The data is then uploaded directly to the computer for storage and retrieval.
2. Data analysis
A. Page large number analysis:
Large number is the most core index in data, refers to PV, UV, click rate, conversion rate and so on. These data can intuitively show the basic situation of the analysis target. By analyzing the large number of pages, the number cruncher can provide some indicator of a page’s problems, such as its effectiveness, revenue, or appeal.
The numerical lazy supports the simultaneous analysis of two indicators, which can discover the law of large numbers and sort out 27 special data scenarios. After automatic judgment, the system will output conclusions and trace the causes.
B. Detailed analysis of floors
Just like Big Numbers, the floor analysis function shows the basic data information of each floor to the user by subdividing the floor ID of the page data, and finally analyzes the data to draw some scene conclusions.
Floor analysis supports 11 special scenarios. The four-quadrant function flatten all floors so that users can intuitively view the distribution area of each floor. When users discover a problem and prepare to trace the source, the core data of the pit in the floor is displayed to help users quickly locate the cause of the problem.
C. User portrait
As the name implies, user portrait is to provide user characteristics of each activity to help users understand the situation of users, so that in business, corresponding strategies can be pushed according to different groups.
D. Multi-objective comparison
The main change from traditional data tools is that data can be compared between different targets to analyze the relationships between certain activities.
Users can select two targets for comparison and view the index difference between the two active pages. Similarly, similar to Large number, users can find problems and get corresponding suggestions by comparing the two activities.
When multiple targets are compared, the major indexes of all activities to be compared will be displayed in the form of tables.
E. Policy verification
Policy validation is the core function of counting Lazy, which is the biggest difference between counting Lazy and other data tools. At present, sloth divides the verification goals into four general directions, and each general direction subdivides the specific goals of different fitting business. After selecting the verification goals, the system will give the final achievement results of the goals, and present the possible causes of data changes and analysis directions.
After verifying the target, it can also disassemble the data for the core verification indicators and display the detailed data. Users can trace the reasons that affect the data by analyzing the listed factors.
F. Report generation/conclusion retrieval
Report retrieval is a time-consuming operation in the daily work of the demanders, who need to search through a lot of folders for past documents. Therefore, Html2Canvas is used to generate images directly from the conclusion of the page. During the generation process, users can freely edit the layout and format of the report, saving a lot of time.
At the same time, users can search past analysis reports by entering project keywords in the conclusion search function.
Count the lazy future
Improve analytical expertise
While sloths currently have some data analysis capabilities, at the end of the day, counting sloths are a solution, not a perfect one. At present, most of the analysis conclusions are limited to the subjective experience of the activities we handle. Many analysis conclusions do not verify the confidence degree, and the number of analyzed data samples is too small, so many strategies cannot become effective conclusions.
As mentioned above, real data analysis requires solid professional knowledge including data and quantitative analysis, which requires a lot of learning before practical application. The validity of the results directly affects the final application effect of the user. As a developer not specializing in data analysis, the core evolution direction of the sloth is to improve the rationality of the analysis process and the validity of the conclusion.
Function of decoupling
“Cloud” is all the future trend of China, many partners can also puts forward several lazy part of analysis function for its research and development platform, suggests many users might actually use of data analysis scenarios are not the same, ABtest some user just want to do, for example, some users want to attribution, but lazy is an independent platform, If all the functions are integrated into one platform, the entire platform will become bloated, not only will the user experience become worse, but the development cost of the entire platform will become higher and higher.
So, remove the function decomposition coupling is an essential step that will be different disassembly analysis function, can adapt to any data format analysis scenarios, will not be limited to the analysis of the data format and target, any platform or user access will only need to select a specific function to use, do not need to depend on the number of lazy whole platform, not only conducive to users, It is also more conducive to the development of numeracy laziness.
At present, the ABtest function has been decoupled and other platforms have been connected to start the analysis work. The ABtest function can compare activity indicators of multiple analysis objects to obtain experimental results and provide evidence for users’ subsequent strategies.
More policy
The most important thing to do data analysis is the analysis method. Currently, there are still not enough analysis strategies. In the future, we will build a strategy library through accumulation, output data prediction and excellent strategy suggestions externally, and input the objectives and specific plans of iteration internally, so as to finally provide analysis ability for more businesses.
Write in the last
In an era of increasingly expensive traffic, maintaining growth efficiently and cost-effectively is a core goal for all business teams, and this can’t be done without data analytics. Without data, every thought is “I think”, “should be”, every attempt is trial and error, so the importance of data analysis for any member of a growth team is self-evident. Man is not everything, “knife and cutting wood workers”, the proper use tools to support data analysis can effectively improve work efficiency, and good tools can improve the operational results of the final, and this is the target number of lazy, few lazy will strive to become all users “simple” “fast” “with” little helper.
Concave-convex reveal series article address
1. Past and Future of Bump Laboratory
2. “Concave and convex Technology Revealed · Linglong Intelligent Design Platform · Dream Oriented Design”
3. “Concave and convex Technology Revealed · Deco Intelligent Code · Start the Revolution of Production and Research Efficiency”
4. “The Secrets of Concave-convex Technology · Visualization of Linglong Page · The Road to Growth and Transformation”
5. “Concave and convex Technology Revealed · Quark Design Assets · Creating high-quality Materials of the whole Matrix”
6. “Concave and convex Technology Revealed · Tide R&D Platform · Layout of NEW R&D infrastructure”
7. Revelation of Concave and Convex Technology · Taro · From Span end to Open Span Frame
8. Revelation of Concave-convex Technology · Basic Service System · Building Technology Center of service End
9. “The Secrets of Concave-convex Technology, Technical Improvement and Business Development”
10. “Concave and convex Technology Revealed, Number lazy, growth team good helper
other
Thank you for your attention to concave-convex laboratory readers. In order to provide better content, I hope you can spare a few minutes to complete a small survey. Next year, the content of concave-convex public account is up to you. Click on the direct
Join the concave-Convex Laboratory open, open source, professional, loving and dream family? Click on the direct
For those readers who haven’t followed bump Lab, please follow us. We only push 4 times a month. We cherish every push and won’t let you down. Wechat search “bump lab” can be followed.